FR2462780A1 - METHOD FOR MANUFACTURING A HIGH BREAKING VOLTAGE P-N JUNCTION - Google Patents

Abstract

IONS OF A FIRST IMPURITY (ARSENIC) OF A TYPE DETERMINING A GIVEN CONDUCTIVITY AND HAVING A RELATIVELY LOW DIFFUSION COEFFICIENT ARE DISTRIBUTED IN THE SUBSTRATE TO FORM A LAYER OF A FIRST IMPURITY BY IMPLANTATION ENERGY ENERGY , FOLLOWED BY HEATING FOR THE PARTIAL DIFFUSION OF IMPLANTED IONS. IONS OF A SECOND IMPURITY (PHOSPHORUS DESIGNATED BY PH) OF THE TYPE DETERMINING THE SAME CONDUCTIVITY AND HAVING A RELATIVELY LOW DIFFUSION COEFFICIENT ARE DISTRIBUTED IN THE SUBSTRATE BY A SECOND IMPLANTATION OF IONS TO FORM A SECOND IMAGE LAYER LOCATED MORE DEEP INTO THE SUBSTRATE THAN THE FIRST LAYER. THE FIRST LAYER HAS A SUBSTANTIALLY HIGHER IMPURITY CONCENTRATION THAN THE SECOND. THE FIRST AND SECOND IMPURITIES ARE DIFFUSED BY HEATING THE SUBSTRATE IN A WAY TO BRING THE LAYERS TO MELT OR COMBINE IN A SINGLE LAYER OF IMPURITY CONSTITUTED OF BOTH THE FIRST AND SECOND IMPURITIES. THE RESULTANT IMPURITY LAYER IS SUBSTANTIALLY INERATE TOWARDS LATER THERMAL MANUFACTURING PHASES AND HAS A RELATIVELY PROGRESSIVE DISTRIBUTION PROFILE SUCH AS THE P-N FORMED JUNCTION HAS A RELATIVELY HIGH BREAKAGE VOLTAGE.

Description

The present invention relates to a method

  manufacturing a semiconductor device, and more

  In particular, a process for the production of a P-N junction in a substrate which is practically insensitive to the heat applied during the subsequent manufacturing phases and which has a relatively high breaking stress. One of the basic techniques used in the manufacture of semiconductor devices is training

  a P-N junction. Various processes are known for establishing

  ment of a P-N junction in a substrate. For example, starting from a P-type silicon substrate, it is known to distribute, either by implantation of ions, or by other

  doping techniques, an N-type impurity in the sub-

  trat in order to modify the conductivity of the substrate within the limits of the distribution area. By appropriate selection of the impurity and by adjusting the concentration and

  configuration of the distributed impurity layer, the characteristics

  electrical characteristics of the resulting P-N junction can

  be adjusted between certain required limits.

  A semiconductor device is normally subject after the creation or establishment of the P-N junction, to. manufacturing stages which influence the concentration and configuration of the impurity layer and therefore the electrical characteristics of the P-N junction. By

  example, subsequent heating phases carried out

  above the diffusion temperature of the distributed impurity tend to cause the diffusion of the distributed impurity in

  dispersing and reducing the concentration of the layer of-

  impurity and causing migration of the P-N junction.

  The degree of alteration of the distributed impurity layer

  depends on the concentration of this impurity, on its coeffi-

  diffusion cient and parameters of the heating phases

  later. It is therefore necessary to take into consideration

  ration this alteration of the distributed impurity layer caused by the subsequent heating stages, when

  determines the nature of the original distribution.

  Certain impurities such as arsenic or

  antimony, have relative diffusion coefficients

  This is very low and significant diffusion of these impurities only takes place at relatively high temperatures (11000C). The subsequent thermal manufacturing phases

  which normally take place at a temperature (950 C) lower

  re at the diffusion temperature of these impurities, will have

  therefore little effect on a repaired impurity layer

  tie. For this reason, impurities with diffe- rent coefficients

  relatively weak melting, are desirable for the rea-

  reading of P-N junctions, since when they are

  distributed in a layer of concentration and confi-

  desired guration, they are relatively insensitive to al-

  erosion caused by the subsequent manufacturing phases

your.

  However, impurities with low coefficients

  diffusion clients tend to form a layer having

  a distribution profile with a periphery or limit

  relatively steep. Such a layer with a periphery

  relatively abrupt laugh or limit leads to a junction

  P-N with a relatively low breaking tension.

  This is how there is competition or competition between

  electrical characteristics of the formed junction and its re-

  resistance to subsequent thermal manufacturing stages. If one wishes to precisely adjust the characteristics of the P-N junction, this is done at the expense of a voltage of

high break.

  On the other hand, there are certain impurities,

  such as phosphorus which have a coefficient of diff-

  higher melting and therefore require lower temperatures to achieve high diffusion. Of

  such impurities are distributed in a layer with a pro-

  diffusion wire having a relative periphery or limit

  progressively progressive, but the layer will be much more affected

  ted by subsequent thermal manufacturing stages. Ain-

  if, although it is desirable to have a layer with a distribution profile having a relatively progressive periphery or limit because it results in a relatively high breaking stress, it is extremely difficult, when using such impurities, 'to settle

  the characteristics of the P-N junction with precision

  se of the relatively high effect exerted by the subsequent thermal manufacturing phases on the concentration and configuration of the impurity layer. That is why,

  with impurities such as phosphorus, it is necessary to renounce

  cer to a degree of substantial adjustment in favor of the concen-

  tration and configuration of the impurity layer, so

  to obtain a relatively high breaking tension.

  The most advantageous situation would be to

  use an impurity which can be distributed in a layer

  che with a relatively progressive periphery or limit-

  sive, which would give a P-N junction with a relatively high breaking tension, and which at the same time has a

  low diffusion coefficient such as the manufacturing phases

  subsequent thermal cations have a relatively small effect on the concentration and configuration of the impurity layer. However, the higher the diffusion coefficient

  the weaker, the steeper the periphery of the distributed layer.

  There is therefore always a competition between the characteristics

  electrical junction ticks and ability to adjust

  of the concentration and configuration of the layer of-

  impurity. When manufacturing the semiconductor,

  choose one characteristic at the expense of the other. The pre-

  sente invention therefore proposes mainly to provide a method of manufacturing a P-N junction in a semiconductor device in which a high breaking voltage can be obtained without sacrificing the possibility of adjusting

  the impurity concentration and the depth of the junction.

  The invention also proposes to provide

  a process for developing a P-N junction with a voltage

  relatively high breaking strain, in which two impu-

  different reties are distributed in separate layers

  in the substrate, which are then brought together to form

  sea a single layer of the two impurities.

  Another object of the invention is to provide a method for manufacturing a P-N junction having a

  relatively high breaking stress, in which

  reads a second impurity leading to a layer with a

  do relatively progressive periphery to achieve a profile

with progressive distribution.

  The invention also relates to a method of fa-

  bridging of a P-N junction with a breaking voltage

  laterally high which only implements equipment

  conventional and which more particularly includes

  relatively simple procedures that can be implemented

lives with such equipment.

  According to the present invention, ions of a first impurity of a type determining a given conductivity and having a relatively low diffusion coefficient are distributed in an area of the substrate to form

  -20 a first layer of impurity. Second impure ions-

  tee of the same type determining the conductivity and having a relatively high diffusion coefficient are distributed in the same area of the substrate to form a second layer of impurity. The second layer of impurity is distributed more

  deep into the substrate as the first layer of impu-

  reté. The first and second impurities are diffused in

  view of a combination of the first and second layers

  to form a single layer of impurity

  of both the first and second impurities.

  The first impurity is chosen so as to have a relatively low diffusion coefficient, since the subsequent thermal manufacturing stages

  must have or will have little effect on its distribution-profile

  bution. The second impurity is chosen so as to have a higher diffusion coefficient such that the distribution profile of its layer is relatively progressive. The

  combined layer, because it is made up of the com-

  bination of the two impurities, is therefore relatively insensitive to subsequent heating phases or stages but gives rise to a junction having a high breaking voltage. The impurity concentration of the first impurity layer is adjusted to be significantly higher than the impurity concentration of the second

  impurity layer so as to reduce the tendency of the

  che merged or combined to broadcast during stages of

  subsequent heating. For example, the concentration of the pre-

  first impurity layer may be two orders of magnitude (two points) higher than the concentration of the second

impurity layer.

  The second impurity layer is formed at

  a depth greater than the first layer of impurity.

  By forming the second layer of impurity at a greater depth

  large than the first layer of impurity, when the first

  re and second layers are combined, the portion of the

  che merged or combined near the periphery or boundary is made up of a relatively high percentage of the second

  impurity. Thus, the single impurity layer presents a re-

  relatively progressive partition at its limit and at the same

  time, because of the higher concentration of the first

  re impurity in the mass of the layer. The configuration of

  this will be almost insensitive to the alteration due to the sta-

  subsequent thermal manufacturing.

  The first impurity is preferably repaired.

  tied by implantation, then by partial diffusion of the ions

  implanted to form the first layer of impurity. Any-

  times, the first layer of impurity can be formed by d'-

  other doping methods, for example by diffusion. The

  second layer of impurity is obtained by implantation of-

  ions of the second impurity. Once the two layers

  impurities are formed, the substrate is heated to redis-

  tribute and merge the first and second impurities by diffusion so that a layer of impurity results

  unique composed of both impurities.

  When using implantation techniques

  tion of ions, the two impurities are preferably implanted at a substantially identical acceleration energy. Being

  given however that the second impurity is chosen in my-

  to have a lower atomic number than the first

  impurity, the layer of the second impurity will have a concentration

  maximum tration at a depth greater than the depth

  of the maximum concentration of the layer of the first im-

  purity. It is for this reason that implantation at a substantially identical acceleration voltage leads to the

  intermediate layered configuration sought.

  It is desirable, but not essential,

  to carry out the stages of implantation of ions through

  diary of a relatively thin oxide layer formed on the surface of the substrate. The relatively thin oxide layer

  is practically transparent to the implanted ions but prevents

  a deterioration of the substrate surface which could

  result from implantation at accelerating energies

  lifted. The above objects as well as others which are iniqedatequ will appear later, / obtained according to the present invention which relates to a process for the manufacture of a P-N junction having a relatively high breaking stress exposed

  in the description below, with reference to the appendices, o

  the same references designate the same elements.

In these drawings:

  Fig. 1 is a graph illustrating the different

  between a relatively steep concentration profile

  and a relatively gradual concentration profile.

  Fig. 2 is a graph illustrating the ten-

  failure sions resulting from the concentration profiles

shown in figure 1.

  Fig. 3 is a graph of the dis-

  tribution of the first and second impurities inside

  of the substrate before the diffusion stage.

  Fig. 4 is a graph of the curve of con-

  centering in impurity after the diffusion stage.

  Fig. 5 is a sectional view of a substrate

  wearing an ion implantation mask.

  Fig. 6 is a sectional view of a substrate

  after implantation of the first impurity.

  Fig. 7 is a sectional view of a substrate

  after partial diffusion of the first impurity.

  Fig. 8 is a sectional view of a substrate

  after implantation of the second impurity.

  Fig. 9 is a sectional view of a substrate after diffusion of the first and second impurities in sight

  the formation of a single layer of impurities.

  Figure 1 shows the difference between a

  impurity layer with a periphery or boundary

  aggressive or non-abrupt, so a profile curve of dis-

  gradual or progressive tribution and a layer of impurity

  with an abrupt periphery or separation and therefore

  follow a steep distribution profile curve.

  In Figure 1, the abscissa represents the depth from

  of the surface of a P> type substrate and the ordinate shown

  feels the concentration of an impurity (P or N) which is re-

  part below the surface of the substrate.

  solid lines represent a layer of impurity with a pe-

  relatively progressive riphery or separation. The curve

  in dashed lines represents a layer of impurity with a peri-

  relatively abrupt separation or separation.

  In Figure 2, we see the representation gra-

  phique of the breaking stresses resulting from P-N junctions

  which result from impurity layers having the characteristics

  ques of the curves of figure 1. In figure 2, the abscissa represents the breaking voltage, and the ordinate the magnitude of the current passing through the junction P-N. The curve in solid lines in Figure 2 corresponds to the electrical characteristics of a P-N junction made with a layer

  impurity with a relative periphery or separation-

  progressively as shown in solid lines in FIG. 1. The dashed curve in FIG. 2 represents the electrical characteristics of a PN junction obtained from a layer of impurity having a relatively abrupt periphery or limit as shown by the dashed curve of figure 1. It is clearly seen from figure 2 that a junction with a higher breaking tension results from the use of a layer of impurity with a curve of

  relatively progressive distribution profile.

  In Figure 3, the distribution profile curves of the first and second layers are shown.

  impurity at an intermediate point in the process according to the

  before the combination of impure layers-

  you. In Figure 3, the abscissa represents the depth from

  shot from the surface of a P-type substrate, and the ordinate re-

  shows the concentration of type N impurities distributed there. Curve 10 represents the distribution profile

  of the first impurity after its implantation in the subs-

  trat but before its partial distribution. The vertical line

  12 which divides curve 10 into two equal parts representing

  te the depth, from the surface of the substrate, of the peak

  of the profile curve 10. The dashed curve 10 'represents

  te the distribution profile of the first impurity after

  that its distribution is complete, that is to say after

  partial ion ions implanted during an intermediate diffusion step. It should be noted that the step of

  intermediate diffusion brings the distribution profile curve

  bution10 to extend to an extent determined by countries

  rameters of the intermediate diffusion stage.

  Curve 14 represents the distribution profile

  bution of the second impurity after its implantation in the substrate. The vertical line 16 which divides the curve 14 into two equal parts represents the peak of the distribution curve of the second impurity. Line 16 is farther from the surface of the substrate than line 12, thus showing that the peak of the distribution curve of the second impurity is at a greater depth in the

  substrate as the peak of the distribution curve of the pre-

  my own impurity. In other words, the second layer of im-

  purity is located deeper inside the subs-

  trat as the first layer of impurity. In addition, it can be seen from FIG. 3 that the second impurity layer has an impurity concentration which is substantially less than

  that of the first layer of impurity.

  After the first and second impurity layers have formed inside the substrate, it is

  heated for the diffusion of the first and second im-

  purities, thus causing the first and second layers

  me impurities to merge into a single impurity layer whose distribution profile curve is represented by the curve 18 of Figure 4. In Figure 4.1, the abscissa represents the depth from the surface of the P-type substrate,

  while the ordinate represents the concentration of the impure-

  type N combined tees. As shown in Figure 4 a-

  near the diffusion phase which combines the first and two

  xth layers, these completely lose their identities

  individual and form a single layer of corresponding impurity

  laying on N-type impurities and exhibiting a

  relatively progressive concentration profile.

  The configuration of the profile curve

  distribution 18 of the fused or combined impurity layer

  born, in particular its border or progressive limit is largely due to the molecular characteristics of the

  second impurity, namely the diffusion coefficient re-

  laterally high of the second impurity and the fact that the second impurity can be implanted deeper in the substrate than the first. On the other hand, the insensitivity

  relative configuration of the merged layer or

  heated in subsequent heating phases is due in large

  part of the molecular characteristics of the first im-

  purity, namely its relatively low diffusion coefficient and the fact that the first impurity is distributed in a substantially higher concentration than the second impurity.

  Figures 5 to 9 illustrate the operating phases

  racks thanks to which the process of the pre-

  invention. As seen in Figure 5, the subs-

  type P silicon 20 has a surface 22 on the

  what is an ion implantation mask 24 formed.

  that ion implantation 24 can be constituted by dio-

  silicon oxide or other similar material. The mask 24

is obtained in the known manner.

  We develop or deposit a cover

  che of silicon dioxide on the surface 22 of the substrate 20 in a thickness of about 5000 Angstrams. A non-photosensitive layer (in English "photo-resist") not shown, is then formed on the surface of the layer of dioxide of

  silicon 24. A mask is placed on this non-photo-

  sensitive. The mask is transparent in all its parts, except in its part located above the area of the substrate where the P-N junction must form. The whole is then exposed to light and developed. Unexposed parts of the non-photosensitive layer and the dioxide dioxide layer

* underlying silicon 24 are then etched or attacked

  for exposure of the substrate surface 22-to-

  above the areas where the P-N junction must form. A relatively thin oxide layer 26 of thickness is then made to develop on the exposed surface 22 of the substrate.

  about 200 Angstroms so that the whole appears

  as shown in Figure 5.

  It should be noted that the silicon dioxide

  silicon, under a thickness of about 5000 Angstr6ms, is o-

  Paque the implantation of ions under low energies.

  However, when in a relatively thin layer of about 200 Angstroms, the silicon dioxide is relatively transparent to the implantation of ions. When

  assembly is exposed to ions during the im-

  planting, the thick oxide parts of the 24 em layer

1l

  thus prevent the implantation of ions in the substrate, a-

  when the relatively thin oxide layer 26 allows the

  planting of ions in the substrate below it.

  The relatively thin oxide layer 26 serves in the implantation areas to prevent damage to the surface of the

  substrate 22 which often occurs when ions are im-

  planted at relatively high acceleration energies.

  In Figure 6, we see the substrate as it

  appears after implantation of the first impurity. The pre-

  impurity is chosen so as to have an ato-

  relatively high radiation and a relative diffusion coefficient

  tively low, which requires a relatively high temperature to obtain significant diffusion. We choose

  preferably arsenic or antimony ions for the first

  re impurity. Arsenic, for example, has an atomic number

  than 33 and requires a temperature of around 11000C to obtain a substantial diffusion. Given that

  subsequent processing phases take place at a temperature

  somewhat lower, about 9500C, these have

  little effect on the location of implanted arsenic ions.

  It is preferable to implant the ar-

  senic as deeply as possible, that's why the im-

  ion planter is set near its accelerating energy

  maximum tion, approximately 100 Kev. We can however

  also use lower energies. The implant dose

  typical tation is approximately 5 x 1015 / cm2, which translates

  by a concentration of arsenic ions of approximately 5 x 1019 / cm3.

  The result of arsenic implantation is a layer of implanted arsenic ions 28 formed below the thin layer

oxide 26.

  If arsenic ions have been introduced by implantation as is preferable, the arsenic layer 28 / s partially in the silicon substrate 20 under atmosphere of ambient N2 or ambient O2 / N2 or ambient O2 at temperatures of typical in the vicinity of

  11000C. The duration of this partial diffusion stage is re-

  laterally short. When the partial diffusion phase is completed, the diffused arsenic layer 28 ′ appears as

shown in Figure 7.

  It is also possible to boost the subs-

  trat with ions of the first impurity by usual diffusion methods. When this option is chosen, the diffusion of arsenic ions takes place during doping and no

  separate partial release phase is not required.

  Phosphorus ions are now implanted

  through the same thin oxide layer 26 at maximum energy

  male of the ion implanter, approximately 100 Kev, but at a significantly lower dose than for implantation

  arsenic. Preferably, the phosphorus concentration

  sultante is about two points below the concentration

  tion in implanted arsenic. An implantation dose of phos-

  typical phore of 5 x 101 / cm is preferable, Aqui results in a phosphorus concentration of approximately 5 x 10 17 / cm3

  in the substrate. At the same acceleration energy as

  used for arsenic implantation, phosphorus is

  implanted at a greater depth below the surface

  ce 22 of the silicon substrate, because phosphorus possesses

  of a lower atomic number (15) than arsenic. The neck-

  implanted phosphorus che 30 is therefore deeper-

  lie in the substrate than the implanted arsenic layer28.

  After implantation of phosphorus, the assembly appears as

shown in figure 8.

  After the implantation of phosphorus, the whole-

  ble is exposed to conditions of diffusion at temperatures

  tures approximately equal to the diffusion temperature of the first impurity, that is to say about 11000C,

  in an oxidizing atmosphere for a predetermined period.

  It is also possible to carry out part of this diffusion phase in a nitrogen atmosphere or in a

  atmosphere consisting of a combination of oxygen and azo-

  you. The result of this heating is that the arsenic layer 28 ′ and the phosphorus layer 30 fuse or combine into a single N-type layer 32 of the two arsenic ions and

  phosphorus. The broadcast duration is determined by the pro-

  Founder of final junction sought, taking into account the

  its concentration of implanted ions. Many com-

  appropriate combinations of duration and temperatures can

  be used for broadcast. After the broadcast, the as-

  semblage appears as shown in figure 9.

  After formation of the P-N junction, the as-

  semblage is subjected to various additional operating phases

  nelles. Some of these subsequent operating phases include

  take a heating of the whole but these phases of heating

  fage usually takes place around 9500C and thus has little effect on the concentration or configuration of the fused layer 32 due to the low diffusion coefficient and the concentration of arsenic ions. However,

  the fused layer 32 has a profile curve

  relatively progressive centration, i.e. a relatively progressive periphery or limit due to phosphorus ions

more deeply implanted.

  It is understood that the present invention

  tion relates to a method of manufacturing a P-N junction in a substrate. which has a relative breaking voltage

  tively high but which allows the manufacturer to control

  or fine-tune the characteristics of this ring-

  tion P-N. This result is obtained by implanting a

  relatively high concentration of a first impurity which has a high atomic number and a low diffusion coefficient to form a first layer of impurity, by implantation of a lower concentration of a second

  impurity with a lower atomic number and a coef-

  higher diffusion factor to form a second

  che deeper than the first, and by heating the sub-

  strat for the diffusion of the impurities distributed for

  cause the layers to merge into a single layer

  both the first and second impurities. In this way,

  the merged layer has the property of the first impu-

  reté, namely that it gives the manufacturer the possibility of fine-tuning the concentration and configuration of the layer, as well as the characteristic of the second

  impurity, a relatively progressing distribution profile-

  sif. Only one embodiment of the present invention has been described above by way of example, but it is clear that many modifications and variations

can be made.

Claims (43)

  1. Method for forming an impurity layer   having a relatively pro- impurity distribution profile   compressive in a substrate, this process being characterized by the following stages:   a) distribution of first impulse ions   reté (As), of a type determining a given conductivity and having a relatively low diffusion coefficient   in the substrate (20) to form there a first layer of   purity (28) having a maximum impurity concentration at a first depth (12);   b) distribution of ions of a second impulse   reté (Ph) of a type determining said given conductivity and having a relatively high diffusion coefficient. in the substrate (20) to form a second impurity layer   (30) having a maximum impurity concentration at two xth depth (16); and   c) dissemination of the first and second impulse   selected (As, Ph) to cause the combination of said pre-   first and second layers (28.30) for training   a single impurity layer (32) composed of both   say first and second impurities (As, Ph).   2. Method according to claim 1, charac-   terized in that the second depth (16) is greater than the first depth (12).   3. Method according to claim 1, charac-   terized in that the distribution stage of the first im-   purity (As) includes the stage of distribution of the first impurity (As) at a first concentration (10), and o the stage of distribution of the second impurity (Ph) comprises   the stage of distribution of the second impurity to a second   xth concentration (14), said first concentration (10)   being higher than the second concentration (14).   4. Method according to claim 3, charac-   terized in that said first concentration (10) is su-   about two points less than said second concentra- tion (14).   5. Method according to one of claims 1   to 4, characterized in that said first impurity (As) pos-   a first atomic number and that said second impurity   reté (Ph) has a second atomic number, the first (12) and second (16) depths being a function, at least   in part, said first and second atomic numbers.   6. Method according to claim 5, charac-   terized in that said first atomic number is higher   lifted only the second atomic number.   7. Method according to any one of the res-   dications 1 to 6, characterized in that the stage of diffusion   sion comprises the stage of heating the substrate (20) to a   sufficient temperature to cause substantial diffusion   tial of said first impurity (As).   8. The method of claim 7, charac-   terized in that said temperature is at least equal to the   sufficient temperature to cause substantial diffusion   tial of said second impurity (Ph).   9. The method of claim 8, charac-   terized in that said temperature is higher than the tem-   sufficient temperature to cause substantial diffusion   tial of said second impurity (Ph).   10. Method according to any one of the re-   vendications 1 to 9, characterized in that the stage of dif-   further includes the step of adjusting the heating time to adjust the configuration of said layer single impurity (32).   11. Method according to any one of the re-   vendications 1 to 10, characterized in that the stage of dis-   tribution of said first impurity (As) includes the sta-   implantation of ions of said first impurity (As) in said substrate (20).   12. The method of claim 11, ca-   characterized in that the implantation step comprises the stages of formation of an oxide layer (24) on the surface   of the substrate, and implantation of said ions through the said oxide layer (24).   13. The method of claim 12, ca-   characterized in that the distribution stage of said pre-   the first impurity (As) further comprises the stage of heating the substrate (20) for the partial diffusion of the ions of said first impurity (As) after said implantation.   14. Method according to any one of the re-   vendications 1 to 13, characterized in that the stage of dis-   tribution of said first impurity (As) includes the sta-   of diffusion of ions of the first impurity (As) in the substrate (20).   15. Method according to any one of the re-   vendications 1 to 14, characterized in that the stage of dis-   tribution of said second impurity (Ph) includes the sta-   of implantation of ions of said second impurity (Ph) in said substrate (20).   16. A method according to any of the   vendications 1 to 15, characterized in that said first impurity (As) is arsenic.   17. Method according to any one of the re-   vendications 1 to 15, characterized in that said first impurity is antimony.   18. Method according to any one of the re-   vendications 1 to 17, characterized in that said second- impurity (Ph) is phosphorus.   19. Method according to any one of the re-   vendications 1 to 18, characterized in that the stage of dis-   tribution of said first impurity (As) includes the sta-   distribution of said first impurity (As) at a 2 dose of approximately 5 x 10 / cm   20. Method according to any one of the re-   vendications 1 to 19, characterized in that the stage of dis-   tribution of said second impurity (Ph) includes the sta-   of distributing said second impurity at a dose approximate 5 x 1013 / cm2.   21. Method according to any one of the re-   vendications 1 to 20, characterized in that the stage of dis-   tribution of said second impurity comprises the stage of distribution of said second impurity at a concentration of approximately 5 × 10 17 / cm 3   22. Method according to any one of the re-   vendications 1 to 21, characterized in that said stages of implantation of the first and second impurities (As, Ph)   are produced at substantially the same accelerating energy tion.   23. The method of claim 22, charac-   terized in that said acceleration energy is about Kev.   24. Method according to any one of the res-   dications 1 to 23, characterized in that the heating stage   ge includes the stage of heating the substrate (20) to   around 1100 ° C in an oxidizing atmosphere.   25. Method according to any one of the res-   dications 1 to 23, characterized in that the heating stage   ge includes the stage of heating the substrate (20) to   about 11000C in a nitrogen atmosphere.   26. A method according to any of the   vendications 1 to 23, characterized in that the heating stage   fage includes the stage of heating the substrate (20) to   around 1100 ° C in an oxygen and nitrogen atmosphere.   27. Method of forming a P-N junction having a relatively high breaking tension in a   substrate, this process being characterized by the following stages before:   a) implantation of ions from a first impulse   reté (As) of a type determining a given conductivity and having an atomic number and a diffusion coefficient   given, at a given acceleration energy, in the sub-   strat (20) of opposite conductivity type to form a first layer (28) having a distribution peak at a first depth (12);   b) implantation of ions of a second impulse   reté (Ph) of a type determining said given conductivity   and having an atomic number and a diffusion coefficient   higher than said first impurity (As), sensi-   said acceleration energy given, in the subs-   trat (20) so as to form a second layer (30) pre-   feeling a distribution peak at a second depth (16); and c) heating the substrate (20) to cause a substantial diffusion of the ions so as to fuse   the first and second layers (28.30) to form a layer   single conductivity type impurity (32)   by both the two impurities (As, Ph) having a pro-   relatively gradual distribution wire.   28. The method of claim 27, charac-   terified in that the ion implantation stage of a first   re impurity (As) includes the concentration adjustment stage   tion of the implanted ions of the first impurity (As). some.   29. Method according to one / of the claims   27 and 28, characterized in that the implantation stage of-   ions of a second impurity (Ph) includes the stage of settling   ge of the concentration of implanted ions of the second im- purity (Ph).   30. Method according to one of claims 27   to 29, characterized in that the concentration of implanted ions   tees of the first impurity (As) is higher than the con-   centration of implanted ions of the second impurity (Ph).   31. Process according to any of the   vendications 37 to 30, characterized in that the dose of implanted ions of the first type (As) is approximately 5 x 10 15 / cm2   32. Method according to any one of the re-   vendications 27 to 31, characterized in that the dose of ions   implanted second impurity (Ph) is approximate- 13 2 ment of 5 x 10 / cm   33. Method according to any one of the re-   vendications 27 to 32, characterized in that said first impurity (As) is arsenic.   34. Method according to any one of the re-   vendications 27 to 33, characterized in that said first impurity is antimony.   35. Method according to any one of the re-   vendications 27 to 34, characterized in that said second im- purity is phosphorus.   36. Method according to any one of claims 27 to 35, characterized in that the type determining the given conductivity is of type N.   37. Method according to any one of the res-   dications 27 to 36, characterized in that the implantation stage   tion of said first impurity (As) further comprises the stage of heating the substrate (20) for diffusion   partial of said first impurity (As).   38. Method of manufacturing a junction   P-N having a relatively high breaking tension, process   characterized by the following stages: a) masking of the surface (22) of the substrate (20) at. the exception of places on the substrate surface is   located above the areas where the ring must be formed tion P-N;   b) growth of a relative oxide layer   thinly (26) on the surface of the unmasked places;   c) implantation of first impulse ions   reté (As), at a given acceleration energy, having a relatively low diffusion coefficient and a number   relatively high atomic, through said layer of oxy-   of (26), in the substrate (20) to create a first layer impurity (28);   d) implantation of ions of a second im-   purity (Ph), at said given acceleration energy, having   a relatively high diffusion coefficient and a nom-   relatively low atomic bre, through the oxide layer   (26), towards the substrate (20) to create a two   xth impurity layer (30); and e) heating the substrate (20) for the diffusion of the first (As) and second (Ph) impurities to merge the first (28) and second (30) impurity layers to form a single impurity layer (32) of said first (As) and second (Ph) impurities having a   relatively progressive impurity profile.   39. Method according to claim 38, charac-   terized in that said oxide layer (26) has a thickness about 200 Angstroms.   40. Method according to one of claims   38 or 39, characterized in that the masking stage comprises   the stage of formation of a thick oxide layer (24) of-   about 5000 Angstroms and removal of the part of the-   said thick oxide layer (26) at the places above the which will be formed the P-N junction.   41. Method according to any one of the re-   vendications 38 to 40, characterized in that the concentration   of said first impurity (As) in the first layer of-   impurity (28) is greater than the concentration of the second-   impurity (Ph) in said second impurity layer (30).   42. Method according to any one of the re-   vendications 38 to 41, characterized in that said first   layer (28) is implanted at a depth less than this the second layer (30).   43. Method according to any one of the re-   vendications 38 to 42, characterized in that the heating stage   fage includes the stage of heating the substrate (20) to   the diffusion temperature of said first impurity (As).