FinFET semiconductor device having increased gate height control
A semiconductor device includes a silicon-on-insulator (SOI) substrate having a buried oxide (BOX) layer, and a plurality of semiconductor fins formed on the BOX layer. The plurality of semiconductor fins include at least one pair of fins defining a BOX region therebetween. Gate lines are formed on the SOI substrate and extend across the plurality of semiconductor fins. Each gate line initially includes a dummy gate and a hardmask. A high dielectric (high-k) layer is formed on the hardmask and the BOX regions. At least one spacer is formed on each gate line such that the high-k layer is disposed between the spacer and the hardmask. A replacement gate process replaces the hardmask and the dummy gate with a metal gate. The high-k layer is ultimately removed from the gate line, while the high-k layer remains on the BOX region.
1. A method of fabricating a semiconductor device, comprising:
forming a plurality of semiconductor fins and a plurality of gate lines on a silicon-on-insulator (SOI) substrate, each gate line comprising a gate stack including a dummy gate and a hardmask formed on the dummy gate;
selectively forming a gate stack high dielectric (high-k) layer on the hardmask; and
forming at least one spacer on each gate line, the gate stack high-k layer of each gate line disposed between the at least one spacer and the hardmask.
2. The method of claim 1 , further comprising forming a high-k layer on a buried oxide (BOX) region of the SOI substrate located between each semiconductor fin.
3. The method of claim 2 , further comprising etching the at least one spacer via a first spacer etching process such that a height of the at least one spacer is less than a height of the gate stack high-k layer.
4. The method of claim 3 , wherein the etching the at least one spacer further comprises etching a first spacer via the first spacer etching process.
5. The method of claim 4 , wherein the etching the at least one spacer further comprises forming a second spacer over the first spacer etched according to the first spacer etching process and etching the second spacer via a second spacer etching process such that the first and second spacers have a height less than a height of the gate stack high-k layer.
6. The method of claim 5 , further comprising etching a first portion of the gate stack high-k layer via a high-k etching process such that the dummy gate is exposed.
7. The method of claim 6 , wherein the high-k etching process selectively etches only the gate stack high-k layer while inhibiting etching of the first and second spacers.
8. The method of claim 7 , further comprising etching a remaining gate stack high-k layer disposed on the first and second spacer via the high-k etching process after removing the hardmask.
9. The method of claim 8 , further comprising removing the dummy gate to form a void in the gate stack and forming a metal gate in the void.
10. The method of claim 2 , wherein the forming the high-k layer and the gate stack high-k layer further comprises selectively growing HfO 2 on each hardmask and each BOX region while inhibiting growth of the HfO 2 on the semiconductor fins.
11. A method of preventing spacer divots in a gate stack of a SOI semiconductor finFET, the method comprising:
simultaneously forming a first high dielectric (high-k) layer on a buried oxide (BOX) region located between pairs of semiconductor fins and a second high-k layer on a hardmask disposed on a dummy gate of the gate stack;
forming at least one spacer on each gate line such that the second high-k layer is disposed between the at least one spacer and the hardmask;
etching the at least one spacer via a first etching process such that a height of the at least one spacer is less than a height of the second high-k layer; and
etching a first portion of the second high-k layer via a high-k etching process that is different from the first etching process to expose the hardmask.
12. The method of claim 11 , wherein the at least one spacer withstands etching from the high-k etching process.
13. The method of claim 12 , further comprising:
removing the hardmask; and
etching a remaining portion of the second high-k layer disposed adjacent the at least one spacer via the high-k etching process after removing the hardmask such that the first high-k layer remains on the BOX region.
14. The method of claim 13 , further comprising:
removing the dummy gate to form a void in the gate stack; and
forming a metal gate in the void.
15. The method of claim 12 , wherein the simultaneously forming the first and second high-k layers comprises growing first and second high-k layers comprising HfO 2 via atomic layer deposition (ALD).
16. The method of claim 15 , wherein the ALD comprises positive patterning ALD.
17. The method of claim 16 , wherein the positive pattern ALD comprises depositing blocking co-polymers at the semiconductor fins to prevent growth of the HfO 2 thereon.
18. The method of claim 15 , wherein the ALD comprises ALD nucleation.
19. The method of claim 18 , wherein the semiconductor fins are formed from silicon (Si) to inhibit growth of HfO 2 thereon, and the BOX region and the hardmask comprise SiO 2 and SiN, respectively, to promote growth of the HfO 2 thereon.