Integrated capacitor and method of producing an integrated capacitor

Abstract

Integrated capacitor including a first electrode structure, a second electrode structure, and an interposed dielectric layer structure. The dielectric layer structure includes a layer combination having an SiO.sub.2 layer, an Si.sub.3N.sub.4 layer, and an Si.sub.xN.sub.y layer. The Si.sub.xN.sub.y layer includes a non-stoichiometric silicon nitride material with an increased proportion of silicon.

Claims

1. An integrated trench capacitor, comprising: a first electrode structure, a second electrode structure, and a dielectric layer structure interposed between the first electrode structure and the second electrode structure, wherein the first electrode structure comprises a semiconductor substrate provided with a trench structure, wherein the dielectric layer structure comprises a layer combination comprising an SiO.sub.2 layer, an Si.sub.3N.sub.4 layer, and an Si.sub.xN.sub.y layer, wherein the Si.sub.xN.sub.y layer comprises non-stoichiometric silicon nitride material that is silicon-rich silicon nitride material, wherein a thickness of the Si.sub.3N.sub.4 layer to a thickness of the Si.sub.xN.sub.y layer is configured to be in a ratio of n to one, wherein n is ranging from 1.5 to 2.5, and wherein the dielectric layer structure is arranged within the trench structure.

2. The integrated trench capacitor as claimed in claim 1, wherein the first electrode structure forms a rear-side electrode arranged at a rear-side contact; and wherein the second electrode structure forms a front-side electrode arranged at a front-side contact.

3. The integrated trench capacitor as claimed in claim 1, wherein a ratio of silicon to nitrogen of the Si.sub.xN.sub.y layer ranges from 0.8 to 1.

4. The integrated trench capacitor as claimed in claim 1, wherein the Si.sub.xN.sub.y layer is arranged separately from, or not directly adjacent to, the SiO.sub.2 layer.

5. The integrated trench capacitor as claimed in claim 1, wherein the dielectric layer structure of the integrated capacitor comprises an effective oxide thickness of at least 1200 nm and a dielectric strength of at least 900 V.

6. The integrated trench capacitor as claimed in claim 1, wherein a thickness of the Si.sub.xN.sub.y layer ranges from 50 nm to 2000 nm, from 50 nm to 1000 nm, from 50 nm to 500 nm, or from 100 nm to 1000 nm.

7. The integrated trench capacitor as claimed in claim 1, wherein the dielectric layer structure comprises a layer combination comprising an SiO.sub.2 layer, two Si.sub.3N.sub.4 layers, and an Si.sub.xN.sub.y layer.

8. The integrated trench capacitor as claimed in claim 7, wherein the Si.sub.xN.sub.y layer is arranged between the two Si.sub.3N.sub.4 layers.

9. The integrated trench capacitor as claimed in claim 1, wherein the dielectric layer structure comprises a layer combination comprising an SiO.sub.2 layer, an Si.sub.3N.sub.4 layer, and two Si.sub.xN.sub.y layers.

10. The integrated trench capacitor as claimed in claim 9, wherein the Si.sub.3N.sub.4 layer is arranged between the two Si.sub.xN.sub.y layers.

11. A method of producing an integrated trench capacitor, comprising: providing a first electrode structure, a second electrode structure, and a dielectric layer structure interposed between the first electrode structure and the second electrode structure, wherein the first electrode structure comprises a semiconductor substrate provided with a trench structure; and producing the dielectric layer structure within the trench structure of the semiconductor substrate, wherein the dielectric layer structure comprises a plurality or a combination of mutually adjoining dielectric layers, at least one of the dielectric layers comprising SiO.sub.2 material, at least one of the dielectric layers comprising Si.sub.3N.sub.4 material, and at least one of the dielectric layers comprising Si.sub.xN.sub.y material, wherein the Si.sub.xN.sub.y material comprises non-stoichiometric silicon nitride material that is silicon-rich silicon nitride material, and wherein a thickness of the Si.sub.3N.sub.4 layer to a thickness of the Si.sub.xN.sub.y layer is configured to be in a ratio of n to one, wherein n is ranging from 1.5 to 2.5.

12. The method as claimed in claim 11, wherein the dielectric layer structure comprises a plurality or a combination of a dielectric layer comprising SiO.sub.2 material, a dielectric layer comprising Si.sub.3N.sub.4 layer material, and a dielectric layer comprising Si.sub.xN.sub.y material.

13. The method as claimed in claim 12, wherein a ratio of silicon to nitrogen of the dielectric layer comprising Si.sub.xN.sub.y material ranges from 0.8 to 1.

14. The method as claimed in claim 11, wherein the dielectric layer structure comprises a higher proportion of the Si.sub.3N.sub.4 material than of any material deviating therefrom.

15. The method as claimed in claim 11, wherein the dielectric layer comprising Si.sub.xN.sub.y material is not arranged directly at the dielectric layer comprising SiO.sub.2 material.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Embodiments of the present invention will be detailed subsequently referring to the appended drawings, in which:

(2) FIG. 1 shows a schematic representation of an integrated capacitor, in accordance with an embodiment of the present invention;

(3) FIG. 2a shows a schematic representation of an integrated capacitor comprising a planar dielectric layer structure comprising four layers in accordance with an embodiment of the present invention;

(4) FIG. 2b shows a schematic representation of an integrated capacitor comprising a dielectric layer structure comprising four layers and a trench structure in accordance with an embodiment of the present invention;

(5) FIG. 3 shows a schematic representation of an integrated capacitor configured as a trench capacitor in accordance with an embodiment of the present invention;

(6) FIG. 4a shows a picture, on a scanning electron microscope, of a first electrode structure comprising a semiconductor substrate provided with a trench capacitor, in accordance with an embodiment of the present invention;

(7) FIG. 4b shows a picture, on a scanning electron microscope, of a cross section of an integrated capacitor as a trench capacitor, in accordance with an embodiment of the present invention;

(8) FIG. 5a shows a schematic representation of a top view of a trench structure of a semiconductor substrate of a first electrode structure of an integrated capacitor, in accordance with an embodiment of the present invention;

(9) FIG. 5b shows a schematic representation of a cross section of a hole structure of a trench structure of a semiconductor substrate of an electrode structure of an integrated capacitor, in accordance with an embodiment of the present invention;

(10) FIG. 5c shows a schematic equivalent circuit diagram of an integrated capacitor, in accordance with an embodiment of the present invention;

(11) FIG. 6a shows a diagram of a distortion of silicon semiconductor wafers of different hole geometries in the event of individual dielectric layers being deposited, in accordance with an embodiment of the present invention;

(12) FIG. 6b shows a table of the dielectric layer structures whose distortion of the silicon semiconductor wafer is shown across the individual deposition processes of the dielectric layer structure in FIG. 6a, in accordance with an embodiment of the present invention;

(13) FIG. 7 shows a diagram of a capacitance/voltage characteristic of an integrated capacitor, in accordance with an embodiment of the present invention;

(14) FIG. 8a shows a diagram of current/voltage characteristics of several integrated capacitors, in accordance with an embodiment of the present invention, as compared to a capacitor having a dielectric made of silicon dioxide and exclusively silicon-rich nitride;

(15) FIG. 8b shows a table of the dielectric layer structures, of the integrated capacitors, whose current/voltage characteristics are depicted in FIG. 8a, in accordance with an embodiment of the present invention; and

(16) FIG. 9 shows a block diagram and a schematic representation of a sequence of steps of producing an integrated capacitor, in accordance with an embodiment of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

(17) Before embodiments of the present invention will be explained in more detail below with reference to the drawings, it shall be noted that elements, objects and/or structures in the various figures which are identical or identical in function or in action have been provided with identical or similar reference numerals, so that the descriptions of said elements that are provided in the various embodiments are interchangeable and/or mutually applicable.

(18) FIG. 1 shows an embodiment of an integrated capacitor 100 comprising a first electrode structure 110 and a second electrode structure 120. In addition, the integrated capacitor 100 comprises a dielectric layer structure 130 arranged between the first electrode structure 110 and the second electrode structure 120. The dielectric layer structure 130 comprises a layer combination having an SiO.sub.2 layer 132, an Si.sub.3N.sub.4 layer 134, and an Si.sub.xN.sub.y layer 136. The Si.sub.xN.sub.y layer 136 comprises a non-stoichiometric silicon nitride material having an increased proportion of silicon. In accordance with an embodiment, the proportion of silicon of the Si.sub.xN.sub.y layer 136 is increased as compared to the proportion of silicon of the Si.sub.3N.sub.4 layer 134.

(19) In accordance with an embodiment, the first electrode structure 110 may comprise a semiconductor substrate. The semiconductor substrate may comprise p-doped silicon material. Doping of the silicon substrate contributes to increasing conductivity in a targeted manner. In accordance with an embodiment, the semiconductor substrate is doped with boron (B), an element of the third main group of the periodic table.

(20) The second electrode structure 120 comprises polycrystalline silicon material (polysilicon), in accordance with an embodiment. Said material is a highly n-doped polycrystalline silicon, for example. Just like doping of the silicon substrate of the first electrode structure 110, this serves the purpose of increasing conductivity of the high-ohmic silicon. For example, phosphorus and argon, which are elements of the fifth main group of the periodic table, are used for doping.

(21) In accordance with an embodiment, the first electrode structure 110 forms a rear-side electrode arranged at a rear-side contact, and the second electrode structure 120 forms a front-side electrode arranged at a front-side contact.

(22) The core of the present invention lies in using a dielectric layer stack, i. e. the dielectric layer structure 130, consisting of a layer combination comprising an SiO.sub.2 layer 132, an Si.sub.3N.sub.4 layer 134, and an Si.sub.xN.sub.y layer 136. Within this context, the number and sequence of the partial layers 132 to 136, their layer thicknesses and the quantity ratio of silicon to nitrogen (x:y) in Si.sub.xN.sub.y have a decisive influence on the producibility (in particular, reduction of a semiconductor wafer distortion and delamination) and the electric properties (in particular, increase of a breakdown voltage and capacitance density) of the integrated capacitor 100.

(23) The silicon dioxide SiO.sub.2 layer 132 exhibits the advantages of straightforward manufacturing and good insulation properties. In accordance with an embodiment, the SiO.sub.2 layer 132 is arranged, due to lower interface states, within the dielectric layer structure 130, as a dielectric layer at an interface with the first electrode structure 110, e. g. the silicon substrate. So as to at least partly avoid delamination of the dielectric layer structure 130, the silicon dioxide layer 132 is arranged, in accordance with an embodiment, between the first electrode structure 110 and a silicon nitride layer such as the Si.sub.3N.sub.4 layer 134 or the Si.sub.xN.sub.y layer 136, for example.

(24) In accordance with an embodiment, the Si.sub.3N.sub.4 layer 134 comprises stoichiometric silicon nitride Si.sub.3N.sub.4 having a permittivity of 7.5, as a result of which it is possible to achieve higher layer thicknesses and, thus, increased dielectric strength of the integrated capacitor 100 while achieving higher capacitance as compared to using exclusively silicon dioxide within the dielectric. With stoichiometric silicon nitride, the stoichiometric ratio of the elements of silicon (Si) and nitrogen (N) is predefined as three to four.

(25) In accordance with an embodiment, manufacturing of semiconductor devices comprising stoichiometric silicon nitride gives rise to mechanical tensions scaling with the layer thickness. So as to bypass limitation of the thickness of the dielectric layer on account of mechanical load, e. g., low-stress silicon nitride Si.sub.xN.sub.y is integrated into the dielectric layer structure 130. Thus, the dielectric strength of the integrated capacitor 100 is increased by means of the low-stress silicon nitride since the increase in the thickness of the dielectric results in a shift of the dielectric breakdown toward higher voltages.

(26) The Si.sub.xN.sub.y layer 136 comprises non-stoichiometric silicon nitride material, also referred to as low-stress silicon nitride Si.sub.xN.sub.y. The non-stoichiometric silicon nitride is composed of a modified ratio of silicon to nitrogen in relation to the Si.sub.3N.sub.4 material. By using low-stress silicon nitride, larger layer thicknesses may be implemented since the intrinsic tensions of the layer are smaller than intrinsic tensions in stoichiometric silicon nitride.

(27) In accordance with an embodiment, a ratio of silicon to nitrogen of the Si.sub.xN.sub.y layer 136 ranges from 0.8 to 2, from 0.8 to 1.5, or from 0.8 to 1. With this ratio, the Si.sub.xN.sub.y layer 136 comprises very little mechanical tension within the layer. By combining this Si.sub.xN.sub.y layer 136 with the Si.sub.3N.sub.4 layer 134, the mechanical tension of the entire dielectric layer structure 130 is lower, in accordance with an embodiment, than a sum of the individual tensions of the individual layers of the dielectric layer structure 130.

(28) A disadvantage of the low-stress silicon nitride Si.sub.xN.sub.y is the fact that its defect concentration is higher than in stoichiometric nitride Si.sub.3N.sub.4. Due to the higher defect concentration in the Si.sub.xN.sub.y layer 136, utilization of an Si.sub.3N.sub.4 layer 134 that is in contact with the upper electrode, i.e. the second electrode structure 120, is implemented in accordance with an embodiment, or combinations with a thin Si.sub.xN.sub.y layer 136 in contact with the second electrode structure 120, which have a low leakage current. However, it is to be taken into account that a thicker Si.sub.xN.sub.y layer 136 results in a higher leakage current, for example. By means of the inventive combination of Si.sub.3N.sub.4 134 and Si.sub.xN.sub.y 136, the advantages of the low leakage current (due to the Si.sub.3N.sub.4 layer 134) and of the reduced mechanical stress (due to the Si.sub.xN.sub.y layer 136) may be united. In accordance with an embodiment, it is to be taken into account here that upon deposition of different silicon nitride layers (e. g. of the Si.sub.3N.sub.4 layer 134 and of the Si.sub.xN.sub.y layer 136), the mechanical stress present in the entire layer stack, i. e. in the entire dielectric layer structure 130, does not match the sum of the individual mechanical tensions. For implementing a low-stress overall dielectric, i. e. the dielectric layer structure 130, it is therefore sufficient to replace a smaller part, in percentage, than might initially be expected by stress-free nitride Si.sub.xN.sub.y.

(29) In accordance with an embodiment, the Si.sub.xN.sub.y layer 136 is arranged separately from, or not directly adjacent to, the SiO.sub.2 layer so as to at least partly avoid delamination of the layers. However, this also depends on the overall thickness of the dielectric layer structure 130 to be implemented and will be problematic, for example, with an Si.sub.3N.sub.4 layer 134 of a thickness of 1000 nm on an Si.sub.xN.sub.y layer 136 of a thickness of 500 nm. All of the layer thicknesses indicated herein may be understood to be TARGET data. ACTUAL data may deviate from said TARGET data. For example, deviations of up to 10%, of up to 5% or of up to 2% from the TARGET data, which are due to, e. g., production- or design-related reasons, may occur.

(30) In accordance with an embodiment, the dielectric layer structure 130 comprises a higher proportion of an Si.sub.3N.sub.4 material than of any material deviating therefrom. For example, the dielectric layer structure 130 may comprise more Si.sub.3N.sub.4 layers 134 than SiO.sub.2 layers 132 and Si.sub.xN.sub.y layers 136. Alternatively, the Si.sub.3N.sub.4 layer 134 may be configured to be thicker than the SiO.sub.2 layer 132 and/or than the Si.sub.xN.sub.y layer 136, with equal expansions of the mutually adjoining faces of the individual layers.

(31) In accordance with an embodiment, an expansion d.sub.Si.sub.3.sub.N.sub.4 131.sub.2 of the Si.sub.3N.sub.4 layer 134, which is perpendicular to mutually adjoining faces of the layers of the dielectric layer structure 130, to an expansion d.sub.Si.sub.x.sub.N.sub.y 131.sub.1 of the Si.sub.xN.sub.y layer 136, which is perpendicular to mutually adjoining faces of the layers of the dielectric layer structure 130, is configured in a ratio N:1, N ranging from 1.5 to 2.5. Thus, the layer stack comprises a higher proportion of Si.sub.3N.sub.4, such as the ratio d.sub.Si.sub.3.sub.N.sub.4 d.sub.Si.sub.x.sub.N.sub.y=2:1. The expansions 131.sub.1 to 131.sub.3, which are perpendicular to mutually adjoining faces of the layers of the dielectric layer structure, are the thicknesses of the individual layers 132 to 136, for example. In accordance with an embodiment, therefore, the thickness 131.sub.2 of the Si.sub.3N.sub.4 layer 134 is double the thickness 131.sub.1 of the Si.sub.xN.sub.y layer 136, or, if the integrated capacitor 100 comprises a dielectric layer structure 130 having more than the three layers 132 to 136, an overall thickness of all of the Si.sub.3N.sub.4 layers 134 may be double an overall thickness of all of the Si.sub.xN.sub.y layers 136.

(32) So as to further increase the capacitance of the integrated capacitor 100, a surface area enlargement may be implemented in that the first electrode structure 110 comprises a semiconductor substrate provided with a trench structure.

(33) The advantage of the inventive implementation is, in accordance with an embodiment, in realizing monolithically integrated capacitors 100 comprising high dielectric strength and/or high capacitance density
on account of the combined utilization of Si.sub.3N.sub.4 134 and Si.sub.xN.sub.y 136. The advantage results from a reduction of the mechanical stress, with a simultaneously small leakage current, that is due to the dielectric layer stack, i. e. the dielectric layer structure 130. This enables implementing low-cost silicon capacitors with clearly increased dielectric strength. Optional features of the invention will be summarized in Table 2.

(34) TABLE-US-00002 TABLE 2 (FEATURES OF EMBODIMENTS OF THE INVENTION): Feature Potential embodiment Surface enlargement K 2 ≤ K ≤ 20 Number N of nitride layers 2 ≤ N ≤ 3 Order of layers 1 SiO.sub.2 − Si.sub.3N.sub.4 − Si.sub.xN.sub.y Order of layers 2 SiO.sub.2 − Si.sub.3N.sub.4 − Si.sub.xN.sub.y − Si.sub.3N.sub.4 Order of layers 3 SiO.sub.2 − Si.sub.xN.sub.y − Si.sub.3N.sub.4 − Si.sub.xN.sub.y Layer thickness ratio d.sub.Si.sub.x.sub.N.sub.y ≤ 0.5 .Math. d.sub.Si.sub.3.sub.N.sub.4 Quantity ratio in Si.sub.xN.sub.y 0.8 ≤ x : y ≤ 1

(35) In accordance with an embodiment of Table 2, the layer thickness d.sub.Si.sub.x.sub.N.sub.y of the Si.sub.xN.sub.y layer amounts to a maximum of 50% of the layer thicknesses of all Si.sub.3N.sub.4 layers. Within this context, a minimum layer thickness d.sub.Si.sub.x.sub.N.sub.y of the Si.sub.xN.sub.y layer may correspond to a lower limit of producibility. For example, the minimum layer thickness d.sub.Si.sub.x.sub.N.sub.y of the Si.sub.xN.sub.y layer may amount to, e. g., 3% of the layer thicknesses of all of the Si.sub.3N.sub.4 layers. This means that the following relationship may apply for the layer thickness ratio, in accordance with an embodiment: 0.03.Math.d.sub.Si.sub.3.sub.N.sub.4≤d.sub.Si.sub.x.sub.N.sub.y≤0.5.Math.d.sub.Si.sub.3.sub.N.sub.4.

(36) A surface structure for increasing the surface area, and the architecture of the dielectric layer stack (number and thicknesses of the partial layers) may be determined via a cross-section analysis performed by means of a scanning electron microscope. For analyzing the quantity ratios of silicon and nitrogen and, eventually, the order of the partial layers, methods such as energy-dispersive X-ray spectroscopy or secondary-ion mass spectroscopy are possible.

(37) Due to its extremely low parasitic series inductance, the silicon capacitor described is extremely suitable as a back-up capacitor or attenuating element for voltage peaks or high-frequency oscillations (RC snubber) in switching applications involving very short switching times.

(38) In accordance with an embodiment, the integrated capacitor 100 may be operated at operating voltages of up to 900 V or up to 1200 V on account of the combination of the Si.sub.3N.sub.4 layer 134 and the Si.sub.xN.sub.y layer 136.

(39) Each of FIG. 2a and FIG. 2b shows a schematic representation of an inventive integrated capacitor 100 in accordance with an embodiment of the present invention. The integrated capacitor 100 in FIG. 2a and FIG. 2b may comprise the same features and functionalities as those of the integrated capacitor 100 of FIG. 1, the integrated capacitor 100 of FIGS. 2a and 2b differing from the integrated capacitor of FIG. 1 in that the dielectric layer structure 130 comprises three instead of two silicon nitride layers 133.sub.1 to 133.sub.3 in addition to the SiO.sub.2 layer 132. The alternative dielectric layer structure 130 is arranged between the first electrode structure 110 and the second electrode structure 120. The integrated capacitors shown in FIG. 2a and in FIG. 2b differ from each other in that the integrated capacitor 100 in FIG. 2a may represent a plate capacitor, and the integrated capacitor in FIG. 2b may represent a trench capacitor.

(40) What is characteristic for dielectrics is low conductivity and, therefore, high resistivity. For employing the material in a capacitor, high permittivity results in a high capacitance value. The breakdown field strength of the dielectric material is a measure of the breakdown strength of the dielectric and, therefore, of the dielectric strength of the capacitor 100. In accordance with an embodiment, in the integrated capacitor 100, the dielectric is represented by the dielectric layer structure 130. The dielectric of the inventive integrated capacitor 100 thus is present as a multi-layer system comprised of different dielectric materials.

(41) In accordance with an embodiment, the silicon dioxide layer 132 is located at an interface with a small p-doped silicon substrate (i. e. with the first electrode structure 110 in FIG. 2a and/or with the substrate 113 in FIG. 2b). This is followed by a combination of three layers 133.sub.1 to 133.sub.3, e. g. of comparable thicknesses, of two silicon nitrides 134, 136. Thus, the silicon nitride layers 133.sub.1 to 133.sub.3 may comprise stoichiometric silicon nitride Si.sub.3N.sub.4 134 and/or low-stress silicon nitride Si.sub.xN.sub.y 136, which exhibit different electrical and mechanical properties.

(42) In accordance with an embodiment, for example, layers 133.sub.1 and 133.sub.3 each comprise an Si.sub.3N.sub.4 layer 134, and the layer 133.sub.2 comprises an Si.sub.xN.sub.y layer 136. Alternatively, any other two layers among the three layers 133.sub.1 to 133.sub.3 may represent the Si.sub.3N.sub.4 layer 134, and a third layer of the three layers 133.sub.1 to 133.sub.3 may represent the Si.sub.xN.sub.y layer 136. Thus, e. g., two of the three silicon nitride layers 133.sub.1 to 133.sub.3 are Si.sub.3N.sub.4 layers 134, and one of the three silicon nitride layers 133.sub.1 to 133.sub.3 is an Si.sub.xN.sub.y layer 136. For example, the first layer 133.sub.1 may be an Si.sub.3N.sub.4 layer 134, the second layer 133.sub.2 may be an Si.sub.xN.sub.y layer 136, and the third layer 133.sub.3 may be an Si.sub.3N.sub.4 layer 134, such as in the previously described order of layers 2 in Table 2.

(43) In accordance with an embodiment, alternatively, two of the three silicon nitride layers 133.sub.1 to 133.sub.3 each comprise an Si.sub.xN.sub.y layer 136, and one of the three silicon nitride layers 133.sub.1 to 133.sub.3 comprises an Si.sub.3N.sub.4 layer 134, such as the previously described order of layers 3 in Table 2, wherein the first silicon nitride layer 133.sub.1 and the third silicon nitride layer 133.sub.3 comprise an Si.sub.xN.sub.y layer 136, and the second silicon nitride layer 133.sub.2 comprises the Si.sub.3N.sub.4 layer 134.

(44) In accordance with an embodiment, the dielectric layer structure 130 may comprise further silicon nitride layers 133.sub.1 to 133.sub.3 so as to further increase the thickness of the dielectric of the integrated capacitor 100 and, thus, also to increase the dielectric strength of the integrated capacitor 100.

(45) FIG. 3 shows a schematic representation of an integrated capacitor 100 in accordance with an embodiment of the present invention. In accordance with an embodiment, a first electrode structure 110 of the capacitor 100 comprises a semiconductor substrate provided with a trench structure. The surface of the trench structure of the first electrode structure 110 has, e. g., a dielectric layer structure 130 of the integrated capacitor 100 arranged thereat. Thus, the dielectric layer structure 130 is arranged between the first electrode structure 110 and a second electrode structure 120.

(46) In accordance with an embodiment, the first electrode structure 110 forms a rear-side electrode arranged at a rear-side contact 112, and the second electrode structure 120 forms a front-side electrode arranged at a front-side contact 122.

(47) On account of the trench structure of the first electrode structure 110, the surface thereof is increased by the factor of K. The concept of a silicon trench capacitor (or silicon RC snubber) is known, inter alia, from [2]. Thus, in accordance with

(48) $\begin{matrix} C = ϵ_{0} .Math. ϵ_{r} .Math. \frac{K .Math. A}{d}, & (1) \end{matrix}$

(49) a higher capacitance density is achieved as compared to a planar capacitor. In the formula, C corresponds to the capacitance value, ϵ.sub.0 corresponds to the electric field constant, ϵ.sub.r corresponds to permittivity, and A corresponds to the surface of the dielectric, and d corresponds to the effective oxide thickness of the dielectric; the effective overall thickness d of the dielectric may include the entire effective silicon nitride thickness 139 and the effective thickness 131.sub.3 of the SiO.sub.2 layer 132. Components of said technology have so far been published to have operating voltages of up to 600 V only. In accordance with an embodiment, the integrated capacitor 100 (which may also be referred to as a silicon trench capacitor) comprising the specific combination of an Si.sub.3N.sub.4 layer and an Si.sub.xN.sub.y layer in the dielectric layer structure 130 presents a concept of operating voltages of up to 900 V or up to 1200 V. The core of the concept is a multi-layer dielectric (i. e., the dielectric layer structure 130) comprising at least one layer of silicon-rich silicon nitride for reducing the mechanical stress.

(50) Scaling of the dielectric strength of the capacitor 100 (e. g. voltage class of 900 V) is effected by the effective oxide thickness of the dielectric 130. A breakdown voltage U.sub.max follows a relationship of a material-specific critical field strength E.sub.crit and a given effective oxide thickness d of the dielectric 130:
U.sub.max=E.sub.crit.Math.d.sub.eff (2)

(51) However, as the thickness of the dielectric layer increases, so does its intrinsic mechanical tension. On account of the specifically developed dielectric layer structure 130, the mechanical tension present in the capacitor 100 and/or in the layers of the dielectric layer structure 130 may be kept small as the overall thickness 139 is increased, whereby, during manufacturing of the capacitor 100, pronounced distortions or breakage of the semiconductor substrate or delamination of the dielectric 130 itself may be at least partly prevented.

(52) As can be seen from equations (1) and (2), the breakdown voltage of the capacitor 100 is directly proportional to the layer thickness 139 of the dielectric 130, whereas the capacitance is inversely proportional to it—breakdown voltage and capacitance density exhibit mutually inverse behaviors in terms of their respective dimensioning as a function of d. To achieve the original capacitance density (that was present prior to scaling of the dielectric strength) by means of a corresponding surface area enlargement is not possible if SiO.sub.2 and Si.sub.3N.sub.4 are used exclusively since the mechanical stress additionally highly scales with the surface of the dielectric. The specific combination of Si.sub.3N.sub.4 and Si.sub.xN.sub.y, however, effectively reduces mechanical stress.

(53) In accordance with an embodiment, e. g., a semiconductor substrate, or a layer of polysilicon, serves as an electrode 110, 120. The contacts 112, 122 are implemented with aluminum, for example.

(54) FIG. 3 is a schematic cross section through an integrated capacitor 100, wherein, in accordance with an embodiment, the trench structure of the first electrode structure 110 is implemented by means of rectangular recesses. Even though the dielectric layer structure 130 is shown to be one single layer, the latter may comprise several layers such as an SiO.sub.2 layer, an Si.sub.3N.sub.4 layer, and an Si.sub.xN.sub.y layer.

(55) In the following, further embodiments of FIG. 3 will be presented in different words: The capacitor 100 is, e. g., a silicon capacitor which may be successfully manufactured and characterized to have a dielectric strength of 1200 V. It was possible to reduce the high mechanical loads acting during processing by using low-stress silicon nitride in a dielectric stack, i. e. in the dielectric layer structure 130. In accordance with an embodiment, stoichiometric and low-stress silicon nitride were combined to form a multi-layer system, i. e. the dielectric layer structure 130 comprising three layers, at least one layer consisting of low-stress silicon nitride.

(56) With the aid of a hexagonal hole structure, the surface area of the capacitors 100 was enlarged, in accordance with an embodiment. Any combinations of the layer stacks may be implemented, in accordance with an embodiment, to have a small hole depth L10, or selected layer stacks may be implemented to have a hole design L20 of a larger depth.

(57) In accordance with an embodiment, the different material combinations in the dielectric layer structure 130 hardly have any influence on the capacitance values of the silicon capacitors 100. However, in accordance with an embodiment, it was possible to clearly increase the capacitance per area unit by about 80% because of an enlarged surface area caused by a deeper hole design L20 as compared to a smaller hole depth L10. The curve of the current/voltage characteristic of the components, i. e. of the capacitors 100, depends on the different dielectric layer stacks 130. This follows a system according to which, in accordance with an embodiment, a larger proportion of stoichiometric silicon nitride results in a larger maximum voltage that is achieved with a current flow of, e. g., 10 mA. The lower dielectric strength in layer stacks 130 which have a higher proportion of low-stress silicon nitride is attributed to, e. g., tunnel mechanisms within the dielectric. Due to the assumed higher trap concentration in the low-stress silicon nitride as compared to the stoichiometric silicon nitride, said charge transport mechanisms will dominate, e. g., already as from relatively low field strengths. The trap concentration was determined by means of temperature-dependent current/voltage measurements as a function of the electric field strength. In accordance with an embodiment, said measurements of the low-stress silicon nitride largely match those of the stoichiometric silicon nitride.

(58) With silicon capacitors 100 which exhibit a dielectric layer stack 130 of silicon dioxide, two stoichiometric silicon nitride layers, and a low-stress silicon nitride layer, it was possible to achieve a maximum voltage of 1575 V at 10 mA.

(59) To complete electrical characterization with regard to the different dielectric layer stacks 130 it is useful to determine the load limits of the silicon capacitors 100 as well as their long-range stabilities. In accordance with an embodiment, a series resistance of the silicon capacitor is independent of the architecture of the dielectric layer stack.

(60) By using the low-stress silicon nitride, e. g., a degree of freedom is obtained with regard to the mechanical strains. On account of the findings from metrological documentation of the distortion of the silicon semiconductor wafer, i. e. of the first electrode structure 110, an enlargement of the surface area by means of deeper holes with a simultaneous layer composition having a hole design L30 (surface area enlargement more pronounced than with L20) and larger is possible. This goes hand in hand with increased capacitance per area unit. The hole depth may be further increased by an increased proportion of low-stress silicon nitride. Thus, silicon capacitors having high capacitances per area unit may be implemented. However, in accordance with an embodiment, the dielectric strengths of components having high proportions of low-stress silicon nitride are lower.

(61) In accordance with an embodiment, the capacitor 100 represents an optimization with regard to a minimum proportion of low-stress silicon nitride and, thus, a maximum possible dielectric strength.

(62) Further optional details of the trench structure of the first electrode structure 110 will be explained by means of FIGS. 4a, 4b, 5a, 5b, and 5c.

(63) FIG. 4a shows a picture of a breaking edge of a silicon semiconductor wafer 110 comprising a hole structure 111.sub.1 to 111.sub.5 following a dry-etching process, said wafer 110 serving, in accordance with an embodiment of the present invention, as the first electrode structure of the integrated capacitor. The hole structure 111.sub.1 to 111.sub.5 may be implemented, for example, by means of an alternating dry-etching process such as an ASE (advanced silicon etching, reactive ion etching) process. As depicted in FIG. 4a, for example, this process will result in recesses 111.sub.1 to 111.sub.5. The recesses 111.sub.1 to 111.sub.5 may comprise cylindrical shapes with a sphere-shaped rounding at the end in the semiconductor substrate 110. Since FIG. 4a is a schematic cross section of the semiconductor substrate 110, the recesses 111.sub.1 to 111.sub.5 are depicted as rectangular recesses having rounded corners.

(64) FIG. 4b shows a picture, on the scanning electron microscope, of a cross section of an integrated capacitor 100 in accordance with an embodiment. Therefore, a first electrode structure 110, a second electrode structure 120, and an interposed dielectric layer structure 130 become clear in FIG. 4b. Moreover, a front-side contact 122 is arranged on the second electrode structure 120.

(65) By means of pictures on the scanning electron microscope as are shown, e. g., in FIG. 4a and/or FIG. 4b, structural parameters of the hole geometries of the trench structure of the first electrode structure 110 may be determined.

(66) The trench structure of the semiconductor substrate 110 may comprise a multitude of recesses which may be identified, inter alia, in FIG. 4b by the depressions 111.sub.1 to 111.sub.n in the front-side contact 122, n being a positive integer. In accordance with an embodiment, the recesses defining the trench structure of the semiconductor substrate 110 are circular holes. In accordance with the embodiment of the integrated capacitor 100 that is presented in FIG. 4b, the semiconductor substrate 110 comprises a hexagonal arrangement of the circular holes, which is indicated by the depressions 111.sub.1 to 111.sub.n present on the front-side contact 122. This type of arrangement of holes provides, e. g., a very pronounced increase in the capacitance value of the integrated capacitor 100.

(67) FIG. 5a is a schematic top view of a hexagonal hole structure as depicted for the integrated capacitor 100 of FIG. 4b. Moreover, FIG. 5b shows a cross section through a hole of the trench structure of the integrated capacitor along a cutting edge Q-Q in FIG. 5a. FIG. 5a and FIG. 5b serve to define parameters for calculating a surface area enlargement of the integrated capacitor by means of the hexagonally arranged hole structure. In said determining step one assumes, e. g., that the layer deposition of the dielectric materials of the dielectric layer structure 130 across the hole 111.sub.1 to 111.sub.7 is even. For calculating the enlargement factor of the surface area of the integrated capacitor, one takes the mean of a thickness d 138 of the dielectric layer structure 130. In accordance with FIGS. 5a and 5b, this serves as an orientation point for a distance a between the holes 111.sub.1 to 111.sub.7, a diameter D of the holes 111.sub.1 to 111.sub.7, and a depth h of the holes 111.sub.1 to 111.sub.7.

(68) A capacitance C.sub.hole structure in an equilateral triangle 102 is composed of a capacitance of half a cylinder C.sub.half cylinder, added to a capacitance of a planar face between the holes C.sub.planar−C.sub.semicircle and to a capacitance of a bottom in the hole C.sub.bottom. Thus, the following applies, in accordance with an embodiment, for the enlargement factor K of the surface area:

(69) $k = \frac{c_{hole structure}}{c_{planar}} = 1 + \frac{4 π a}{\sqrt{3} a^{2}} .Math. (\frac{(d - D)}{2} + \frac{h - d}{\ln (\frac{D}{D - 2 d})}) .$

(70) In accordance with an embodiment, the structural parameters of the hole geometry may be predetermined. For example, the distance a between the holes may range from 1 μm to 5 μm, 2 μm to 3 μm, or 2.4 μm to 2.8 μm. The diameter D of the holes may range, in accordance with an embodiment, from 3 μm to 10 μm, from 4.5 μm to 6.5 μm, or from 5 μm to 6 μm. The depth h of the holes may range from 5 μm to 50 μm, from 10 μm to 40 μm, or from 10 μm to 35 μm. For example, a first hole design L10 may be implemented with a distance a of 2.75 μm, a diameter D of 5.35 μm, and a depth h of 12.3 μm. Alternatively, a hole design L20 may be implemented with a distance a of 2.48 μm, a diameter D of 5.59 μm, and a depth h of 22.5 μm. In addition, in accordance with an embodiment, a hole design L30 having a depth of about 30 μm may be realized.

(71) FIG. 5c shows an equivalent circuit diagram in accordance with an embodiment of an integrated capacitor 100 comprising partial capacitances of different layers and structures of a dielectric layer structure 130 and the individual resistances. The capacitance value measured is composed of individual capacitances of a parallel circuit along the dielectric layer boundaries and of a series circuit in accordance with the layer architecture. Within this context, e. g., the influences of a resistive voltage divider are neglected.

(72) The dielectric layer structure 130 depicted in FIG. 5c may comprise the same features and properties as those of the dielectric layer structure 130 of FIG. 2a or FIG. 2b. For example, the integrated capacitor 100 in FIG. 5c comprises an SiO.sub.2 layer 132, a first silicon nitride layer 133.sub.1, a second silicon nitride layer 133.sub.2, and a third silicon nitride layer 133.sub.3. Moreover, the integrated capacitor 100 of FIG. 5c comprises, in accordance with an embodiment, a first electrode structure 110, a second electrode structure 120, a rear-side contact 112, and a front-side contact 122.

(73) During deposition of dielectric layers on the silicon substrate (the first electrode structure 110), mechanical tensions arise which affect the substrate. Said internal tensions of the deposited layers may be attributed to thermally induced tensions, on the one hand, and to intrinsic tensions, on the other hand. The different expansion coefficients of silicon substrate and of deposited layers 132 and 133.sub.1 to 133.sub.3 are mainly responsible for thermally induced tensions and are due to, e. g., the large temperature difference between a process temperature and an ambient temperature. Intrinsic tensions are to be attributed, inter alia, to foreign atoms, which substitute for atoms of the layer material or occupy interstitial positions. The lattice mismatch due to the different lattice constants between the different materials is, e.g., a further cause of intrinsic tensions.

(74) The distortion of a silicon semiconductor wafer serving as a first electrode structure of a multiplicity of embodiments of the integrated capacitor 100 is a measurable quantity which allows making a statement about the extent of the inner tension within the dielectric layer structure 130. One distinguishes, e. g., between tensile stress, which causes concave distortion and/or a positive radius of curvature, and compressive stress, which causes convex distortion and/or a negative radius of curvature. Distortion of the silicon semiconductor wafer is measured both in parallel with and orthogonally to the flattening at the edge of the silicon semiconductor wafer. Its distortions exhibit congruent behavior, which is why what is documented in FIG. 6a are exclusively the maximum distortions of parallel measurements of the silicon semiconductor wafers for the sake of increased clarity.

(75) FIG. 6a shows a diagram depicting the distortion 114 of the silicon semiconductor wafers of different hole geometries upon deposition of the individual dielectric layers. In accordance with the embodiments depicted in FIG. 6a, the capacitors are inventive capacitors comprising trench structures, as in embodiments of FIGS. 3 to 5c, for example. The table of FIG. 6b depicts the layer architecture of the respective dielectric layer structures 130.sub.1 to 130.sub.7 examined in FIG. 6a. The differences between the hole designs L10, L20 and L30 was already discussed above in connection with FIGS. 5a and 5b. In accordance with an embodiment, FIG. 6a depicts distortions ranging from 0 μm to 200 μm (the axis of the distortion 114 is split up, e. g., in steps of 50 μm). FIG. 6a depicts only the distortions of specific embodiments, and it is clear that with alternative dielectric layer structures, distortions within other ranges may also be implemented. It is to be pointed out that the capacitor described herein is not limited to the parameters depicted in FIG. 6a and FIG. 6b

(76) As can be seen from the measurements of the wafer distortion, processing that has failed because of mechanical stress does not necessarily correlate with wafer distortion. For example, the wafers having the design 130.sub.1 will break even though they do not exhibit the maximum distortion. However, wafer distortion represents a limit for those process steps wherein the wafer may be “attached” to the chuck under vacuum (e. g. polyimide) or wherein a robot conveys the wafers.

(77) In accordance with an embodiment, the inventive capacitor represents a concept for silicon capacitors having a dielectric strength of 1200 V. Within this context, one has found, with regard to the mechanical properties of the dielectric layer stack (i. e. of the dielectric layer structure), that the silicon semiconductor wafers having the dielectric layer stack with the architecture (cf. 130.sub.4 (E) in FIG. 6a and FIG. 6b) of 330 nm of silicon dioxide, 500 nm of low-stress silicon nitride, and 1000 nm of stoichiometric silicon nitride are highly delaminated following deposition of the second stoichiometric silicon nitride and therefore were not able to be processed further. In addition, it is to be noted that the silicon semiconductor wafer having the dielectric reference layer stack with the architecture (cf. FIG. 6a and FIG. 6b 130.sub.1 (A, small depth)) of 330 nm of silicon dioxide and 1500 nm of stoichiometric silicon nitride will burst, which is due to the high intrinsic tension that is mainly found in stoichiometric silicon nitride. Thus, an optimized combination of SiO.sub.2 layer, Si.sub.3N.sub.4 layer, and Si.sub.xN.sub.y layer, as is described in the dielectric layer structure in accordance with 130.sub.2 (H, small depth), 130.sub.3 (C, small depth), 130.sub.5 (C), 130.sub.6 (C, large depth) and 130.sub.7 (H, large depth), is advantageous for the integrated capacitor with regard to a reduction of delamination and straightforward producibility of the integrated capacitor.

(78) Upon establishing the different dielectric layer stacks of the delaminated and broken silicon semiconductor wafers it becomes clear that significant distortion of the silicon semiconductor wafer is not necessarily causally related to a high intrinsic tension. This becomes clear essentially when comparing the maximum value of the largest expansion of the distortion between the broken silicon semiconductor wafer (130.sub.1 (A, small depth), comprising a distortion of, e. g., 80 μm) to the dielectric layer stack having a deep hole design with 1000 nm of stoichiometric silicon nitride and 500 nm of low-stress silicon nitride (130.sub.6 (C, large depth), comprising a distortion of, e. g., 200 μm). One possible cause is the interaction of the coefficient of thermal expansion and modulus of elasticity. Likewise, it is also conceivable that a relaxation of the crystal structure has already occurred by the time of measurement of the distortion.

(79) FIG. 7 depicts a capacitance/voltage characteristic (C(U) characteristic) of the inventive capacitor. The capacitance/voltage characteristic of an MIS capacitor exhibits a characteristic curve due to the voltage dependency of the width of the space-charge region within the substrate.

(80) For capturing the C(U) characteristic, a direct voltage 200 of −40 V to +40 V is applied (the axis of the direct voltage 200 being split up, e. g., in steps of 10 V), which is superposed by an alternating voltage having an amplitude of, e. g., 10 mV and 100 kHz. The capacitive and resistive portions of the impedance is determined from the amplitude and phase of the current flow. Upon the start at the negative voltage up to the positive voltage, the forward characteristic and the backward characteristic are captured upon reverse passing through the voltage ramp from +40 V to −40 V.

(81) In accordance with an embodiment, FIG. 7 depicts capacitances 210 within a range from 0.75*10.sup.−9 F to 1.1*10.sup.−9 F in steps of 0.05*10.sup.−9 F for the hole design L10, and within a range from 1.55*10.sup.−9 F to 1.9*10.sup.−9 F in steps of 0.05*10.sup.−9 F for the hole design L20. FIG. 7 depicts only the C(U) characteristic of specific embodiments, and it is clear that with alternative dielectric layer structures, C(U) characteristics may also be implemented within other ranges. It is to be noted that the capacitor described herein is not limited to the parameters depicted in FIG. 7 and FIG. 8b.

(82) The capacitance/voltage measurements were performed, by way of example, on several non-charged capacitors (see FIG. 8b for parameters with regard to the capacitors analyzed) on several silicon semiconductor wafers. FIG. 7 shows the C(U) characteristics of three components from the center of, respectively, a silicon semiconductor wafer having the hole design L10 and of a component from the center of a silicon semiconductor wafer having the hole design L20.

(83) With the negative voltage of, e. g., −40 V, the p-MIS capacitors are operated in accumulation, in accordance with an embodiment. In an accumulation mode of the capacitor, the characteristic approximates a capacitance value (see FIG. 7). However, said approximation does not take place with all of the capacitors examined. With the components, e. g. 130.sub.2 (H, small depth), of the silicon semiconductor wafers having the dielectric layer combinations comprising a stoichiometric silicon nitride layer as the second or third silicon nitride deposition, the value of the capacitance increases, e. g., as the voltage decreases, so that the capacitance value cannot be directly determined, in the accumulation mode of the capacitor, by means of this measurement setup.

(84) If one compares the results between the hole designs L10 and L20, it may be found that the components, e. g. 130.sub.5 (C), located on the silicon semiconductor wafers having the hole design L20 exhibit higher capacitances than the components, e. g. 130.sub.2 (H, small depth), 130.sub.3 (C, small depth), and 130.sub.8 (I, small depth), located on the silicon semiconductor wafers having the hole design L10. The reason is the larger surface area that is due to the deeper holes of the hole design L20 as compared to the surface area of the hole design L10.

(85) An overall permittivity of a dielectric stack is higher when it has a higher proportion of low-stress silicon nitride. As a result, the capacitance also increases. A larger thickness of the dielectric stack entails a lower capacitance value. Accordingly, the capacitance value of the silicon semiconductor wafer 130.sub.2 (H, small depth) (see FIG. 6b and FIG. 7) is increased because of a pronounced discrepancy between a smaller thickness and a higher overall permittivity.

(86) For determining the dielectric strength and for identifying different charge transport mechanisms within the capacitor, capturing a current/voltage characteristic of the inventive capacitors is suitable. The dielectric strength of a capacitor is, e. g., dependent on the thickness of the dielectric (i. e. of the dielectric layer structure) and on the electric field strength. An irreversible breakdown of the dielectric takes place when a critical field strength E.sub.crit is exceeded at a breakdown voltage U.sub.BD present. The maximally achieved voltage U.sub.IV,max ist determined at a current flow of 10 mA. With none of the measured layer stacks, a dielectric breakdown can be found up to the maximally achieved voltage. The breakdown voltage is therefore higher than the maximally achieved voltage, so that the producibility of capacitors having dielectric strengths of 1200 V can be successfully confirmed.

(87) The current/voltage characteristic of a component from the center of each silicon semiconductor wafer is depicted in FIG. 8a. In other words, FIG. 8a depicts measured output characteristics of Si capacitors with a dielectric composed of silicon dioxide and silicon-rich nitride (130.sub.8 (I, small depth)) and a combination of silicon-rich and stoichiometric nitride (130.sub.2 (H, small depth), 130.sub.3 (C, small depth), 130.sub.5 (C)) of equal or similar overall thicknesses. In accordance with an embodiment, FIG. 8a shows the current measured at the component as a function of the voltage applied. Within this context, the current is depicted in logarithmic terms, e. g. within a range from 10.sup.−9 A to 10.sup.−2 A, and the voltage is depicted, e. g., within a range from 0 V to 1600 V, in steps of 200 V. FIG. 8a depicts the current as a function of the voltage of specific embodiments only, and it is clear that with alternative dielectric layer structures, currents within different ranges (e. g. within different current and/or voltage ranges) may be implemented. It is to be noted that the capacitor described herein is not limited to the parameters depicted in FIG. 8a and FIG. 8b. An advantageous embodiment of the present invention is represented by the capacitor having the dielectric layer structure in accordance with 130.sub.3 (C, small depth).

(88) FIG. 8b shows the different layer stacks that are analyzed in FIG. 8a, with the maximally achieved voltage and the overall layer thickness of the dielectric.

(89) FIG. 8a clearly shows the higher voltages achieved with the dielectric multi-layer systems comprising two stoichiometric silicon nitride layers (130.sub.5 (C), 130.sub.3 (C, small depth)) as compared to the dielectric layer stack consisting exclusively of low-stress silicon nitride (130.sub.8 (I, small depth)) and as compared to the dielectric layer variations comprising a stoichiometric silicon nitride layer (130.sub.2 (H, small depth)). Thus, the integrated capacitor described herein is optimized as compared to general capacitors having a layer architecture of 130.sub.8. A comparison of the current/voltage characteristic of the components having a stoichiometric silicon nitride layer (130.sub.2 (H, small depth)) reveals that the contact between stoichiometric silicon nitride and the underlying silicon dioxide is advantageous in terms of a higher maximally achieved voltage. The contact between low-stress silicon nitride and the front-side electrode, which might have a negative effect on the current/voltage characteristic on account of the higher defect density present in low-stress silicon nitride as compared to stoichiometric silicon nitride (130.sub.2 (H, small depth), 130.sub.3 (C, small depth)), is not confirmed by measurements.

(90) The influence of the enlarged surface area of the components that is due to deeper holes becomes clear, with regard to the current/voltage characteristic, in that the maximally achieved voltage of the components having a hole design L20 (130.sub.5 (C)) is lower as compared to those components of the same dielectric layer stack which have the hole design L10 (130.sub.2 (H, small depth), 130.sub.3 (C, small depth)). Possible reasons are the additional vacancies that are present at the interface with the electrode because of the enlarged surface area, on the one hand. On the other hand, the overall thickness of the dielectric is smaller since, in accordance with an embodiment, the deposition rate of the dielectric layers during manufacturing is not adapted despite the larger surface area.

(91) In accordance with an embodiment, a silicon capacitor (such as, e. g., 130.sub.5 (C)), the trench structure of which comprises a hole depth of about 20 μm and which comprises a dielectric layer stack of 330 nm of silicon dioxide, 500 nm of stoichiometric silicon nitride, 500 nm of low-stress silicon nitride, and 500 nm of stoichiometric silicon nitride, offers an optimized concept in terms of capacitance and dielectric strength. Thus, one achieves a capacitance, per area unit, of 133 pF/mm.sup.2 and a dielectric strength of 1450 V.

(92) FIG. 9 schematically shows individual consecutive process steps of a method 300 of manufacturing the inventive silicon capacitors in accordance with an embodiment of the present invention. Within this context, particular attention is to be paid to the fact that FIG. 9 schematically depicts exclusively processing of the relevant front side of the silicon semiconductor wafer and that depositions performed on the rear side and their back-etching processes are not included.

(93) At the beginning of the process, in order to realize a trench capacitor 100 in accordance with an embodiment, silicon substrate 110 representing a first electrode structure is structured with holes 111.sub.1 to 111.sub.3. By means of a lithography 310, the hole structure is transferred to the surface and, on the basis thereof, is etched 320 into the depth of the substrate. This is followed by the different processes for depositing 330 the individual layers 132 and 133.sub.1 to 133.sub.3 of the dielectric 130, i. e. of the dielectric layer structure, and for depositing 340 the electrode 120, i. e. the second electrode structure. Moreover, the method optionally includes producing 350 a front-side contact 122 and a rear-side contact 112. The electrode 120, consisting of, e. g., polycrystalline silicon, and the front-side contact 122 made of, e. g., aluminum, define dimensions of the component. In order to avoid air sparkover between the front 122 and rear-side contacts 112 of the singulated components, polyimide is optionally applied on the edges of the components.

(94) In accordance with an embodiment, the inventive capacitor 100 is built on a silicon semiconductor wafer (i. e. silicon substrate 110) p-doped with boron, which wafer comprises, e. g., a diameter of 150 mm, a thickness of 675 μm, and a layer resistivity of 9±0.3 ohm*cm. By applying a lithography 310, e. g., a hexagonal hole structure as described in FIG. 4b and FIG. 5a is produced on the surface of the silicon 110.

(95) Etching 320 of the holes 111.sub.1 to 111.sub.3 is effected, for example, by means of an ASE (advanced silicon etching) process, which is an alternating dry-etching process. On the basis of the lithography 310 for producing a hole mask 312, one alternates, in etching 320 the holes 111.sub.1 to 111.sub.3, between an ion etching step and a passivating step, in accordance with an embodiment. For passivation, a protective layer is deposited, for example, in between the individual, non-fully anisotropic etching steps, which protective layer serves to maintain the etching direction and to protect the walls, which have already been etched, of the hole 111.sub.1 to 111.sub.3 against further material removal.

(96) In accordance with an embodiment, the dielectric 130 consists of a silicon dioxide layer 132 and three subsequent silicon nitride layers 133.sub.1 to 133.sub.3 of different properties. They are referred to as silicon nitride1 133.sub.1, silicon nitride2 133.sub.2, and silicon nitride3 133.sub.3, in accordance with their order of deposition. Initially, the SiO.sub.2 layer 132 is deposited 330 on the silicon substrate at 1050° C., for example, by means of thermal oxidation based on a natural reaction of silicon (Si) with oxygen (O.sub.2) to yield silicon dioxide (SiO.sub.2).

(97) A silicon nitride layer 133.sub.1 to 133.sub.3 is produced, e. g., by means of a chemical reaction of dichlorosilane (SiCl.sub.2H.sub.2) with ammonia (NH.sub.3). Depositions 330 of the different silicon nitride layers 133.sub.1 to 133.sub.3 take place, e. g., by means of a low-pressure chemical vapor deposition (LPCVD) at a low pressure (about 17 Pa-27 Pa) as the atmospheric pressure and at a temperature of between 700° C. and 800° C.

(98) The chemical reaction equation for depositing 330 stoichiometric silicon nitride (Si.sub.3N.sub.4) is 3 SiCl.sub.2H.sub.2+4NH.sub.3.fwdarw.Si.sub.3Na+6 HCl+6H.sub.2. As side products of the reaction, hydrogen chloride (HCl) and hydrogen (H.sub.2) are formed.

(99) During deposition 330 of low-stress silicon nitride (Si.sub.xN.sub.y), the ratio of the reaction gases dichlorosilane and ammonia is changed. The higher the proportion of dichlorosilane as compared to ammonia, the higher the proportion of silicon will be in the resulting silicon nitride. As a result, what is deposited 330 is not silicon nitride having a ratio of three to four between silicon and nitride as in the stoichiometric case, but a silicon-rich silicon nitride and/or non-stoichiometric silicon nitride, which is also referred to as a low-stress silicon nitride Si.sub.xN.sub.y.

(100) In the production 330 of the dielectric 130, therefore, in accordance with an embodiment, an SiO.sub.2 layer is initially deposited onto the silicon substrate, followed by combinations of layers 133.sub.1 to 133.sub.3 including at least one Si.sub.3Na layer and at least one Si.sub.xN.sub.y layer.

(101) As the front-side electrode 120, i. e. as the second electrode structure, e. g. polycrystalline silicon is grown 340 by means of an LPCVD method by using monosilane (SiH.sub.4) SiH.sub.4.fwdarw.Si+2H.sub.2. In this process, the polysilicon is deposited on the dielectric layer stack 130 in the holes 111.sub.1 to 111.sub.3 by means of, e. g., pyrolytic decomposition of silane at 600° C. to 650° C. As said deposition in the hole structure continues, the electrode 120 that may be used is formed by the time an increased capacitance value is achieved. In accordance with an embodiment, doping of the polysilicon with phosphorus and argon is effected, during deposition, by an additional gas supply of monophosphane (PH.sub.3) and argon (Ar).

(102) In accordance with an embodiment, a further layer is deposited onto the polysilicon layer, i. e. onto the second electrode structure 120, for forming the front-side contact 122 with, e. g., aluminum. Said further layer is deposited, e. g., by means of a physical vapor deposition (PVD) method. Aluminum is low in resistance and forms reliable contacting of the individual components. Finally, the dimension of the component is optionally transferred to the silicon semiconductor wafer by means of a lithography, and the layers of polysilicon and aluminum that are present between the resulting components are ablated by using a dry-etching process.

(103) For contacting the rear side of the silicon semiconductor wafer 110, in accordance with an embodiment, all of the layers that are additionally grown, or deposited, onto the rear side during the desired layer deposition on the front side, are initially etched back. For contacting the rear side, e. g. a solderable layer stack 112 of chromium, nickel and silver is subsequently vapor deposited.

(104) In accordance with an embodiment, an integrated capacitor comprises a first electrode structure, a second electrode structure, and an interposed dielectric layer structure, the dielectric layer structure comprising a layer combination having an SiO.sub.2 layer, an Si.sub.3Na layer, and an Si.sub.xN.sub.y layer, and the Si.sub.xN.sub.y layer comprising non-stoichiometric silicon nitride material with an increased proportion of silicon.

(105) In accordance with one aspect, the first electrode structure comprises a semiconductor substrate provided with a trench structure.

(106) In accordance with a further aspect, the first electrode structure forms a rear-side electrode arranged at a rear-side contact; and the second electrode structure forms a front-side electrode arranged at a front-side contact.

(107) In accordance with a further aspect, a ratio of silicon to nitrogen of the Si.sub.xN.sub.y layer is from 0.8 to 1.

(108) In accordance with a further aspect, the dielectric layer structure comprises a larger proportion of an Si.sub.3N.sub.4 material than of any material deviating therefrom.

(109) In accordance with a further aspect, a thickness of the Si.sub.3N.sub.4 layer is formed to have a ratio of n to one to a thickness of the Si.sub.xN.sub.y layer, n ranging from 1.5 to 2.5.

(110) In accordance with a further aspect, the Si.sub.xN.sub.y layer is arranged separately from, or not directly adjacent to, the SiO.sub.2 layer.

(111) In accordance with a further aspect, the thickness of the Si.sub.xN.sub.y layer corresponds to a maximum of 33% of the overall thickness of all Si.sub.xN.sub.y layers and Si.sub.3N.sub.4 layers.

(112) In accordance with a further aspect, a thickness of the Si.sub.xN.sub.y layer corresponds to a maximum of 50% of an overall thickness of all Si.sub.3N.sub.4 layers.

(113) In accordance with a further aspect, the dielectric layer structure of the integrated capacitor comprises an effective oxide thickness of at least 1200 nm and a dielectric strength of at least 900 V.

(114) In accordance with an embodiment, a method of producing an integrated capacitor comprises producing a dielectric layer structure within a trench structure of a semiconductor substrate, the dielectric layer structure comprising a plurality or a combination of mutually adjoining dielectric layers, at least one of the dielectric layers comprising an SiO.sub.2 material, at least one of the dielectric layers comprising an Si.sub.3N.sub.4 material, and at least one of the dielectric layers comprising an Si.sub.xN.sub.y material, the Si.sub.xN.sub.y material comprising non-stoichiometric silicon nitride having an increased proportion of silicon.

(115) In accordance with one aspect, the dielectric layer structure comprises a plurality or a combination of an SiO.sub.2 layer, an Si.sub.3N.sub.4 layer, and an Si.sub.xN.sub.y layer.

(116) In accordance with a further aspect, a ratio of silicon to nitrogen of the Si.sub.xN.sub.y layer ranges from 0.8 to 1.

(117) In accordance with a further aspect, the dielectric layer structure comprises a higher proportion of the Si.sub.3N.sub.4 material than of any material deviating therefrom.

(118) In accordance with a further aspect, a thickness of the Si.sub.3N.sub.4 layer is formed to have a ratio of n to one to a thickness of the Si.sub.xN.sub.y layer, n ranging from 1.5 to 2.5.

(119) In accordance with a further aspect, the Si.sub.xN.sub.y layer is not arranged directly adjacent to the SiO.sub.2 layer.

(120) One embodiment relates to a method of any of the above-described embodiments.

(121) A further embodiment relates to a device of any of the above-described embodiments.

(122) A further embodiment relates to a production method of any of the above-described embodiments.

(123) Even though some aspects have been described within the context of a device, it is understood that said aspects also represent a description of the corresponding method, so that a block or a structural component of a device is also to be understood as a corresponding method step or as a feature of a method step. By analogy therewith, aspects that have been described in connection with or as a method step also represent a description of a corresponding block or detail or feature of a corresponding device.

(124) While this invention has been described in terms of several embodiments, there are alterations, permutations, and equivalents which fall within the scope of this invention. It should also be noted that there are many alternative ways of implementing the methods and compositions of the present invention. It is therefore intended that the following appended claims be interpreted as including all such alterations, permutations and equivalents as fall within the true spirit and scope of the present invention.

REFERENCES

(125) [1] S. Banzhaf et al, “Stress reduction in high voltage MIS capacitor fabrication”, 19th International Symposium POWER ELECTRONICS Ee2017, Novi Sad, Serbia, 2017. [2] S. Berberich, M. März, U.S. Pat. No. 7,738,226. [3] T. Erlbacher, H. Schwarzmann, A. J. Bauer, “Improving module performance and reliability in power electronic applications by monolithic integration of RC-snubbers”, Proc. Internat. Symp. On Power Semicond. Dev. & ICs (ISPSD), Brugge, 283-286. [4] J. vom Dorp, T. Erlbacher, V. Lorentz, A. J. Bauer, H. Ryssel, L. Frey, “Integrierbare Bauelemente zur Erhöhung der Betriebssicherheit elektronischer Systemkomponenten im Automobil”, Automotive meets Electronics 2010, Dortmund pp. 72-77. [5] F. Krach, T. Heckel, L. Frey, A. J. Bauer, T. Erlbacher and M. Maerz, “Innovative monolithic RC-snubber for fast switching power modules”, 9th International Conference on Integrated Power Electronics Systems CIPS2016, Nuremberg, Germany, 2016, pp. 1-6. [6] S. Banzhaf et al., “Post-trench processing of silicon deep trench capacitors for power electronic applications”, 28th International Symposium on Powersemiconductor Devices and ICs ISPSD 2016, Prague, Czech Republic, 2016. [7] Hakim et al., “Nonlinear capacitors integration”, Proc. Internat. Semiconductor Conference 2000, Sinaia, pp. 303-306 vol:1,2000. [8] J. vom Dorp, T. Erlbacher, A. J. Bauer, H. Ryssel, L. Frey, “Dielectric layers suitable for high voltage integrated trench capacitors”, J. Vac. Sci. Technol. B 29(1), 01AB04, 2011. [9] F. Krach, H. Schwarzmann, A. J. Bauer, T. Erlbacher, and L. Frey, “Silicon nitride, a high potential dielectric for 600 V integrated RC-snubber applications”, Journal of Vacuum Science and Technology B, 33, pp. 01A1112, 2015. [10] J. Gardeniers, H. Tilmans and C. Visser, “LPCVD silicon-rich silicon nitride films for applications in micromechanics, studied with statistical experimental design”, J. Vac. Sci. Technol. A 14(5), September/October 1996. [11] V. Prodanovic et al., “Optimization of silicon-rich silicon nitride films for electron multiplication in timed photon counters”, EUROSENSORS 2015.

Integrated capacitor and method of producing an integrated capacitor

Assignee

Inventors

Cpc classification

Classification Explorer

H01G4/40

ELECTRICITY

Classification Explorer

H01L29/945

ELECTRICITY

Classification Explorer

H01L28/40

ELECTRICITY

Classification Explorer

H01L29/66181

ELECTRICITY

International classification

Classification Explorer

H01L29/94

ELECTRICITY

Classification Explorer

H01L29/66

ELECTRICITY

Abstract

Claims

Description