Method of forming a fine pattern of a semiconductor device using a double patterning technique

1. Field of the Invention

The present invention relates to the manufacturing of semiconductor devices. More particularly, the present invention relates to a method of forming a fine pattern of a semiconductor device using a double patterning technique, e.g., using two generally coplanar hard mask patterns.

2. Description of the Related Art

A highly integrated semiconductor device referred to as a nanoscale device is formed of patterns whose features have widths and pitches (critical dimensions) below the resolution limits of photolithography. One such pattern of a highly integrated semiconductor device is an isolation layer which is used to define active regions of the device. The isolation layer is formed by producing a trench/recess in a substrate and filling the trench/recess with an insulating material. However, it is difficult to fill the trench/recess adequately with insulating material when sections of the trench/recess are narrow and the pitch of the sections of the trench/recess is fine, i.e., it is difficult to form an isolation layer whose critical dimensions are on the order of nanometers.

An object of the present invention it to provide a method of forming a pattern of a semiconductor device, whose topographic features are narrower and have a pitch smaller than that which can be formed using photolithography alone, i.e., an object of the present invention is to provide a method of forming a pattern not limited by the resolution of photolithography processes.

Another object of the present invention is to provide a method of forming a trench isolation layer for defining an active region, and by which the layer has excellent gap-filling characteristics.

Still another object of the present invention is to provide a relatively simple method of forming a fine pattern of a semiconductor device.

Still another object of the present invention is to provide a method of forming a trench isolation layer for defining an active region, which ensures that the isolation layer is not damaged when an etch mask used in the method is removed.

According to an aspect of the present invention, there is provided a method of forming a pattern of a semiconductor device in which a double patterning technique is used. A first hard mask layer is formed on a substrate, and a first mask pattern is formed on the first hard mask layer. Then, a buffer layer is formed over both sidewalls of each topographical feature of the first mask pattern. Next, a second mask pattern is formed adjacent to the first mask pattern with segments of the buffer layer interposed therebetween. In this respect, each topographical feature of the second mask pattern is interposed between adjacent ones of each respective pair of the topographical features of the first mask pattern. A first hard mask pattern which exposed portions of the substrate is then formed by etching the first hard mask layer using an etch mask selected from the group comprising the first mask pattern, the second mask pattern, and the buffer layer. A trench is formed in the substrate by etching the exposed portions of the substrate using the first hard mask pattern as an etch mask, and an isolation layer is formed in the trench. The isolation layer formed of an insulating material that is different in composition from that of the first hard mask pattern.

The first mask pattern and the second mask pattern may be formed of the same material as that of the first hard mask layer.

The first hard mask pattern may be formed by etching the first hard mask layer using the first mask pattern and the second mask patterns together as an etch mask. The first hard mask pattern may be formed by etching the first hard mask layer using the buffer layer as an etch mask.

The method may also include removing selected portions of the first mask pattern and the second mask pattern to form a trimmed first mask pattern and a trimmed second mask pattern, respectively, before the first hard mask pattern is formed. In this case, the first hard mask pattern is formed by etching the first hard mask layer using the trimmed first mask pattern and the trimmed second mask pattern together as an etch mask.

The method may include removing selected portions of the buffer layer to form a trimmed buffer layer, after the second mask pattern is formed but before the first hard mask pattern is formed. In this case, the first hard mask patterns is formed by etching the first hard mask layer using the trimmed buffer layer as an etch mask.

The method may also include forming a second hard mask layer before the first mask pattern is formed. In this case, the second hard mask layer is formed of a material different from that of the first hard mask layer. Also, recesses may be formed in a top surface of the second hard mask layer between the topographical features of the first mask patterns. In this case, the buffer layer may be formed in such a manner that a conformal segment of the buffer layer covers both opposite sidewalls of two adjacent topographical features of the first mask pattern and a lower surface of the second hard mask layer which delimits the bottom of the recess between the opposite sidewalls of the two adjacent topographical features of the first mask patterns.

These and other features, objects and advantages of the present invention will be more clearly understood from the following detailed description of the preferred embodiments thereof made in conjunction with the accompanying drawings in which:

FIG. 1 is a layout of active regions that can be realized by a method of forming a pattern of a semiconductor device according to the present invention;

FIGS. 2A through 2Q are cross-sectional views taken along line II-II′ of FIG. 1 , illustrating one embodiment of a method of forming a pattern of a semiconductor device according to the present invention; and

FIGS. 3A through 3H are cross-sectional views illustrating another embodiment of a method of forming a pattern of a semiconductor device according to the inventive present invention.

The present invention will now be described more fully with reference to the accompanying drawings. Like reference numerals are used to designate like elements throughout the drawings. Also, in the drawings, the thicknesses and widths of layers and regions are exaggerated for clarity. Furthermore, various features and regions in the drawings are illustrated schematically. Accordingly, the present invention is not limited by the relative sizes of or distances between features/regions shown in the attached drawings. Finally, it is to be noted that the term “features” as used throughout the specification refers to topographic features which are repeated periodically in a layer of the device, i.e., which together constitute a particular pattern.

Referring to FIG. 1 , a plurality of active regions 110 of a semiconductor device are defined by an isolation layer 120 . Each of the active regions 110 is in the form of an island having a short axis X 1 and a long axis Y 1 which intersect each other at a right angle. The short axis X 1 and the long axis Y 1 may respectively extend in directions that are parallel or oblique to the x-axis and the y-axis of FIG. 1 . A process of forming the active regions 110 on a substrate 200 will be explained with reference to FIGS. 2A through 2Q .

Referring to FIG. 2A , a pad oxide layer 202 is formed on the substrate 200 . A first hard mask layer 210 is formed on the pad oxide layer 202 . The substrate 200 may be a general semiconductor substrate such as a silicon substrate.

The first hard mask layer 210 is used as an etch mask when the substrate 200 is etched in order to define the active regions 110 . If the substrate 200 is a silicon substrate, the first hard mask layer 210 may be a polysilicon layer. The first hard mask layer 210 may be formed to a thickness of about 300 Å to about 1000 Å.

Referring to FIG. 2B , a second hard mask layer 220 and a first mask layer 230 are sequentially formed on the first hard mask layer 210 . If the first hard mask layer 210 is a polysilicon layer, the second hard mask layer 220 may be a silicon oxide layer. For example, the second hard mask layer 220 may be a medium temperature oxide (MTO) layer having a thickness of about 300 Å to about 600 Å. The first mask layer 230 may be a polysilicon layer.

Referring to FIG. 2C , a third hard mask layer 232 , an anti-reflection layer 234 , and a photoresist pattern 236 , which are necessary to pattern the first mask layer 230 using photolithography, are sequentially formed on the first mask layer 230 . The third hard mask layer 232 may be an amorphous carbon layer (ACL). The anti-reflection layer 234 may be a layer of SiON. The photoresist pattern 236 and more particularly, the features which constitute the photoresist pattern 236 , may have a pitch greater than the pitch of the fine pattern that is to be formed on the substrate 200 .

Referring to FIG. 2D , an anti-reflection layer pattern 234 A and a third hard mask pattern 232 A are formed by sequentially etching the anti-reflection layer 234 and the third hard mask layer 232 using the photoresist pattern 236 as an etch mask. Next, a first mask pattern 230 A is formed by etching the first mask layer 230 using the third hard mask pattern 232 A as an etch mask. Also, at this time, recessed portions 220 R of the second hard mask layer 220 may be formed by etching parts of the second hard mask layer 220 , which are exposed between the features of the first mask pattern 230 A, to a depth “d”.

Each of the features of the first mask pattern 230 A may have a linear shape that extends in a direction parallel to the long axis Y 1 of each of the active regions 110 ( FIG. 1 ). The features of the first mask pattern 230 A may be repeatedly formed on the substrate 200 at a first pitch P. The width PW 1 of each of the features of the first mask pattern 230 A in the x-axis direction may be equal to the width AW of each of the active regions 110 ( FIG. 1 ) in the x-axis direction.

By the time the third hard mask pattern 232 A is formed, part of the photoresist pattern 236 may be consumed.

Referring to FIG. 2E , top surfaces of the features of the first mask pattern 230 A are exposed by removing the third hard mask pattern 232 A, the anti-reflection layer pattern 234 A, and the photoresist pattern 236 . In this respect, the third hard mask pattern 232 A, the anti-reflection layer pattern 234 A, and the photoresist pattern 236 may be removed by ashing and stripping.

Referring to FIG. 2F , a conformal buffer layer 240 is formed over the exposed surfaces of the first mask pattern 230 A and the recessed portions 220 R of the second hard mask layer 220 . The buffer layer 240 may be an oxide layer or a nitride layer formed by atomic layer deposition (ALD). In any case, as a result, recesses 242 are formed at the top surface of the buffer layer 240 in such a manner that each of the recesses 242 is disposed between each respective pair of adjacent features of the first mask pattern 230 A. Due to the buffer layer 240 , the features of a second mask pattern 250 A (see FIG. 2H ) that are to be formed in the recesses 242 can have the same widths and heights as the features of the first mask pattern 230 A that are to be used as an etch mask in a subsequent process.

In this respect, the buffer layer 240 can uniformly cover top surfaces and sidewalls of the first mask pattern 230 A and the recessed portions 220 R of the second hard mask layer 220 . For example, the thickness of the buffer layer 240 may be the same as the depth “d” ( FIG. 2D ). Also, the width RW of each of the recesses 242 in the x-axis direction may be equal to the width AW of each of the active regions 110 ( FIG. 1 ) in the x-axis direction.

Furthermore, the buffer layer 240 is preferably formed of a material having etching characteristics similar to those of the material from which the second hard mask layer 220 is formed. That is, the buffer layer 240 may be formed of the same material as that used to form the second hard mask layer 220 or the buffer layer 240 may be formed of a material that is different from but has similar etching characteristics as the material from which the second hard mask layer 220 is formed. For example, each of the second hard mask layer 220 and the buffer layer 240 may be an oxide layer.

Referring to FIG. 2G , a second mask layer 250 is formed over the buffer layer 240 on the substrate 200 to such a thickness that the recesses 242 are filled by the second mask layer 250 . The second mask layer 250 may be formed of the same material as that of the first mask layer 230 . For example, the second mask layer 250 may be a polysilicon layer.

Referring to FIG. 2H , a second mask pattern 250 A is formed in the recesses 242 by removing an upper portion of the second mask layer 250 . More specifically, the second mask layer 250 is etched. For example, the second mask layer 250 is wet etched. As a result, those portions of the buffer layer 140 covering the first mask pattern 230 A are exposed between respective features of the second mask pattern 250 A formed by removing an upper portion of the second mask layer 250 . The etch rate of the second mask layer 250 may be adjusted so that the upper surface of the resulting second mask pattern 250 A is level with the upper surface of the first mask pattern 230 A.

Each of the features of second mask pattern 250 A may have a linear form that extends longitudinally in the same direction in which the features of the first mask pattern 230 A extend longitudinally, e.g., parallel to the long axis Y 1 of each of the active regions 110 ( FIG. 1 ). The width PW 2 of each of the features of the second mask pattern 250 A in the x-axis direction may be equal to the width AW of each of the active regions 110 in the x-axis direction. Also, the features of the second mask pattern 250 A may be substantially horizontally aligned with the features of the first mask pattern 230 A. That is, the upper surfaces of the features of the second mask pattern 250 A may lie within the same horizontal plane as the upper surfaces of the features of the first mask pattern 230 A, and the lower surfaces of the features of the second mask pattern 250 A may lie within the same horizontal plane as the lower surfaces of the features of the first mask pattern 230 A.

Referring to FIG. 2I , those portions of the buffer layer 240 which cover the first mask pattern 230 A are removed to expose top surfaces of the features of the first mask pattern 230 A For example, the buffer layer 240 is dry etched. If the buffer layer 240 is an oxide layer, the etching gas employed in the dry etching process may be CxFy (wherein each of x and y is an integer of 1 to 10), or a mixture of CxFy, O 2 , and Ar. More specifically, the CxFy may be C 4 F 6 or C 4 F 8 . Alternatively, the buffer layer 240 is wet etched. For example, if the buffer layer 240 is an oxide layer, an etchant containing fluorine (F), such as diluted HF (DHF), NH 4 F, or a combination thereof, may be used in order to selectively wet etch the buffer layer 240 . Again, an appropriate selection of materials ensures that the buffer layer 240 has a significantly greater etch rate than the etch rates of the first mask pattern 230 A and the second mask pattern 250 A.

In any case, both the top surfaces of the features of the first mask pattern 230 A and the top surfaces of the features of the second mask pattern 250 A are exposed on the substrate 200 as a result.

Note, though, the process described with reference to FIG. 2I may be omitted in certain circumstances.

Referring to FIG. 2J , a fourth hard mask layer 262 , an anti-reflection layer 264 , and a photoresist pattern 266 are sequentially formed to cover the first mask pattern 230 A, the second mask pattern 250 A, and the remnants of the buffer layer 240 . For example, the fourth hard mask layer 262 may be an ACL. The anti-reflection layer 264 may be formed of SiON.

The photoresist pattern 266 defines a plurality of openings 266 H. The width HW 1 of each of the openings 266 H in the x-axis direction may be equal to or greater than the width PW 1 of each of the features of the first mask pattern 230 A in the x-axis direction. Alternatively, the width HW 1 of each of the openings 266 H in the x-axis direction is equal to or greater than the width PW 2 of each of the features of the second mask pattern 250 A in the x-axis direction.

Referring to FIG. 2K , the anti-reflection layer 264 and the fourth hard mask layer 262 are sequentially etched, using the photoresist pattern 266 as an etch mask, to form an anti-reflection layer pattern 264 A and a fourth hard mask pattern 262 A. At this time, part of the photoresist pattern 266 may be consumed. In any case, and as a result, openings 262 H are formed between the features of the fourth hard mask pattern 262 A. The openings 262 H expose respective features of the first mask pattern 230 A and the second mask pattern 250 A. The second hard mask layer 220 may also be exposed through the plurality of openings 262 H. Furthermore, if the openings 262 H are each wide enough, the remnants of the buffer layer 240 disposed around the exposed features of the first mask pattern 230 A and the second mask pattern 250 A can be exposed through the openings 262 H.

Next, those features of the first mask pattern 230 A and the second mask pattern 250 A which are exposed through the openings 262 H are etched away using the fourth hard mask pattern 262 A and the remnants of the buffer layer 240 as an etch mask. As a result, a trimmed first mask pattern 230 B and a trimmed second mask pattern 250 B of features having shapes in the form of islands similar to those of the active regions 110 ( FIG. 1 ) are obtained.

Referring to FIG. 2L , top surfaces of the trimmed first mask pattern 230 B and the trimmed second mask pattern 250 B are exposed by removing the fourth hard mask pattern 262 A, the anti-reflection layer pattern 264 A, and the photoresist pattern 266 . In this respect, the fourth hard mask pattern 262 A, the anti-reflection layer pattern 264 A, and the photoresist pattern 266 may be removed by ashing and stripping.

Referring to FIG. 2M , the remnants of the buffer layer 240 are anisotropically etched using the trimmed first mask pattern 230 B and the trimmed second mask pattern 250 B as an etch mask. As a result, the second hard mask layer 220 is exposed between the features of the trimmed first mask pattern 230 B and the features of the trimmed second mask pattern 250 B. The exposed portion of the second hard mask layer 220 is anisotropically dry etched to form a second hard mask pattern 220 A which exposes the first hard mask layer 210 .

If each of the second hard mask layer 220 and the buffer layer 240 is an oxide layer, the remnants of the buffer layer 240 and the second hard mask layer 220 can be dry etched at a rate higher than the trimmed first mask pattern 230 B and the trimmed second mask pattern 250 B which are formed of polysilicon, using a gas mixture of CxFy (wherein x is an integer of 1 to 6 and y is an integer of 3 to 8) and O 2 . For example, C 3 F 8 , C 4 F 6 , C 4 F 8 , or C 5 F 8 mixed with O 2 in a volumetric ratio of 1:1 may be used as an etching gas. If necessary, Ar may be added to the etching gas. Also, the dry etching may be performed in a plasma atmosphere that is obtained by exciting the etching gas.

If each of the second hard mask layer 220 and the buffer layer 240 is a nitride layer, the remnants of the buffer layer 240 and the second hard mask layer 220 can be dry etched at a higher rate than the trimmed first mask pattern 230 B and the trimmed second mask pattern 250 B which are formed of polysilicon, using CHxFy (wherein x and y are each an integer of 1 to 3 and x+y=4). For example, the etching gas can be CH 2 F 2 , CH 3 F, or a mixture thereof. If necessary, O 2 may be added to the etching gas. Also, the dry etching may be performed in a plasma atmosphere that is obtained by exciting the etching gas.

Referring to FIG. 2N , a first hard mask pattern 210 A is formed by anisotropically etching the first hard mask layer 210 using the second hard mask pattern 220 A and the remnants of the buffer layer 240 as an etch mask. If the trimmed first mask pattern 230 B and the trimmed second mask pattern 250 B are of the same material as the first hard mask layer 210 , the trimmed first mask pattern 230 B and the trimmed second mask pattern 250 B covering the second hard mask pattern 220 A and the remnants of the buffer layer 240 may be removed while the first hard mask layer 210 is etched.

A gas mixture of HBr and O 2 may be used as an etching gas to dry etch the first hard mask layer 210 . As an example, the HBr and O 2 may be supplied at flow rates having a ratio of about 10:1 to about 30:1. Also, He may be added to the etching gas. More specifically, HBr may be supplied at a flow rate of about 100 sccm to about 300 sccm, O 2 may be supplied at a flow rate of about 5 sccm to about 30 sccm, and He may be supplied at a flow rate of about 50 sccm to about 200 sccm. If necessary, Cl 2 or a gas mixture of HBr and Cl 2 may be used instead of HBr. Also, the first hard mask layer 210 may be etched in a plasma atmosphere that is obtained by exciting the etching gas.

Referring to FIG. 2O , a trench 280 is formed in the substrate 200 by etching the pad oxide layer 202 to expose those portions of a top surface of the substrate 200 spanning the features of the first hard mask pattern 210 A, and etching the exposed portions of the substrate 200 using the first hard mask pattern 210 A as an etch mask. When viewed from above, the trench 280 may have the same shape as that of the isolation layer 120 ( FIG. 1 ). That is, the shape of the opening in the upper surface of the substrate 200 formed by the trench 280 may be the same as the shape of the isolation layer 120 in horizontal section.

Furthermore, the trench 280 may have first sections 280 A and second sections 280 B spaced laterally from the first sections, and in which the width of each first section 280 A is smaller than the width of each second section 280 B. Referring back to FIG. 1 , there are two different distances AD 1 and AD 2 between adjacent active regions 110 of FIG. 1 in the x-axis direction, wherein the distance AD 1 is smaller than the distance AD 2 . The width TW 1 of a first section 280 A of the trench 280 (the dimension of the first section 280 A in the x-axis direction) corresponds to the distance AD 1 , and the width TW 2 of a second section 280 B of the trench (the dimension of the second section 280 B in the x-axis direction) corresponds to the distance AD 2 .

Also, as shown in FIG. 2O , the second hard mask pattern 220 A and remnants of the buffer layer 240 are not left on the first hard mask pattern 210 A of the resultant structure in which the trench 280 is formed. However, if necessary, the second hard mask pattern 220 A or both the second hard mask pattern 220 A and the remnants of the buffer layer 240 may be left on the first hard mask pattern 210 A after the trench 280 is formed.

Referring to FIG. 2P , an isolation layer 282 is formed in the trench 280 by depositing an insulating material in the trench 280 and on the first hard mask pattern 210 A and planarizing the resultant structure using chemical mechanical polishing (CMP) until the first hard mask pattern 210 A is exposed. The isolation layer 282 may correspond to the isolation layer 120 shown in FIG. 1 .

The isolation layer 282 may be formed of an insulating material that is different from that from which the first hard mask pattern 210 A is formed. For example, the isolation layer 282 may include an oxide layer liner contacting the substrate 200 in the trench 280 , and gap-filling insulating material including a nitride filling the remainder of the trench 280 . Alternatively, the isolation layer 282 may include an oxide layer liner contacting the substrate 200 in the trench 280 , a nitride layer liner formed on the oxide layer liner, and gap-filling insulating material of hydropolysilizane-based inorganic SOG, such as TOSZ, formed on the nitride layer liner and filling the remainder of the trench 280 .

Referring to FIG. 2Q , the first hard mask pattern 210 A is removed from the resultant structure including the isolation layer 282 . The active regions 110 are defined by the isolation layer 282 .

According to the method shown in and described with respect to FIGS. 2A through 2Q , the isolation layer 282 and the first hard mask pattern 210 A that are used as an etch mask to form the trench 280 are formed of different materials. Accordingly, the isolation layer 282 can be prevented from being damaged while the first hard mask pattern 210 A is removed after the isolation layer 282 is formed. In particular, when the trench 280 needs to have very small dimensions suitable for a highly integrated semiconductor device, a nitride must be used as insulating material for filling the trench 280 because nitrides have good gap-filling properties. To this end, according to an embodiment of the present invention, the gap-filling insulating layer is a nitride layer and the first hard mask pattern 210 A is a polysilicon layer. Thus, the isolation layer 282 is not damaged when the hard mask pattern 210 A is removed.

Moreover, according to an embodiment of the present invention, the first hard mask pattern 210 A used as an etch stop layer during CMP (the process used to form the isolation layer 282 ) is a polysilicon layer. Accordingly, only polysilicon and oxide layers having a high etch selectivity can be used as etch masks for forming the trench 280 . Therefore, the double patterning process of forming the isolation layer 282 for defining the active regions 110 is made relatively simple.

Another method of forming a pattern of a semiconductor device according to the present invention will now be described with reference to FIGS. 3A through 3H . In this regard, the layers/features, etc. shown in and described with respect to the embodiment of FIGS. 3A through 3H and which are similar to those shown in and described with respect to the embodiment FIGS. 2A through 2Q are denoted by the same reference numerals, and the steps/processes used to form such layers/features will not be described in detail. Basically, though, the method of the embodiment of FIGS. 3A through 3H is different from the method of the embodiment of FIGS. 2A through 2Q in that the remnants of the buffer layer 240 , instead of the first mask pattern 230 A and the second mask pattern 250 A, are used as an etch mask to pattern the first hard mask layer 210 .

Referring to FIG. 3A , the first mask pattern 230 A and the second mask pattern 250 A are formed in such a manner that the remnants of the buffer layer 240 are disposed between the features of the first mask pattern 230 A and the features of the second mask pattern 250 A as described with reference to FIGS. 2A through 2I .

Next, a trimmed mask pattern is formed on the first mask pattern 230 A, the second mask pattern 250 A, and the remnants of the buffer layer 240 in a manner similar to that described with reference to FIG. 2J . However, the remnants of the buffer layer 240 are divided by the trimmed mask layer into a plurality of features whose surfaces have the same shapes as those of the active regions 110 ( FIG. 1 ), unlike in the method of FIG. 2J .

In FIG. 3A , a fourth hard mask layer 362 , an anti-reflection layer 364 , and a photoresist pattern 366 are sequentially formed as the trimmed mask layer to cover the first mask pattern 230 A, the second mask pattern 250 A, and the remnants of the buffer layer 240 . The fourth hard mask layer 362 , the anti-reflection layer 364 , and the photoresist pattern 366 may be similar to the fourth hard mask layer 262 , the anti-reflection layer 264 , and the photoresist pattern 266 described with reference to FIG. 2J , respectively. However, the relative positions of the openings 366 H defined by the photoresist pattern 366 are different from those of the openings 266 H in the photoresist pattern 266 described with reference to FIG. 2J .

Referring to FIG. 3B , an anti-reflection layer pattern 364 A and a fourth hard mask pattern 362 A are formed by sequentially etching the anti-reflection layer 364 and the fourth hard mask layer 362 using the photoresist pattern 366 as an etch mask in a manner similar to that described with reference to FIG. 2K . As a result, a plurality of openings 362 H are formed between the features of the fourth hard mask pattern 362 A. Portions of the remnants of the buffer layer 240 are exposed through the openings 362 H formed between the features of the fourth hard mask pattern 362 A. If each of the plurality of openings 362 H is wide enough, the features of the first mask pattern 230 A and the second mask pattern 250 A disposed around the exposed portions of the remnants of the buffer layer 240 may also be exposed through the openings 362 H.

Also, part of the photoresist pattern 366 may be consumed while the fourth hard mask pattern 362 A is formed.

Next, the exposed portions of the remnants of the buffer layer 240 are etched using the fourth hard mask pattern 362 A, the first mask pattern 230 A, and the second mask pattern 250 A as an etch mask. As a result, portions of the second hard mask layer 220 which are exposed. Then the exposed portions of the second hard mask layer 220 are etched. Accordingly, a trimmed buffer layer 240 C and a trimmed second hard mask layer 220 C are formed on the substrate 200 , and portions of the first hard mask layer 210 are exposed through the openings 362 H.

Referring to FIG. 3C , the fourth hard mask pattern 362 A, the anti-reflection layer pattern 364 A, and the photoresist pattern 366 are removed to expose the top surfaces of the trimmed buffer layer 240 C, the first mask pattern 230 A, and the second mask pattern 250 A.

Referring to FIG. 3D , the first mask pattern 230 A and the second mask pattern 250 A are etched using the trimmed buffer layer 240 C as an etch mask.

If the first mask patterns 230 A and the second mask pattern 250 A are formed of polysilicon, and the trimmed buffer layer 240 C and the trimmed second hard mask layer 220 C are oxide layers, the first mask pattern 230 A and the second mask pattern 250 A may be removed by wet etching or dry etching. An etchant containing NH 4 OH may be used to wet etch the first mask pattern 230 A and the second mask pattern 250 A. For example, NH 4 OH, H 2 O 2 , and H 2 O mixed in a volumetric ratio of 4:1:95 may be used as the wet etchant. Alternatively, an etching gas containing CF 4 may be used to dry etch the first mask pattern 230 A and the second mask pattern 250 A. For example, a gas mixture of CF 4 and O 2 , or a gas mixture of CF 4 , O 2 , N 2 , and HF may be used to isotropically chemically dry etch (CDE) the first mask pattern 230 A and the second mask pattern 250 A.

As a result, a plurality of vertically extending parts 240 V and a plurality of horizontally extending parts 240 H of the trimmed buffer layer 240 C are exposed as shown in FIG. 3D .

Referring to FIG. 3E , the trimmed second hard mask layer 220 C and the horizontally extending parts 240 H of the trimmed buffer layer 240 C are removed by an anisotropic dry etching process. As a result, the top surface of the first hard mask layer 210 is exposed between the vertically extending parts 240 V of the trimmed buffer layer 240 C. During this process, the vertically extending parts 240 V of the trimmed buffer layer 240 C are consumed from the top. Subsequently, a mask of remnants of the trimmed second hard mask layer 220 C and remnants of the vertically extending parts 240 V of the trimmed buffer layer 240 C are left on the first hard mask layer 210 .

Referring to FIG. 3F , the first hard mask layer 210 is anisotropically dry etched using the aforementioned mask as an etch mask. Thus, a first hard mask pattern 210 C is formed.

Referring to FIG. 3G , the pad oxide layer 202 is etched to expose the top surface of the substrate 200 between features of the first hard mask pattern 210 C, and the exposed portions of the substrate 200 are etched using the first hard mask pattern 210 C and the etched pad oxide layer 202 as an etch mask. Accordingly, a trench 380 is formed in the substrate 200 . The trench 380 may include a first section 380 A, and a second section 380 B which is wider than the first section 380 A.

Referring to FIG. 3H , insulating material is deposited in the trench 380 and over the first hard mask pattern 210 C. The resultant structure is planarized using CMP until the first hard mask pattern 210 C is exposed in the same manner as described with reference to FIG. 2P and FIG. 2Q . The first hard mask pattern 210 C is then removed.

As a result, an isolation layer 382 is formed in the trench 380 . The active regions 110 A are defined on the substrate 200 by the isolation layer 382 .

According to the method of FIGS. 3A through 3H , the isolation layer 382 and the first hard mask pattern 210 C that are used as an etch mask to form the trench 380 are formed of different materials, and the first hard mask pattern 210 C is removed after the isolation layer 382 is formed. Therefore, the device isolation later 382 can be prevented from being damaged. In particular, the insulating layer filling the trench 380 is a nitride layer and the first hard mask pattern 210 C is a polysilicon layer. Thus, the isolation layer 382 will not be damaged when the hard mask pattern 210 C is removed.

Furthermore, the first hard mask pattern 210 C is used as an etch stop during CMP to form the isolation layer 382 is a polysilicon layer, and only polysilicon layers and oxide layers having a high etch selectivity are used as etch masks for forming the trench 380 . Therefore, the double patterning technique of forming the isolation layer 382 which defines the active regions 110 A is relatively simple.

Finally, although the present invention has been particularly shown and described with reference to the preferred embodiments thereof, it will be understood that various changes in form and details may be made therein without departing from the true spirit and scope of the invention as defined by the following claims.