NO136159B - DOOR HOLDER DEVICE. - Google Patents

Abstract

Door holder device.

Description

Procedure for coating electrical circuit elements with a single-phase class material.

The present invention relates to a

new species of glassy material. This material

ale has certain physical and chemical properties that point to a number of areas of application. An important area is when encapsulating

ling of electrical circuit elements and interconnected larger units of such elements

ter.

Certain new types of glass have emerged

met by experiments, e.g. the arsenic-sulphur-thallium glass. This glass differs from most ordinary glasses in that it has a very low "melting point"; Typical glasses in this system have a viscosity of 30-

pois at a temperature of approx. 125° C. The

These properties indicate that the material can be used as an encapsulation material for dip-coating of semiconductors that are otherwise destroyed electrically or physically at the melting point for the usual glass types. These glasses can, due to their properties, be used with particularly good results when encapsulating circuit elements and thereby improve the elements' electrical properties.

It has been established that there, as a result

of the process that is generally attributed

ves getter effect, it will be achieved that semiconductor devices encapsulated in such a material are improved with respect to leakage current at a reversed voltage; this occurs either shortly after immersion or after a heat or effect maturation treatment. Extensive investigations of the aforementioned glass material have not shown any deviations from this result. Continuous heat or effect hardening treatments for the glass-coated an-

orders have led to further improvements

gives.

Although these glass materials seem to be a solution to most encapsulation problems in this connection, it has been established that they are affected by a min-

dre disadvantage: the glasses do not show any undesirable properties in the temperature ranges at which most devices are used;

the coefficient of expansion for the glass and most metallic or semiconducting materials

however, they do not match and as a result cracks can occur at very low temperatures. In devices which have been encapsulated in this way and subjected to cyclic temperature tests with a low specific temperature of — 40° C, small cracks will appear in most compositions before forty cycles are completed. It is of course unnecessary to state that the glasses do not lose their usefulness as a result of this property. As the electrical properties are excellent, the glasses can also be used from an economic point of view, even when

des a metal or plastic container as an outer capsule. From an economic point of view, such enclosed devices should only be used when the external temperatures become so

low as mentioned.

As it is desirable to find a glass composition that has the properties of the arsenic-sulphur-thallium material without producing heat-shock effects at very low temperatures,

scientific investigations have been carried out in the hope of finding an additive material that could overcome this

oddity. These investigations have led to the mixtures which are the object of the present invention.

According to the invention, it has been established that halogens, iodine and bromine have largely the same effect on the arsenic-sulphur glass as thallium, which effect was thought to be associated only with this material. As a result, it has been established that the elements arsenic-sulphur with iodine or bromine form single-phase glasses within certain critical composition ranges. It has further been found that these compositions are soluble in the previously mentioned vitreous material. A number of solvent materials which have been intentionally added have had one or another effect on the physical and chemical properties of the iodine system according to the invention. Without modifications, glass in this new system has viscosities of 30 pois in the range from approx. 50° C to approx. 400° C. The areas where the iodine and bromine glasses become softer are noteworthy, as the iodine glass's softening point occurs from approx. 17° C up to around 200° C. The bromine system's softening point can be even lower.

As mentioned, the properties of the glass materials can be changed by adding specific materials. It is e.g. It has been established that a small amount of lead increases the plastic yield point at the same time as the softening point is raised to a much lesser extent. An addition of antimony or partial replacement of the arsenic with this has a similar effect. If part of the sulfur is replaced with selenium, the wetting properties of the materials on ceramic or glazed surfaces of other materials are increased. The effects that have been ascertained from the addition of other materials shall be specified in more detail below.

As mentioned earlier, the thallium glasses have a marked effect on the electrical properties of the devices on which they are placed. These results have been associated with the getter effect, and some evidence for this has been found. In any case, it is clear that the thallium glasses somehow bind ionic impurities on the surface of electrical devices, which would otherwise tend to move under the influence of an external or internal field, so that the electrical characteristics at the same time would change. Such a change in the electrical properties is of course not desirable, and in the case of semiconductor devices, this has been a significant source of error. in

It has been established that the glassy materials according to the invention have

a similar effect on electrical elements that come into contact with these materials. It is a fact, which is clearly evident from the data to be given, that the electrical properties of encapsulated devices are improved to an even more pronounced degree when the materials according to the invention are used in the earliest manufacturing steps. As a result, leakage current can be reduced by one order of magnitude or more in the case of semi-conducting transducer devices that are applied with an opposite bias, simply by immersing them in a liquid composition of this nature; this is clear from the following tables. The increased degree of getter effect which occurs with ■ the material according to the invention in relation to the previously known, is believed to be connected with a greater flow property of the new material. The getter effect generally seems to be so great that devices with a suitable encapsulation are improved to such an extent that the electrical properties are better than any that have previously been

achieved. There have been data which seem to indicate that the electrical properties of certain devices are for the first time limited by the body as such rather than by breakdowns on the surface.

Although the encapsulation usually results in properties that are otherwise rarely encountered, there are cases where further improvements occur by effect curing or by what has been found to be its full equivalent, namely heat curing. As a result, it has become clear that the very small number of devices that do not follow the usual tendency for improvements that similar devices show when treated in the same bath, can be given these improvements by a heat maturation treatment for a shorter time.

It has been stated that the encapsulation is carried out by immersion. This is of course a simple laboratory procedure to achieve faster results and without the need to use complicated process equipment. It is to be expected, however, that when producing or immersing elements on a larger scale, methods based on evaporation deposition can be used or that the material has been shaped into units in advance. An enclosure using ready-made material, which will be described in more detail with reference to figures 7A and 7B, can be used in connection with a simple cylindrical part of a material according to the invention which is first placed around one or more wires and which is then brought to flow out, so that all or part of the device is encapsulated at an increased temperature. Vapor depositions can be used, not only in connection with the coating of individual elements, but also to provide protective layers on larger interconnected units consisting of such elements. As an example of such larger units, printed circuits and printed circuit boards can be mentioned. In this connection, it has been found that the vitreous material according to the invention can be placed by evaporation both on hot and cold elements. Coatings placed in this way have been examined, and it has been established that the glass materials according to the invention form a stable glass coating in the same way as when they are placed on the elements by other methods.

The properties that have been achieved in the experiments have primarily been used in connection with semi-conducting transducer devices. Many of the experiments that were carried out were done in connection with silicone devices, where wires were attached by thermal compression. The experiments were carried out on this class of materials as the materials' sensitivity to contamination on the surface with regard to the electrical properties is marked. However, it is well known that other electrical elements often change their electrical properties for the same reason. It is e.g. known that the resistance value of resistors changes with time. This change is attributed to ionic impurities, particularly produced under the influence of an electric field. Similar properties are found in other devices such as e.g. capacitors, inductors and the like.

It thus appears that electrical elements of all kinds can be encapsulated or otherwise brought into contact with materials according to the present invention in the desired manner. Even where stabilization of such elements is not required by binding the ions, the unbreakable moisture barrier that occurs will be a sufficient reason for the material to be used. The fact that the materials in the following will mostly be discussed in connection with semi-conducting transducer devices is only for the sake of simplicity and must only be taken as an example.

The term "encapsulation" is used in this connection in a fairly general way; although it is often stated that the glass coating can serve as an outer enclosure between the element's surface and the surrounding atmosphere, there are cases where other or additional enclosures should also be used. This can be justified by a desire for greater mechanical strength and the materials according to the invention can thus be encapsulated by other means, such as e.g. metal cans.

In addition to the fact that these materials can be used as the final seal, they can also be used as a cleaning agent which is then removed. In this connection, the expensive and more complicated vacuum baking treatment that is now used before the devices are encapsulated in a box can be replaced by bringing the device into contact with a molten, preferably amorphous composition that is as liquid as possible.

It will be clear to those skilled in the art that the term "encapsulation" even in its strictest sense need not mean total encapsulation, as many elements only require that a smaller part which is sensitive to surface effects be encapsulated. Such a sensitive surface can consist of the area determined by a small number of minority-carrying diffusion lengths of an active connection or by the contact part between an element and an electrical connection. As a result, it may be necessary or desirable to cover only one surface or part of a surface of a particular device. Such a limited coating can be carried out by vapor deposition through a mask using known methods. The glass according to the invention wets all the metals that have been tested. Although these materials, like other glass compositions, have a coefficient of expansion that is significantly greater than the coefficient for metals and semi-conducting materials, so that a thermal inequality occurs, this inequality is compensated by the properties that can be associated with the low softening point of the glasses, whereby all the glasses that have been examined have undergone periodic heat treatment between - 40° C and 110° C without cracks occurring. At a single period of the test for the coated devices, the treatment was started at room temperature; the temperature was then reduced to — 40° C. and then raised to 110° C., after which it was again lowered to room temperature. The rate of temperature change was constant over the entire period and the total time was 2 hours per period. This period was repeated at least 30 times.

The invention shall be described in more detail with reference to the drawing, where fig. 1 shows a close-up composition diagram of the glass regions for compositions of the arsenic-sulphur-iodine system, fig. 2 a 30-point viscosity diagram of the glass system according to fig. 1, fig. 3 a ternary composition diagram according to fig. 1, which shows the softening temperatures for certain compositions in the indicated glass-forming region, fig. 4A, 4B and 4C are schematic front views of a typical semiconductor transducer device which is re-encapsulated by a method according to the invention, fig. 5A, 5B and 5C are schematic front views of the same kind of transducer device, this being encapsulated in an alternative way according to the invention; fig. 6 shows a schematic front view of a type of apparatus which has been found suitable for vapor coating with a vitreous material according to the invention, fig. 7A is a perspective view of a semiconductor device and a previously formed part of a material according to the invention before it is heated; fig. 7B shows a perspective view according to fig. 7A after the heating, and fig. 8 a diagram between leakage current and time, where the improvements in the characteristics of nine devices that have been encapsulated in a glass composition according to the invention appear.

Fig. 1 shows a ternary composition diagram of the arsenic-sulphur-iodine system. The area within the straight lines between points 1, 2, 3 and 4 indicates the areas where materials in this system give a single-phase, glassy material. The area within the straight lines 5, 3, 2 and 6 defines the same area, as this is delimited by a minimum content of iodine of 5 percent by weight. An addition of iodine in an amount of only 1 percent by weight to any arsenic-sulfur composition lying between the compositions corresponding to points 1 and 4 has a visible effect, lowering both the 30-poise viscosity point and the softening point. However, a much more significant difference is achieved when 5% by weight of iodine is added, and it can be expected that at least this minimum content will be used in factory production. For certain reasons, it must be assumed that the preferred composition ranges of the vitreous material in the arsenic-sulfur-iodine system are limited to the range that lies within the straight lines between points 5, 3, 2 and 6. For certain purposes, the range that is preferable, is limited by points 7, 8, 9 and 10. This area is formed as follows: the iodine point is connected to the 40% sulfur — 60% arsenic and 70% sulfur — 30% arsenic binary composition points. A line is drawn through point 7 corresponding to an addition of 20% iodine, the stoichiometric mixture, and point 10 corresponding to an addition of 10% iodine, 70% sulfur — 30% arsenic composition. The other side of this region is determined by a line passing through points 8 and 9, which line corresponds to the end of the high iodine glass-forming region.

The area between the two lines joining the iodine point with the binary boundary for arsenic-sulphur indicates a compositional area which, from a theoretical point of view, shows no tendency to pass into solid form as a result of crystallization of the element sulfur on the one hand and arsenic sulphide on the other hand (although samples have shown no instability in either composition within the range between the lines running between points 1, 2, 3 and 4). Points 7 and 10 on these composition lines correspond to the approximate minimum amount of iodine required to produce a 30-poise viscosity point lowering of about 30% on the Celsius scale, in both cases corresponding to a value approximately in accordance with the accepted maximum temperature that can be tolerated by semiconductor devices.

The circled point 11 corresponds to a glass mixture according to the invention which has been used to empirically determine the

specified limits for the system's glass areas.

In reality, several of these mixtures were used when encapsulating the semiconductor devices. The improvements of the

electrical properties that were obtained in

in connection with the use of these, is indicated in this invention. The 30-poise viscosity points and softening points for many of the mixtures are for points 11 indicated in fig. 2 and 3. The circled points 12 correspond to mixtures which have been placed by evaporation, which will be described in more detail below.

The areas which are preferable, and which have been discussed in the previous section, are indicated on the diagram in fig. 1. The basis for the preferred composition ranges is indicated. It is noted that these areas are only preferred for certain purposes. As an example, it can be mentioned that the minimum content of iodine in each individual case was determined mainly due to the reduction of the 30-poise viscosity temperature. A low viscosity when the material is melted is of great interest in connection with placing the material by immersion on delicate objects, where too great a force can result in a wire or a contact point being broken. When applying the material by evaporation, it is not required that the object to be covered or otherwise come into contact with the material is heated to the melting temperature, so that, when the viscosity is of little importance, it may be more advantageous to use a composition within the larger range 5 . other properties may require other compositions to be used. It can be said that the compositions which are preferable within the preferred areas are selected on the basis of the wishes that can be made in the particular areas of application.

The limit compositions corresponding to the numbered points are the following in percentage by weight:

Fig. 2 shows a ternary diagram for the arsenic-sulphur-iodine system on the same coordinates as fig. 1, as the temperatures at which some of the aforementioned compositions have a viscosity of approx. 30 pois. The temperature is indicated in C°. The exact composition points are at the center of the middle number in the indicated temperature. The information found on this figure is of particular interest when encapsulating by immersion of delicate parts. In general, when immersion coating semiconductor devices, it is unnecessary to have viscosities that significantly exceed 30-poise, as a result of the devices' sensitivity. Somewhat higher viscosities can be used when encapsulating larger ones objects or in the case of more stable devices or larger interconnected units. Fig. 3 shows a ternary diagram for the arsenic-sulphur-iodine system plotted on the coordinates according to fig. 1 and 2; the temperatures corresponding to the softening temperatures for these compositions are indicated. The exact composition lies at the center of each of the indicated values. Data for the softening points of the kind indicated here are of greatest interest when designing encapsulated devices that are to be exposed to very low temperatures. It appears from this figure that the softening points for the indicated compositions lie between a lower value of approx. — 17° C to an upper value of approx. 200°C.

As stated above, the information on fig. 3 for the softening points of importance for the getter effect of the vitreous material. It has been established that the getter effect of the iodine compositions is significantly better than with the previously mentioned thallium compositions. This improvement can be associated with the movement characteristics of the getter process. This tendency was of course observed with the thallium system, where a greater getter force was ascertained for the compositions that had low softening points.

Although the glass area in fig. 1 is stated as completely accurate, it is pointed out that the limits are approximate and largely depend on the prevailing manufacturing conditions. The vitreous region can be stretched somewhat by rapid cooling, whereby recrystallization is as small as possible. In accordance with other vitreous compositions, it can be expected that materials lying on the outside of the specified areas can be stabilized in the vitreous state by the addition of one or more stabilizing materials.

For the sake of simplicity, the composition values are given for the pure arsenic-sulfur-iodine system. In accordance with what one might expect, it has been established that smaller amounts of additive or replacement materials can be used without the glassy state being affected. Such additions can also be obtained without it being done on purpose. Substances that change the physical or chemical properties can be added on purpose. It is e.g. been found that different materials can be added to increase the plastic flow temperature. Additions of this kind do not necessarily have to have the same effect on the softness points. An addition of less than 2% by weight of lead to a 30-70% by weight arsenic-sulfur secondary material, to which 10% iodine has been added (24-67-9% by weight arsenic-sulphur-iodine), has e.g. resulted in an increase in the 30-poise viscosity temperature from 311° C to 374° C, while the softening point was only raised by approx. 2° to 55° C. An addition of 5% by weight of antimony to the same basic composition raises the 30-pois point to 370° C and the softening point to 87° C. The lower limit of an addition of antimony which causes a significant effect is 1% or less, and for lead significantly lower, approx. 0.1. An addition of 5% bismuth or tellurium increases the wetting power on ceramics and also results in a significant wetting of silicate glass. Bismuth works in much the same way as antimony in that in the specified range the 30-poise point increases much more than the softening point. Fundamental investigations seem to show that selenium and/or tellurium can be used instead of sulfur and antimony and/or bismuth instead of arsenic up to at least 20 mole percent of the replaced material without the glassy state being affected.

Several of the compositions indicated in fig. 1 has been prepared with bromine instead of iodine. In general, it has been found that the bromine-glass region is essentially the same as for the iodine system. For vitreous materials of this kind, it is typical that the viscosity is somewhat less than for the corresponding iodine system. Several of the bromine glasses are liquid at room temperature and some of these have viscosities over a temperature range of up to approx. 350° C, which goes against the viscosity of water at room temperature. These materials can be used as liquid filling materials and for cleaning operations before the objects are encapsulated in another way, i.e. they are a possible substitute for vacuum baking of semiconductor devices before they are encapsulated in containers. Viscosity and softness-binding point values that lie between those that are characteristic of the glasses of iodine and bromine can of course be obtained by combining the two materials.

The vitreous materials according to the invention and the materials with thallium and similar systems mentioned earlier can be dissolved in each other. A combination of this kind provides yet another method of producing the desired physical properties for the resulting material.

As in the previous experimental

glass systems, the basic ingredients can be replaced to a certain extent. A replacement of iodine with bromine has been mentioned. A replacement of chlorine and fluorine is an obvious possibility, although an absolute replacement with other materials is presumably disadvantageous due to the transport problems associated with the more gaseous materials which also react more easily. A partial or complete replacement of arsenic with antimony or bismuth usually results in a reduced buoyancy. Sulfur can be partially replaced with selenium, whereby the glass-like compositions have an increased wetting power on ceramics or other glass compositions such as e.g. silicates and borosilicates. In the same way, a partial replacement of sulfur with tellurium will greatly increase the dielectric constants for a particular arsenic-iodine composition.

The maximum content of non-intended materials depends on how

the position must be used. When encapsulating or otherwise placing the material on semiconductor devices, it is common to adhere to a very high degree of purity. When covering less sensitive devices, e.g. transformers, inductors, capacitors and the like, such a high degree of purity is not required during manufacture. In most areas of application, the glasses according to the invention are assumed to have their value as a result of their getter properties.

As mentioned above, it has been established that relatively large amounts of 20% by weight or more of a number of non-ionic materials can be added to the glasses without their glass properties being affected, (although such additions have the desired effect by changing the physical properties) . Materials of this kind include gallium, indium, thallium, antimony and bismuth.

It has been shown that a relatively large degree of ionic impurities can be tolerated in the compositions, even if the materials are to be placed in direct contact with the surfaces of very sensitive electrical circuit elements. Nevertheless, it is assumed that in the event of a larger production, the standard purity generally required for the devices to be encapsulated or treated in this way will be taken into account. As the getter effect is affected by the total amount of contaminants present, the use of very clean materials will probably result in even greater improvements to the objects.

In the following - to facilitate the understanding of the invention - a suitable method for the production of glassy materials according to the invention will be described. However, the method to be described serves only as an example, as alternative methods are suggested and others are known to those skilled in the art. The method can be used in the production of a ternary composition in accordance with the diagram in fig. 1, i.e. a composition of the arsenic-sulphur-iodine system. The same method can be used in the production of glass in the arsenic-sulphur-bromine system and for glass where arsenic has been partially replaced with antimony, or where bismuth and sulfur have been replaced with tellurium or selenium. As will also appear, the glasses according to the invention are not limited to an addition of two or three elements, but can contain more than one element at any point, such as e.g. both sulfur and selenium or arsenic and antimony.

In order to facilitate the storage etc. of the starting material, binary compositions of e.g. arsenic and sulfur and arsenic and iodine. An alternative method is to prepare the final mixture of the three elementary materials.

Procedure: The starting materials are iodine, sulfur in powder form and arsenic in metallic form. The iodine material is weighed and the quantities of sulfur and arsenic required to form the desired composition are calculated in relation to this.

The sulfur is weighed and placed in a test tube with a loose stopper and held over a Bunsen burner. The sulfur is heated until it melts into a thick, rubbery consistency. The test tube can be left open if an inert shielding gas is used. Arsenic in the form of small metal parts and iodine in the form of crystals are weighed and added to the molten sulphur. The tube is heated with its contents until an exothermic reaction occurs at a temperature of approx. 500—600° C.

The contents of the tube are mixed by oscillating until all the metallic arsenic dissolves.

The tube is then heated until the contents have completely converted to liquid form.

It has been shown that a 50-gram mixture melts into a homogeneous mixture in the course of approx. 20 min.

The molten mixture can be cooled rapidly in liquid nitrogen, whereby it is prevented from adhering to the container.

This method is particularly suitable for production on a laboratory scale. The relevant reaction and melting times can for the most part be determined simply by observation. The relevant temperatures are likewise of little importance and vary in relation to the composition of the mixture and the nature of the reaction. The temperature at which the glassy material is formed is determined e.g. of the energies associated with the exothermic reaction and of the device temperatures that can be accepted for devices when the devices or objects are later used.

The device shown in fig. 4A, 4B and 4C is an n + p diffused junction diode of silicon with a breakdown voltage of 56 V. This device comprises an n<+>p element 21, leaf-shaped leads 22 and 23 of gold<1> which at heat and trykik are attached to gold-plated Kovar electrodes

25 and 26 which in turn pass through a gold-plated Kovar head 24. The electrodes 25 and 26 are electrically isolated from the head 24 by means of a thermally adapted borosilicate glass seal. The manufacture of the head and the adapted vitreous material is well known from other areas. Element's 21st

tolerances are very small and are of the order of 1/40 mm between the wires 22

and 23 and the active compound. Devices of this type are particularly sensitive to degradation of the electrical properties as a result of contamination on the surface.

The device in fig. 4A is shown placed over a container 28 which may consist of chemical porcelain or other glass, metal or ceramic material; the container contains molten glass material 29 according to the invention. The glass material is kept melted by means of preservatives not shown.

In fig. 4B, the device 21 is immersed in the material 29 in the container 28.

Immersion times have been found to be a critical factor in the degree of improvement that can be achieved by encapsulation. The values indicated on fig. 8 car obtained by devices that were immersed for a time of approx. 30 sec. The data set out in Table 2 were obtained with devices that were immersed for a time of 1 y 2 min. It appears that the improvements that are achieved are greater for the devices that were submerged for a longer time. From fig. 8 and table II - it can be seen, however, that the power maturation or the thermal maturation can cause the devices to acquire the same operational characteristics. Where longer immersion times were used (Table II), the heat ripening had not resulted in further improvements to an appreciable degree.

After immersion of the device 21 with associated parts, it is lifted up and the glassy material adhering to the device is allowed to harden.

Fig. 4C shows a device of this kind

after the material has solidified.

Figs. 5A, 5B and 5C show an alternative method of encapsulation by immersion, where the container forms a final encapsulation. In these figures there is a device 40 with the same shape as that shown in figures 4A to 4C; the device comprises element 41, and wires 42 and 43 which are spot welded to electrodes 45 and 46 pass through a top piece 44.

The electrodes 45 and 46 are isolated from the metal upper part 44 and by means of a glass material, such as borosilicate glass. The container 47, which can be made of glass, metal or ceramic, is heated using heating means not shown, and it contains molten glass 48 with a composition according to the invention.

Fig. 5A shows the device 40 before the immersion. Fig. 5B shows the device immersed in the molten glass 48; the glass is kept molten for at least as long as is required for all surfaces to be covered. It appears that the capsule 47 has such a size that it forms a tight connection with the upper part 44. As a rule, it is not necessary for this connection to be hermetically sealed, as the glass forms a tight unit around the device, so that moisture etc. cannot penetrate in.

The material 48 in fig. 5C is solidified, so that the device 40 is encapsulated in the glass-like medium 48.

Figures 4A to 4C and 5A to 5C show encapsulation methods for encapsulating devices with the glasses according to the invention. Other methods can be mentioned, e.g. application with a brush, spraying and vapor deposition. As previously mentioned, it is not required that the entire device be encapsulated. In connection with previously shaped valleys of the glass, it is advantageous that these have such a shape that they fit exactly over one or more wires which are attached to the most voluminous part of the device. Although the aforementioned diode device is a good example of the type of device which is most sensitive to atmospheric influences and which can therefore be encapsulated with great advantage by the aforementioned methods, other devices can also be improved to a significant extent by a similar treatment. Resistors, capacitors, rectifiers, both elemental and oxide rectifiers, inductors, transformers and other circuit elements as well as entire compositions of such elements, can advantageously be encapsulated both by the methods mentioned in connection with fig. 4 and 5 and by means of other methods.

When the devices are to be coated by immersion according to the described method, it is advantageous that the glass is kept at its immersion temperature, which is generally stated to correspond to a viscosity of approx. 30-pois, as with a hot plate. However, since the getter activity of a particular glassy material has been shown to be dependent on temperature, it is preferable that the glassy material be kept at a temperature that approaches the highest that can be allowed for the device to! be encapsulated. When encapsulating n<+>p silicon diodes, it was found to be advantageous that the temperature was between 300 and 350° C. To protect the glass and possibly also the device from oxidation at this high temperature, it may be advantageous to keep it molten material in an inert atmosphere, although it has not been established that air has any degrading effect on the glass. The gases that can be considered advantageous in this connection include nitrogen, helium and argon.

In accordance with usual methods, heat treatment can be used to reduce the forces that can arise in the device in the case of vitreous materials that have softening points that are well above room temperature; this can be achieved by lowering the temperature gradually from the temperature corresponding to the softening point down to room temperature, e.g. over a period of several hours. This heat treatment can be carried out directly in connection with the cooling or cooling down after encapsulation, or at a later stage as a separate treatment.

When the devices to be encapsulated are fragile, it is desirable that the molten sandinig is kept at such a temperature that the viscosity is not much greater than 30 pois. The 30-pois range for these glasses is found at temperatures between 55 and 450° C. The temperatures that can be tolerated depend on the device to be treated. As a rule, the highest temperature that can be tolerated by a semiconductor device, e.g. germanium or silicon or a group III — igroup V diode, triode or tee-trode, usually the one corresponding to the lowest melting point of the soldered tin or the alloying materials present. In contrast to ordinary glass, even the highest temperatures within the 30-poise range are too low to produce any significant change in the shape or properties of the device to be encapsulated. The 30-pois temperatures for some materials are given as examples, in fig. 2.

When it is desirable for the glass to be protected against external influences, this can be achieved by surrounding it with another material, regardless of its contamination. There can e.g. plastic is used, such as polyvinyl chloride, polyethylene or the like. Such an outer casing significantly reduces the risk of cracks forming when the device is exposed to very low temperatures.

When a hermetically sealed seal is to be achieved with the encapsulation, a hermetically sealed connection must be created between the coating and the electrical cables. This is achieved by the materials described, when they are used in connection with the metals aluminium, silver, gold, platinum, tantalum, molybdenum, nickel, tungsten, brass and kovar (an alloy with approx. 53.7% iron by weight, 29 weight percent nickel, 17 weight percent cobalt and 0.3 weight percent manganese). The excellent wetting properties of the vitreous materials on aluminum and alloys containing aluminium, is particularly advantageous as this metal has a temperature coefficient that is closer to that of the glass material than any of the other common metals used as wires or other structural parts is.

The apparatus shown in fig. 6, has been successfully used in laboratory measuring tests with glass encapsulation of the aforementioned compositions. The apparatus consists of a plate 55 and a hermetically sealed dome 56 which is hermetically sealed to the platform 55 by means of a neoprene ring 57. The gas in the apparatus is removed through the pipe 58 which is connected to a vacuum pump not shown. The glaze mixture 59 to be applied by evaporation is placed in a vessel 60 which is in turn placed within the uppermost winding of a resistance coil 62 which is attached to clamps 63 and 64, which in turn are attached to electrode holders 65 and 66; the glass material may consist of powder or have another advantageous form. The electrode supports 65 and 66 are electrically connected to a power source 66 via wires 68 and 69. An object 70 which is to be provided with a coating, e.g. a printed circuit board, is attached to the clamp 71 on the support 72. The apparatus can be provided with means for turning the support 72, so that the object 70 can be moved in front of the glass source 59.

The entire glass-forming area in fig. 1 as well as the aforementioned combined or substituted systems that have been described can be used for vapor coating. It is considered to be a great advantage that these compositions, in contrast to the usual glass compositions, can provide a coating that is relatively thick (0.38/1000 mm or thicker).

Although a homogeneous single-phase glass is obtained when any of the materials that have been mentioned are used, it should be noted that deviations between the compositions of the source material and the deposited materials may occur in certain cases. In this connection, it has been established that the vapor pressure of stoichiometric arsenic sulphide is somewhat greater than that of the iodine mixtures found in the arsenic-sulphur-iodine system. When the material is therefore completely evaporated, the material that first deposits will be rich in As2S.s, and the material that is finally obtained will be richer in iodine. Although this may not be considered significant in most encapsulation processes, a greater effect is obtained from the variation in iodine content as a change in the 30 pois and softening point temperatures; however, the deposited materials can be homogenized by heating the material either during the coating itself or later.

When a larger material source is used, the desired composition for the coatings can be achieved by gradually regulating the composition of the source. As previously mentioned, glasses containing bismuth, tellurium and selenium will have a greater wetting effect. In particular, it has been established that connections with very high adhesion force are formed between glasses containing selenium and a large number of organic and inorganic materials including carbon, ceramic materials including those containing silicon and aluminium, other materials such as borosilicates and polymeric materials comprising halogenated hydrocarbons, such as e.g. perfluorocarbon.

Application of a coating by evaporation and the effect of a change in the distances and other conditions are well known in this field. However, it has been established that with a material source 59 which has a diameter of approx. 1 cm, a coating of uniform thickness is obtained on a device 70 which has a largest dimension of 3 cm and is located at a distance of 15 cm. If the distance between the source and the device is increased, the evenness of the coating will not deteriorate, but the time required to achieve a certain thickness is increased. If the distance is reduced, this can cause the coating to have an uneven thickness, which can be both advantageous and disadvantageous.

Fig. 7A shows a semiconductor device 75 which is attached to a heat-conducting part 76. The electrical connections are made over electrodes 77 and 78. The sensitive part of the device is in this case its upper surface, where the electrode connection 78 is attached, or a part of the device 75 above the heat-conducting part 78. The glass part 79, which can consist of a body which is pressed from powder of one of the aforementioned glassy materials, consists of a short tube which is placed over the line 78 in connection with the device 75's upper surface.

The temperature of the device 79 is raised to the flow temperature and held there for a time sufficient for the glass to flow around the device 75 and form a connection between the glass and the heat-dissipating part 76. The flow temperatures for the glasses lie somewhere between the 30 pois temperature and the softness temperature. As an example, it should be mentioned that a 24-67-9 weight percent arsenic-sulphur-iodine glass material with a softening point of approx. 53° C, has been found to have sufficient flow properties over a temperature range of between 125 and 200° C, whereby the glass flows out over the device in the course of 10 to 15 mm.

It appears from fig. 7B that the previously shaped device 79 has changed shape during the heating, so that the device 75 has been encapsulated and a hermetically sealed connection has been formed with the surface of the heat-conducting part 76.

Although from a laboratory point of view it is considered that encapsulation by immersion is the most appropriate method, it is assumed that for series production it will be appropriate to use pre-formed glass units. By using such glass units, the glass can be placed on the devices immediately after they have been manufactured and a large number can then be encapsulated at the same time by keeping them at a suitable temperature for a shorter time.

In fig. 8 shows curves for power maturation for phosphorus-boron diffused silicon diodes of the type that has been described in connection with fig. 4A to 4C, and which has been encapsulated with glass according to the invention. The methods that have been followed are the usual methods of aging used for lattice devices. In the case of encapsulated devices that are produced on a commercial basis, the maturing methods are used to uncover the latent errors that may be found and at the same time stabilize the device's characteristics, whereby this usually results in the characteristics being somewhat worse than those the device had before the encapsulation. In the case of devices that are placed in tight containers, the ripening will show whether there are major leaks and point to changes in the characteristics as a result of ionic or other sources. These corrosion tests are performed under a variety of different conditions. Diodes can e.g. a bias voltage is applied which is either directed towards or with the direction of conduction. It is known that the greatest possible power maturation is achieved under the conditions that produce the greatest possible heating of the device. As a result, a deterioration of the characteristics and a stabilization of the devices occurs most quickly, when a bias voltage is applied in the conduction direction, so that a very large current flows and the device is heated accordingly.

At various specific times during the power maturation, this bias is removed and a new, oppositely directed bias is applied to the device at the same time as the leakage current is measured. The ordinates on fig. 8 in the form of milliamperes is a measure of this leakage current. It is common to measure the leakage current with an opposite bias voltage which is a relatively large part of the device's breakdown voltage. The device which gave the data shown in fig. 8, was a low-voltage diode with a breakdown voltage of approx. 56 V. The leakage currents were (measured without a counter-directed bias voltage of 50 V.

The device in fig. 8 was encapsulated by immersion in a glassy material which consisted of 24% by weight arsenic, 67% by weight sulfur and 9% by weight iodine. It appears from the curves that the leakage current showed a tendency to decrease for all the devices that were subjected to the test, as the average decrease during the 100-hour test was approx. 5 times. During the tests, the devices were subjected to a bias voltage in the conduction direction which was sufficient to produce a constant current of 200 mA.

The devices were kept immersed in the molten glass for a time of approx. 30 sec. Although it is not shown in the figure, the greatest dei of the improvements in the discharge flow is achieved during the immersion itself. The devices had a leakage current at 40 V reverse bias of approx. 20 in.uA, or greater. The improvement achieved by immersion was of one order of magnitude or greater. As can be seen from the curves, however, this great improvement is not the greatest that can be achieved by the device's characteristics. It is noted that for the characteristics that have been recorded, there is no convergence of the curves or any maximum degree of improvement. It must therefore be assumed that the advanced maturation and/or use of the devices would lead to a further improvement of the operating characteristics and further ionic getter effect. The devices' electrical characteristics could all be accepted according to the usual specifications, and this applied both before and after the immersion. The improvement described cannot therefore be associated with the recycling of discarded units. In this connection, it can be mentioned that the measured data has shown that a discarded device (table II) which had a leakage current at a reverse bias of over 100 m^A, was improved to such an extent that after the immersion and the heat treatment was well within the acceptable limits, so that this discarded unit was recovered. The subsequent heat ripening on

17 hours at 150° C resulted in a further improvement of 2 orders of magnitude, whereby the device obtained the same characteristics as the others that had been produced in the same group. From an economic point of view, the conclusion can be drawn that the economic result achieved with these materials, either as a final encapsulation agent or as a cleaning or filling material, is not limited by the encapsulation's total price. The implied increase in dividends can result in savings that are several. times greater than the total costs of encapsulation. In order to show the improvements in the electrical characteristics that can be obtained by encapsulation, and at the same time to show the equivalence between different glassy materials with regard to this, a further group of devices was subjected to tests; the devices were of the same type as those that have been described in connection with fig. 8. The devices were first cleaned and the leakage current under a reverse bias voltage of 40 V was measured. The devices were then dipped into a molten solution of a glass material which was maintained at a temperature of 300-350°C; the type of capsule that has been described in connection with fig. 5A—5C, were used in this connection, after which the electrical measurements were repeated. The immersion conditions were different from those that prevailed during the immersion of the devices described in connection with fig. 8, mainly in that the immersion time in the molten glass was different. In this case, the devices were immersed for a longer time, iyz min. The results obtained were generally consistent from material to material and represented an improvement of approx. two orders of magnitude. These results are set out in Table II.

The devices were subjected to heat aging at a temperature of 150° C and for a time of 22 hours. The changes in the characteristics that appeared were irregular, except for the devices that had a relatively high leakage current after immersion. The diodes No. 4, 15 and 19 had leakage currents of 0.3, 0.45 and 0.3 m|xA after the thermal aging. A smaller number tested under the same conditions as indicated in Table II failed. The failed devices all had some kind of open electrical connection, most likely an open wire or contact. It is noted that the vitreous material used for the diodes 1-5 is the same as that used in connection with fig. 8. These have been included in the table to form a basis for a comparison and show that the final values that have been achieved by power maturation can be achieved by immersion alone under the stated conditions.

The maximum leakage current which is permitted according to manufacturing specifications for the devices which were subjected to testing, and whose test results are given in Table II, is 100 muA. It therefore appears that all the devices, except no. 19, fulfill the trading conditions set both before and after the immersion. In this connection, it is interesting to note that the improvement in the leakage current for this discarded unit was of the same order of magnitude as the others, and that the level of the leakage current was reduced to the optimum value (as indicated for the other units) after heat aging.

It is thus clear that an improvement in the characteristics can be produced by heat ripening and effect ripening as well as during encapsulation. The encapsulation itself has in certain cases led to a significant lowering of the leakage current to a value which is the best that has been achieved for these devices to date. The stability of the leakage current during heat aging seems to show that the characteristics have become as good as possible, at least for the prevailing test conditions and for the vitreous material that was used. It appears that this effect is increased by an increased treatment time with the molten material and is thus more prominent when encapsulating by immersion than by evaporation, especially when the coating is placed on a non-heated device, although the same improvement can be achieved by keeping the device heated for a time corresponding to that used for immersion. It also appears that a significant improvement is obtained by heat maturing for a relatively short time at temperatures around 150° C. This effect is increased at higher temperatures even though this can lead to a plastic flow when an unmodified arsenic-sulphur-iodine composition is used without an outer container. This effect can be avoided by using an additive such as e.g. lead, antimony, bismuth or one of the other previously mentioned additives.

Although materials with lower softening points show a greater improvement in the Characteristics when ripening for a specific time and a specific temperature, it is to be expected that devices encapsulated in any of the aforementioned compositions, including those containing additives or where elements are have been replaced with others, so that the plastic flow temperature is increased, will eventually acquire such properties that the characteristics are not affected by ionic contaminants. That this point has been reached is indicated by the plasma breakdown which can generally occur in devices which have the limitation of the body, and which has been observed in devices which have been encapsulated in the vitreous materials. This final value obviously depends on the device as such rather than on the nature of the vitreous encapsulating medlum.

It is emphasized that many of the glassy materials, including all those in which diodes 6-19 (Table II) were encapsulated, were originally intended for thermal tests. These materials were not manufactured in accordance with the purity requirements recommended for the manufacture of semi-conducting devices. The values that have been entered show that a high degree of purity is not required. The first vitreous material in table II was also used when encapsulating the devices in connection with fig. 8, and these were produced in a more careful way. The ionic contamination level for this composition was 0.001%, and for the other compositions approx. 0.005%. It can be seen that the characteristics that were measured did not differ much from one composition to another. Experiments have shown that effect ripening is effective as a means of faster ripening, only as far as the temperature of the device is increased as a result of heating. Heat ripening is therefore equivalent to effect ripening. Both types of ripening are therefore considered to be an accelerated test method that reveals the deviations in the characteristics that can be obtained during use and, to a lesser extent, during storage ripening.

The properties of the vitreous materials are otherwise very good. The internal resistance of glass in the arsenic-sulphur-iodine system is in the range 1012 to 1010 ohm/cm. The dielectric constants for the glasses in the iodine and bromine systems lie in the range 4 to 12 and 4 to 10, respectively. The dielectric loss for these materials at a mega-Hz is approx. 0.0005 and 0.0003 respectively.

It is assumed that the improvements in the characteristics that have been achieved are a consequence of the encapsulation medium itself and not of the specific encapsulation method used.

The idea of using the described glass materials as getter materials has also been discussed. They can thus be used instead of vacuum baking or other cleaning methods in connection with encapsulation or other packaging. Here, only a local wetting of the device is required, as a hermetic seal is not necessary, but can be achieved with the help of an outer container. Such a partial overdraft produces

also the improvements that have been mentioned.

When the primary thing is to clean or get a getter effect, the material can consist of dry

powder that is placed in the devices, after which the encapsulated device finally

is heated so that the material flows out and

covers the devices in the desired places.

Claims (7)

1. Procedure for overdraft of electrical circuit elements with a single-phase glass-like material, characterized in that the glass-like material with which the element is brought into contact consists of arsenic, sulfur and one or both of the elements iodine and bromine. 2. Method according to claim 1, characterized in that a ternary arsenic-sulphur-iodine composition is used which lies within the area on the ternary composition diagram of these three elements which is determined by the straight lines connecting the following composition points expressed in weight percent. 3. Method according to claim 2, characterized in that the aforementioned area is determined by the straight lines connecting the following compositional points expressed in weight percent. 4. Method According to claim 1, characterized in that the contact between the materials is achieved by immersing the element in the molten glass-like material. 5. Method according to claim 1, characterized in that the material is placed in contact with the element by steam coating the glassy material under partial vacuum. 6. Method according to claim 1, characterized in that the glassy material in the form of a predetermined unit is placed on the element, after which the temperature of the material is raised so that it flows out and covers at least part of the element. 7. Method according to claim 1, characterized in that the vitreous material also contains one or more of the elements: lead, antimony, bismuth, tellurium and selenium.