Semiconductor device and manufacturing method thereof

This application is based upon and claims the benefit of priority from Japanese Patent Application No. 2015-174923, filed Sep. 4, 2015, the entire contents of which are incorporated herein by reference.

Embodiments described herein relate generally to a semiconductor device and a manufacturing method thereof.

Graphene sheets are a novel carbon material exhibiting quantized conduction (ballistic conduction) in the same manner as carbon nanotubes, and are attracting attention as a next-generation technology for fabricating low-resistance interconnects to replace metal interconnects. Because the mean free path of electrons in a graphene sheet is very long, that is, about 100 nm to 1 μm, graphene sheets are highly advantageous in electrical conduction in a long interconnect. A graphene sheet is formed on an exposed face of a catalyst layer by CVD or the like. Because an edge portion of a graphene sheet has the lowest electrical resistance in the graphene sheet, a method for forming a plurality of edges in a graphene sheet is required.

FIG. 1 is a perspective view illustrating a schematic structure of a semiconductor device according to a first embodiment;

FIG. 2 is a cross-sectional view taken along line II-II of FIG. 1 ;

FIG. 3 is a diagram illustrating the relationship between the linewidth of a graphene sheet and electrical resistance;

FIGS. 4, 5, 6, and 7 are cross-sectional views illustrating the schematic structures of the semiconductor device according to the first embodiment;

FIGS. 8, 9, 10, 11, 12, 13, 14, and 15 are cross-sectional views illustrating a manufacturing process of the semiconductor device according to the first embodiment;

FIG. 16 is a perspective view illustrating a schematic structure of a semiconductor device according to a second embodiment;

FIG. 17 is a cross-sectional view taken along line XVII-XVII of FIG. 16 ; and

FIGS. 18, 19, 20, and 21 are cross-sectional views illustrating the schematic structures of the semiconductor device according to the second embodiment.

In general, according to one embodiment, a semiconductor device includes an underlayer formed on a substrate, a catalyst layer disposed on the underlayer and extending in an interconnect length direction. The device further includes an upper graphene layer formed on an upper face of the catalyst layer; and side graphene layers provided on two respective side faces of the catalyst layer, the two side faces extending in the interconnect length direction.

Embodiments will be explained hereinafter with reference to drawings.

First Embodiment

FIG. 1 is a perspective view illustrating a schematic structure of a semiconductor device according to a first embodiment. FIG. 2 is a cross-sectional view taken along line II-II of FIG. 1 .

As illustrated in the drawings, the semiconductor device according to the first embodiment includes a semiconductor substrate 10 on which semiconductor elements such as transistors and capacitors are formed, an underlayer 11 formed on the semiconductor substrate 10 , a first contact via 12 that is embedded and formed in the underlayer 11 , a liner layer 13 that is formed on the underlayer 11 , a catalyst layer 14 that is formed on the liner layer 13 , a mask layer 15 that is formed on an upper face of the catalyst layer 14 , an upper graphene layer 16 u that is formed on an upper face of the mask layer 15 , and side graphene layers 16 s formed on the both respective side faces of the catalyst layer 14 . The side graphene layers 16 s are formed on the two respective side faces of the catalyst layer 14 that extend in an L-direction. Specifically, one catalyst layer 14 is provided with two side graphene layers 16 s.

The catalyst layer 14 has a bottom face located on the side of the underlayer 11 and the liner layer 13 , an upper face opposed to the bottom face, and two side faces extending along the interconnect length direction L (second direction orthogonal to a first direction). The catalyst layer 14 has a height in the interconnect height direction W 2 (third direction orthogonal to the first and the second directions). The height of the catalyst layer 14 has an influence on the height of the graphene layer 16 . The catalyst layer 14 has the upper face on which the upper graphene layer 16 u is formed, and the two side faces on which the side graphene layers 16 s are formed. None of the three faces contact the underlayer 11 or the liner layer 13 . Only the bottom face of the catalyst layer 14 contacts the underlayer 11 and the liner layer 13 .

In the drawings, the catalyst layer 14 is a pillar, and the length of the upper face of the catalyst layer 14 is substantially equal to the length of the bottom face of the catalyst layer 14 in the interconnect width direction W 1 (first direction). For this reason, the two side faces extending in the L-direction of the catalyst layer 14 are parallel to each other (W 2 direction). The catalyst layer 14 is not limited to such a configuration, but may be a rectangle, a circle, or a polygon, for example, or may take the form of a hollow cylindrical structure or a layered structure.

The material of the catalyst layer 14 is preferably simple metal such as cobalt (Co), nickel (Ni), iron (Fe), ruthenium (Ru), and copper (Cu), or an alloy or a carbide containing one or more of these elements. Otherwise, the catalyst layer 14 may have a layered structure in which layers containing these materials are stacked. The thickness of the catalyst layer 14 is preferably 0.5 nm or more, to prevent discontinuous formation of the graphene layer 16 .

The graphene layer 16 includes one or more upper graphene layer 16 u , and two or more side graphene layers 16 s . The upper graphene layer 16 u and the side graphene layers 16 s are layers that are cut out of one graphene layer 16 , and include the same material. In addition, each of the upper graphene layer 16 u and the side graphene layers 16 s has a layered structure in which single-layer graphene sheets are stacked, and each of them functions as an interconnect.

Among the side faces of the upper graphene layer 16 u , each of two side faces extending along the L-direction includes an edge face (catalyst-noncontacting edge face) 16 unc of the graphene layer 16 that does not contact the catalyst layer 14 . Neither of the two catalyst-noncontacting edge faces 16 unc of the upper graphene layer 16 u contacts the catalyst layer 14 . Each non-catalyst-contacting edge face 16 unc includes edges of the single-layer graphene sheets included in the upper graphene layer 16 u.

An upper face of each of the side graphene layers 16 s that does not contact the catalyst layer 14 includes an edge face (catalyst-noncontacting edge face) 16 snc of the graphene layer 16 that does not contact the catalyst layer 14 . In addition, a side face of each of the side graphene layers 16 s that contacts the catalyst layer 14 includes an edge face (catalyst-contacting edge face) 16 sc of the graphene layer 16 that contacts the catalyst layer 14 .

Specifically, each of the side graphene layers 16 s is provided with a catalyst-contacting edge face 16 sc and a catalyst-noncontacting edge face 16 snc . Each of the catalyst-noncontacting edge face 16 snc and the catalyst-contacting edge face 16 sc includes edges of respective single-layer graphene sheets included in the side graphene layer 16 s.

In each of the side graphene layers 16 s , each of the single-layer graphene sheets is formed to have an L-shape from the catalyst-contacting edge face 16 sc side to the catalyst-noncontacting edge face 16 snc side. Specifically, the side including the catalyst-contacting edge face 16 sc and extending in the W 1 direction is connected with the side including the catalyst-noncontacting edge face 16 snc and extending in the W 2 direction to form an L-shape.

To increase the connection area with a contact via 20 described later, the catalyst-noncontacting edge face 16 unc of the upper graphene layer 16 u preferably has a small distance from the catalyst-noncontacting edge face 16 snc of each side graphene layer 16 s . In addition, the height (W 2 direction) of the catalyst-noncontacting edge face 16 unc of the upper graphene layer 16 u is preferably equal to the width (W 1 direction) of the catalyst-noncontacting edge face 16 snc of each side graphene layer 16 s . It is also preferable that the length (L-direction) of the catalyst-noncontacting edge face 16 unc of the upper graphene layer 16 u is preferably equal to the length (L-direction) of the catalyst-noncontacting edge face 16 snc of each side graphene layer 16 s.

Generally, because a localized electron state exists at edges of a graphene sheet, quantized conduction (ballistic conduction) more easily occurs at edges of a graphene sheet than in other portions of a graphene sheet, and edges of a graphene sheet have a very low electrical resistance. For this reason, the electrical resistance becomes lower as the number of edges of a graphene sheet increases. Because the catalyst-noncontacting edge faces 16 unc and 16 snc of the upper graphene layer 16 u and the side graphene layers 16 s include many edges of graphene sheets, the catalyst-noncontacting edge faces 16 unc and 16 snc have a very low electrical resistance and serve as low-resistance conduction paths.

In the structure illustrated in the drawings, at least four catalyst-noncontacting edge faces 16 unc and 16 snc are obtained from one graphene layer 16 formed on one catalyst layer 14 .

The graphene layer 16 including a plurality of edges of graphene sheets as described above has a lower electrical resistance than does a graphene sheet that does not include a plurality of edges of graphene sheet. This structure reduces the height (W 2 direction) of the graphene layer 16 .

The liner layer 13 is an auxiliary layer to improve the crystallinity of the catalyst layer 14 . The liner layer 13 increases adhesion between the catalyst layer 14 and the graphene layer 16 , and enables the graphene layer 16 to uniformly grow on the catalyst layer 14 . The liner layer 13 also prevents metal included in the catalyst layer 14 from diffusing into the underlayer 11 and the contact via 12 . The material of the liner layer 13 is, for example, tantalum (Ta), titanium (Ti), ruthenium (Ru), tungsten (W), aluminum (Al), or a nitride (such as tantalum nitride (TaN) or a oxide containing these elements. The liner layer 13 may have a structure in which layers containing these materials are stacked.

An edge seal layer 17 is provided on the catalyst-noncontacting edge faces 16 unc and 16 snc of the upper graphene layer 16 u and the side graphene layers 16 s . The edge seal layer 17 prevents intercalants from separating from the catalyst-noncontacting edge faces 16 unc and 16 snc of the upper graphene layer 16 u and the side graphene layers 16 s . The edge seal layer 17 is, for example, a silicon oxide film, or a silicon nitride film.

The uppermost face 16 uu of the upper graphene layer 16 u is covered with a face seal layer 18 . The face seal layer 18 prevents the intercalants from separating from the uppermost face 16 uu of the upper graphene layer 16 u . The face seal layer 18 is, for example, a silicon oxide film, or a silicon nitride film.

In addition, as indicated in a broken line in FIG. 1 , a modifying layer 19 may be provided on the catalyst-noncontacting edge faces 16 unc and 16 snc of the upper graphene layer 16 u and the side graphene layers 16 s . The modifying layer 19 modifies (edge face processing) the catalyst-noncontacting edge faces 16 unc and 16 snc , to dope intercalants into the upper graphene layer 16 u and the side graphene layers 16 s . Because the modifying layer 19 modifies edges of the graphene sheets included in the catalyst-noncontacting edge faces 16 unc and 16 snc , the modifying layer 19 is not included in the uppermost face 16 uu of the upper graphene layer 16 u . The modifying layer 19 is not limited to a layer (continuous film), but may be molecules or a monoatomic layer. The modifying group is, for example, oxygen, nitrogen, or boron. The intercalants are impurities doped into the upper graphene layer 16 u and the side graphene layers 16 s by intercalation or the like. The intercalants are, for example, halogen, a metal halide, or alkali metal.

Doping intercalants into the graphene layer 16 by intercalation or using modifying groups enables further reduction of the electrical resistance of the graphene layer 16 .

The liner layer 13 , the mask layer 15 , the edge seal layer 17 , the face seal layer 18 , and the modifying layer 19 may be omitted.

The semiconductor substrate 10 is, for example, a silicon semiconductor substrate. The underlayer 11 is an interlayer insulating layer mainly including silicon oxide, silicon nitride, and an airgap. The contact via 12 is formed of, for example, copper, aluminum, tungsten, or an alloy containing one of more of these elements.

FIG. 3 is a diagram illustrating the relationship between the electrical resistance (γΩ/μm) and the respective widths (Wnm) of the graphene sheet interconnect and the metal interconnect.

As illustrated in FIG. 3 , when the width of the graphene sheet interconnect is less than or equal to a predetermined width (for example, 10 nm), the graphene sheet interconnect exhibits a lower electrical resistance than does the metal interconnect (for example, a copper [Cu] interconnect). By contrast, when the graphene sheet interconnect has a width greater than the predetermined width, the graphene sheet interconnect has a higher electrical resistance than does the metal interconnect.

For this reason, the graphene sheet interconnect having a width greater than the predetermined linewidth is preferably cut so as to have a linewidth that is less than or equal to the predetermined linewidth using mask processing or the like. In a thick graphene sheet interconnect, graphene sheet edge faces that do not contact the catalyst layer 14 increases by two faces for each cut.

In the drawings, the graphene layer 16 having a width greater than the predetermined linewidth has four catalyst-noncontacting edge faces 16 unc of the upper graphene layer 16 u . Specifically, the graphene layer 16 having a width greater than the predetermined linewidth has six catalyst-noncontacting edge faces 16 unc and 16 snc of the upper graphene layer 16 u and the side graphene layers 16 s.

In the case of the present example, a plurality of edge faces can be cut out of the graphene layer 16 having a width greater than the predetermined linewidth. For this reason, it is unnecessary to provide a new layer to divide the thick graphene layer 16 . The thick graphene layer 16 having a plurality of edge faces has a lower electrical resistance than does a thick graphene layer that does not have a plurality of edge faces.

In any case, the current in the graphene layer 16 flows in the L-direction.

The following is explanation of a method for connecting the graphene layer 16 with the contact via 20 in the case where the mask layer 15 is conductive or no mask layer 15 is provided. Because the catalyst layer 14 , the mask layer 15 , and the graphene layer 16 are conductive, the contact via 20 can be electrically connected with all the portions of the graphene layer 16 .

In FIG. 4 , each contact via 20 is connected to one face on the uppermost face 16 uu of the upper graphene layer 16 u (first connection method). The first connection method is advantageous in miniaturization, because the connection area between the contact via 20 and the graphene layer 16 is small.

In the case of FIG. 4 , each contact via 20 is not in contact with the catalyst-noncontacting edge faces 16 unc and 16 snc of the graphene layer 16 . However, as described above, because the graphene layer 16 includes a plurality of catalyst-noncontacting edge faces 16 unc and 16 snc , the graphene layer 16 has a lower electrical resistance than does a graphene layer having no catalyst-noncontacting edge faces 16 unc and 16 snc.

In FIG. 5 , each contact via 20 is connected to two faces, that is, one of the two catalyst-noncontacting edge faces 16 unc included in the upper graphene layer 16 u , and the catalyst-noncontacting edge face 16 snc of one side graphene layer 16 s (second connection method).

In FIG. 6 , each contact via 20 is connected to three faces, that is, the uppermost face 16 uu of the upper graphene layer 16 u , the catalyst-noncontacting edge face 16 unc of the upper graphene layer 16 u , and the catalyst-noncontacting edge face 16 snc of one side graphene layer 16 s (third connection method).

In FIG. 7 , each contact via 20 is connected to five faces, that is, the uppermost face 16 uu of the upper graphene layer 16 u , two catalyst-noncontacting edge faces 16 unc of the upper graphene layer 16 u , and the catalyst-noncontacting edge faces 16 snc of the two side graphene layers 16 s . Specifically, the contact via 20 covers the uppermost face 16 uu of the upper graphene layer 16 u and all the catalyst-noncontacting edge faces 16 unc and 16 snc included in the graphene layer 16 (fourth connection method).

As described above, edge portions of a graphene sheet have the lowest resistance in the graphene sheet. For this reason, the catalyst-noncontacting edge faces 16 unc and 16 snc have a lower electrical resistance than does the uppermost face 16 uu of the upper graphene layer 16 u . Specifically, the electrical resistance decreases as the number of graphene sheets that contact the contact via 20 increases and the number of edges of the graphene sheets that contact the contact via 20 increases.

In the first connection illustrated in FIG. 4 , the contact via 20 contacts a face of the uppermost graphene sheet included in the uppermost face 16 uu of the upper graphene layer 16 u . By contrast, in the second to fourth connections illustrated in FIG. 5 to FIG. 7 , the contact via 20 at least contacts edges of a plurality of graphene sheets included in the catalyst-noncontacting edge faces 16 unc and 16 snc . The second to fourth connection methods achieve lower resistance than that in the first connection method, because each contact via 20 is electrically connected with a plurality of catalyst-noncontacting edge faces 16 unc and 16 snc . The second to fourth connections achieve a lower electrical resistance than does the first connection, because the second to fourth connections have a greater number of graphene sheets that contact the contact via 20 than does the first connection, and each contact via 20 electrically contacts edges of the respective graphene sheets.

In the third connection illustrated in FIG. 6 , each contact via 20 is connected with the catalyst-noncontacting edge face 16 unc of the upper graphene layer 16 u and the catalyst-noncontacting edge face 16 snc of the side graphene layer 16 s , in the same manner as the second connection illustrated in FIG. 5 . In addition, in the third connection illustrated in FIG. 6 , because the contact via 20 is also connected with the uppermost face 16 uu of the upper graphene layer 16 u , the third connection has a lower electrical resistance than does the second connection.

In the fourth connection illustrated in FIG. 7 , the number of edges of graphene sheets that electrically contact the contact via 20 is greater than that in the third connection illustrated in FIG. 6 . For this reason, the fourth connection has a lower electrical resistance than does the third connection.

Because the electrical resistance decreases as the electrical connection area between the contact via 20 and the graphene layer 16 increases, the electrical resistance decreases in the order of the first connection method, the second connection method, the third connection method, and the fourth connection method. In addition, because the third and fourth connection methods have larger connection areas between the contact via 20 and the graphene layer 16 than do the first and second connection methods, the third and fourth connection methods enable reduction of the alignment margin failure caused when the contact via 20 is processed. Although the drawings illustrate the state where the face seal layer 17 , the edge seal layer 18 , and the modifying layer 19 existing between the contact via 20 and the graphene layer 16 are removed, the face seal layer 17 , the edge seal layer 18 , and the modifying layer 19 are not necessarily removed when they are conductive. Specifically, when the face seal layer 17 , the edge seal layer 18 , and the modifying layer 19 are conductive, the face seal layer 17 , the edge seal layer 18 , and the modifying layer 19 may be provided between the contact via 20 and the graphene layer 16 .

When the mask layer 15 is insulating, the contact via 20 is required to be connected to both the upper graphene layer 16 u and the side graphene layer 16 s . In such a case, the graphene layer 16 and the contact via 20 are connected using one of the second to fourth connection methods.

The following is explanation of a method for manufacturing the semiconductor device according to the first embodiment, with reference to FIG. 8 to FIG. 15 .

As illustrated in FIG. 8 , an underlayer 11 is formed on a semiconductor substrate 10 , a first contact via 12 is embedded and formed in the underlayer 11 , a liner layer 13 is formed on the underlayer 11 , and a catalyst layer 14 is formed on the liner layer 13 .

Next, as illustrated in FIG. 9 , a pattern of a mask layer 15 is formed on the catalyst layer 14 .

Thereafter, as illustrated in FIG. 10 , the liner layer 13 and the catalyst layer 14 are patterned in accordance with the pattern of the mask layer 15 by a processing method such as RIE. In the patterning, the mask layer 15 may be left on the catalyst layer 14 , or entirely removed from the catalyst layer 14 .

Next, as illustrated in FIG. 11 , a graphene layer 16 is formed on the catalyst layer 14 . The graphene layer 16 is grown on base points of projections and depressions formed on the face of the catalyst layer 14 , in particular, projections and depressions formed on the side faces of the catalyst layer 14 .

The innermost single-layer graphene sheet that contacts the catalyst layer 14 is formed along one of side faces extending in the L-direction of the catalyst layer 14 (W 2 direction), runs on the upper face of the catalyst layer 14 and the mask layer 15 (W 1 direction), and thereafter reaches the other side face of the catalyst layer 14 (W 2 direction). Accordingly, the innermost single-layer graphene sheet is grown to roll the catalyst layer 14 and the mask layer 15 , with the catalyst layer 14 and the mask layer 15 serving as the axis.

Next, an outer single-layer graphene sheet is grown to cover the innermost single-layer graphene sheet. As described above, growth to cover the inner single-layer graphene sheet with an outer single-layer graphene sheet is repeated, to form the graphene layer 16 surrounding the catalyst layer 14 .

During the growth process of the graphene layer 16 , single-layer graphene sheets included in each side graphene layer 16 s are stacked in the perpendicular direction (W 1 direction) on the side face of the catalyst layer 14 extending in the L-direction, and single-layer graphene sheets included in the upper graphene layer 16 u are stacked in the height direction (W 2 direction) of the catalyst layer 14 .

The effect of obstructing growth of graphene sheets increases in the vicinity of the underlayer 11 and the liner layer 13 . For this reason, the growth of graphene sheets is finished when growth of the outermost single-layer graphene sheet is obstructed by the underlayer 11 and the liner layer 13 .

As described above, the graphene layer 16 is formed to cover the catalyst layer 14 and the mask layer 15 .

Alternatively, the graphene layer 16 may be formed to surround the catalyst layer 14 and the mask layer 15 .

Alternatively, the graphene layer 16 may be formed to roll the catalyst layer 14 and the mask layer 15 from one side face of the catalyst layer to the other side face thereof, with the catalyst layer 14 and the mask layer 15 serving as axis.

The underlayer (insulating layer) 11 that obstructs growth of the graphene layer 16 is provided in a space with the adjacent catalyst layer 14 , to prevent contact between the adjacent graphene layers 16 .

Next, as illustrated in FIG. 12 , an insulating layer 21 is formed on the underlayer 11 to cover the graphene layer 16 and the liner layer 13 , and a mask pattern 22 is formed on the insulating layer 21 . The mask pattern 22 is formed above the catalyst layer 14 . In the W 1 direction, the width of the mask pattern 22 is preferably substantially equal to the width of the catalyst layer 14 opposed to the mask pattern 22 .

Next, as illustrated in FIG. 13 , the insulating layer 21 is patterned along the mask pattern 22 by a processing method such as photolithography, to form an insulating layer pattern 21 p . The insulating layer pattern 21 p functions as a mask used when the graphene layer 16 is processed. In the W 1 direction, the width of the insulating layer pattern 21 p is preferably substantially equal to the width of the catalyst layer 14 opposed to the mask pattern 22 .

Thereafter, as illustrated in FIG. 14 , the graphene layer 16 is processed with the insulating layer pattern 21 serving as a mask. In the processing, part of the graphene layer 16 formed outside the insulating layer pattern 21 p , that is, outside the side faces of the catalyst layer 14 extending in the L-direction is cut out.

As described above, at least one upper graphene layer 16 u and at least two side graphene layers 16 s are formed from one graphene layer 16 .

Next, as illustrated in FIG. 15 , the insulating layer pattern 21 p is removed, and an insulating layer 23 is formed on the underlayer 11 , to cover the upper graphene layer 16 u , the side graphene layers 16 s , and the liner layer 13 . A diffusion barrier layer (not illustrated), such as a silicon nitride layer, may be formed between the underlayer 11 and the insulating layer 23 , to cover the graphene layer 16 ( 16 u , 16 s ) and the liner layer 13 .

Lastly, the contact via 20 is formed in the insulating layer 23 , as illustrated in FIG. 4 to FIG. 7 .

As described above, the semiconductor device according to the first embodiment is finished.

In the above manner, according to the first embodiment, four or more catalyst-noncontacting edge faces 16 unc and 16 snc are cut out from one graphene layer 16 . This structure enables the reduction in electrical resistance to be lower than a graphene layer that does not have four or more edge faces. Accordingly, the resistance of the graphene layer 16 can be reduced, regardless of the processing limit or the width of the graphene layer 16 .

Second Embodiment

The following is explanation of a second embodiment. The basic structure and the manufacturing method of the present embodiment are the same as those of the first embodiment. Therefore, explanation of the matters explained in the first embodiment is omitted.

FIG. 16 and FIG. 17 are plan views illustrating a schematic structure of a semiconductor device according to the second embodiment. FIG. 17 is a cross-sectional view taken along line XVII-XVII of FIG. 15 .

As illustrated in the drawings, the second embodiment also has the structure in which a graphene layer 16 formed on a catalyst layer 14 includes at least four catalyst-noncontacting edge faces 16 unc and 16 snc . In addition, the second embodiment has the structure in which a circular graphene layer 16 is formed on a trapezoidal catalyst layer 14 . The trapezoidal catalyst layer 14 has a tapered shape, and has an upper face having a width less than that of a bottom face thereof in the W 1 direction. Specifically, each of the side faces of the catalyst layer 14 that extend in the L-direction is inclined.

The graphene layer 16 grows to form a circle from one side face of the catalyst layer 14 extending in the L-direction toward the other side face.

Each of single-layer graphene sheets included in each side graphene layer 16 s is formed in an arc shape from a catalyst-noncontacting edge face 16 snc toward a catalyst-contacting edge face 16 sc . Each of single-layer graphene sheets included in an upper graphene layer 16 u is formed in an arc shape from one catalyst-noncontacting edge face 16 unc toward the other catalyst-contacting edge face 16 unc . In the same manner as the first embodiment, the current in the graphene layer 16 flows in the interconnect length direction L.

FIG. 18 to FIG. 21 are diagrams illustrating connection relationships between a contact via 20 and the graphene layer 16 according to the second embodiment. In the same manner as in the first embodiment, when the mask layer 15 is conductive or no mask layer 15 is provided, the contact via 20 is connected to the graphene layer 16 using one of first to fourth connection methods ( FIG. 18 to FIG. 21 ). When the mask layer 15 is insulating, the contact via 20 is connected to the graphene layer 16 using one of the second to fourth connection methods ( FIG. 19 to FIG. 21 ).

The manufacturing method according to the second embodiment is the same as that according to the first embodiment, and explanation thereof is omitted.

As described above, the shape of the graphene layer 16 is changed according to the shape of the catalyst layer 14 . Examples of the shape of the catalyst layer 14 are a rectangle, a circle, a polygon, a hollow cylinder, and a layered structure.

As described above, the second embodiment produces the same effect as that produced by the first embodiment. In addition, according to the second embodiment, a circular graphene sheet is formed on the trapezoidal catalyst layer 14 .

While certain embodiments have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the inventions. Indeed, the novel embodiments described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the embodiments described herein may be made without departing from the spirit of the inventions. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the scope and spirit of the inventions.