A Novel Resist Freeze Process for Double Imaging

Transcription

1 A Novel Resist Freeze Process for Double Imaging David J. Abdallah, Eric Alemy, Srinivasan Chakrapani, Munirathna Padmanaban and Ralph R. Dammel AZ Electronic Materials Somerville, NJ USA 1 st exp 2 nd exp The information contained herein is, as far as we are aware, true and accurate. However, no representations or warranties, either express or implied, whether of merchantable quality, fitness for any particular purpose or of any other nature are hereby made in respect of the information contained in this presentation or the product or products which are the subject of it. In providing this material, no license or other rights, whether express or implied, are given with respect to any existing or pending patent, patent application, trademarks, or other intellectual property right.

2 Dual imaging with photoresist freeze 1 st Positive photoresist coating approaches Broad resist compatibility wafer cycle time Ion Implant and EB cure Ext. vacuum process 2 nd Positive photoresist coating Pitch which is ½ the optical limit Polymer Encasing Thermal Curing UV Exposure > 3 track modules 1 track module 1 track module circumvents many processing steps in double patterning more generally applicable than spacer approaches Vapor Reaction Chamber Works w/ lactone 1 track module AZ has focused on developing a dual imaging approach that uses commercially available positive photoresist and standard track hardware to reduce cost, risk and time to market of double imaging page 2

3 Exposure and VRC conditions for double imaging Nikon NSR-306D (NA: 0.85 dry) Dipole illumination σ o 0.82, σ i % phase shift 1:1 90nm L/S 1 st Positive Resist image AX2110P 90nm FT PAB 100 o C/60s, PEB 110 o C/60s 1C5D 37nm FT, 200 o C/60sec Dev. 2.6N TMAH aqueous solution, 60s puddle w/ DI rinse 180nm Nitrogen gas pressure regulator Exhaust Lid Chamber Flow meter Hot Plate Nitrogen gas manifold valve Bubbler 90nm (45nm L/S) 2 nd Positive Resist image Same as 1 st exposure w/ out BARC Exposure field placement is shifted to place features in between features of the first exposure Freeze process variables: freeze material flow rate temperature time page 3

4 Freeze liquids screening by solvent resistance test Freeze liquid selection Results of solvent resistance test after VRC processed images # Freeze liquid BP ( o C) VRC Time (min) CD before soak (nm) CD after soak* (nm) AX 2110P 45nm pitch 85nm Ref No VRC processing No Image 1 1,2-Diaminoethane ,2-Diaminoethane nm 3 1,3-Diaminopropane ,3-Diaminopropane ,5-Diamino-2- methylpentane ** 1,5-Diamino-2- methylpentane ** 7 1-Aminopentane ** 8 N-Methylbutylamine VFI 9 Triethylamine VFI 10 Acetic acid No image 11 Water No image *ArF Thinner used for solvent soak ** visual difference in film after soaking VFI = very faint image. VRC conditions; flow rate = 2.5L/min, temperature = 180 o C page 4

5 FT-IR analysis of AX 2110P resin being frozen with DAE Temperature/Time dependence VRC temp. 180 o C Absorbance VRC temp. 100 o C Absorbance Wavenumber (cm -1 ) Wavenumber (cm -1 ) sec 2min 8min min 1hr W b ( 1) Tg of resin in 2110P is 175 o C * n n R 100 o C 170 o C 180 o C DAE, flow rate 3L/min Temperature uniformity w/ Lab VRC Δ15 o C * page 5

6 SEM images exhibiting process margins for both exposures First exposure VRC Second exposure μm, 49nm +8% Dose BF 50nm μm, 46nm NH 2 H 2 N Conditions DAE, flow 3L/min, 180 o C/2min μm, 59nm +8% Dose BF 52nm μm, 47nm 52mJ, BF 53nm Post VRC: 200oC/60s, Soak ArF Thinner As will be shown Post VRC processing not necessary w/ 180 o C 52mJ, BF 57nm -8% Dose BF 62nm -8% Dose BF 60nm page 6

7 FT-IR difference spectra analysis FT-IR analysis Δ Absorbance Difference spectra after 2 minutes 3L/min, 180 o C/2min N2 only N2 w/ DAE vapors * n n * R NH H 2 N 2 * n n R HN H H NH * Wavenumber (cm -1 ) * n R * n Decreased absorptions lactone carbonyl stretching absorptions at 1796 cm-1 (meth)acrylate carbonyl absorption at 1743 cm-1 (slight) Increase absorptions amide carbonyl stretching absorptions at 1693 cm-1 amide band 2 around 1535 cm-1 broad hydroxyl absorptions at 3425 cm-1 (H-bonding) C- at 1068 cm-1 NH2 scissoring 1545 cm-1. page 7

8 Determining across wafer crosslinking uniformity by FT-IR Crosslinking % (degree of lactone reactions with amino groups) Abs. lactone carbonyl (1797cm-1) ( ) x 100 Abs. ester carbonyl (1727cm-1) Absorbance For a blanket AX2110P film Abs. lactone carbonyl (1797cm-1) 0.94 = Abs. ester carbonyl (1727cm-1) (not processed through the VRC) Wavenumber (cm-1) Crosslinking reduces the lactone carbonyl absorption relative to the ester carbonyl. Uniformity is based on the ratio of the peak heights page 8

9 Across wafer crosslinking of blanket coated AX2110P no VRC processing (normal variation) PR VRC condition for AX2110P; 180oC w/ DAE 3L/min Absorbance Wavenumber (cm-1) Wavenumber (cm-1) 1400 Absorbance Wavenumber (cm-1) Wavenumber (cm-1) 1400 Data is in % lactone reacted Data is in % lactone reacted Reference wafer has about 2% variation (ratio method is valid) Greater than 10% crosslinking of the lactones is needed for freezing since 2 minutes of VRC is required so the film at the wafer edge freezes, however, in order to achieve this nearly 50% of the lactones are crosslinked at the center of the wafer where the temperature is the hottest using our VRC chamber Temperature uniformity w/ Lab o C Δ15 o C page 9

10 Etch rates and optical indices of original and frozen AZ AX2110P Etch analysis Etch rate (nm/s) of AZ AX 2110P after coating, baking, and freezing no VRC 180 o C/N 2 /120s 180 o C/DAE/120s CF4/ CF Measure on blanket coatings, VRC flow rate 3L/min 2 to 5% increase in etch rate for frozen resist Time (s) Etch conditions 2 (SCCM) CF4 (SCCM) N2 (SCCM) AR (SCCM) Pressure (Pa) Top power (W) wafer power (W) CF4/ CF VASE analysis VRC conditions Bake time (min) Bake temp ( 0 C) Freeze liquid Flow rate (L/min) Vase analysis vase ft (nm) % FT lose Cauchy A Cauchy B (μm 2 ) Cauchy C (μm 4 ) densification 2min Mass added + restricted 2min 180 DAE 3 densifiaction SAME n@ 193 k@ 193 page 10

11 Results of E-beam exposure on frozen images After VRC REF After VRC and SEM EXP A After VRC and SEM EXP B and Soak in 300MIF After VRC and SEM EXP B and Soak in ArF Thinner CD = 67nm 59nm 47nm 51nm AZ SEM exposure conditions are harsh to exaggerate the problem E-beam exposure will deprotect polymers leading to some shrinkage Some CD slimming was observed after e-beam exposure which was further trimmed after soaking. Base developer had a bigger impact since the crosslinked polymer is more hydrophilic No missing patterns were observed We will also look into the same experiments with no VRC processing. page 11

12 Process margins comparison of AX 2110P in both exposures Exposure sensitivity CD (nm) Dose used in 2 nd exposure (mj) 1 st Exp 2 nd Exp Frozen image Sensitivity (nm/mj) (ΔmJ, 0.1μm) 1.1 (ΔmJ, 0.1μm) 1.5 (52mJ, 0.1μm) 0.6 Impact of frozen image on 2nd exp. Impact of 2nd exp. on frozen image Focus sensitivity CD (nm) Focus used Focus in (μm) 2 nd 2nd exposure Exposure(μm) 1 st Exp 2 nd Exp Frozen image (52mJ, Δμm) (52mJ, Δμm) (52mJ, 0.1μm) Impact of frozen image on 2nd exp. Impact of 2nd exp. on frozen image page 12

13 Examples of dual imaging with VRC freeze (90nm pitch) st exp 52mJ AZ AX 2110P VRC: DAE, flow 3L/min, 180 o C/2min Post VRC: none 2 nd exp 60mJ AZ AX 2110P 90nmFT 1 st exp 52mJ AZ AX 2110P 1 st exp 52mJ AZ AX 2110P VRC: DAE, flow 3L/min, 180 o C/2min VRC: DAE, flow 3L/min, 180 o C/2min Post VRC: none Post VRC: AZ MIF Developer 2 nd exp 60mJ AZ AX 2110P 110nmFT 2 nd exp 60mJ AZ AX 2110P st exp 45mJ AZ AX 2110P VRC: DAE, flow 3L/min, 100 o C/60min Post VRC: 200oC/60s, Soak ArF Thinner 2 nd exp 60mJ AZ AX 2110P 1 st exp 51mJ AZ AX 2110P VRC: DAE, flow 3L/min, 180 o C/2min Post VRC: none 2 nd exp 51mJ AZ AX 2110P n SiN, no BARC 1 st exp 51mJ AZ AX 2110P VRC: DAE, flow 3L/min, 180 o C/2min Post VRC: none 2 nd exp 57mJ AZ AX 2110P n SiN (K=0.77), no BARC page 13

14 Examples of dual imaging with VRC freeze (90nm pitch) 1 st exp 60mJ AZ AX 2110P VRC: DAE, flow 3L/min, 180 o C/2min Post VRC: none 2 nd exp 60mJ AZ AX 2110P 1 st exp 52mJ AZ AX 2110P VRC: DAP, flow 2.5L/min, 180 o C/2min Post VRC: 200 o C/60s, Soak ArF Thinner 2 nd exp 52mJ AZ AX 2110P 1 st exp 52mJ AZ AX 2110P VRC: 1,5DAP, flow 2.5L/min, 180 o C/4min Post VRC: 200 o C/60s, Soak ArF Thinner 2 nd exp 52mJ AZ AX 2110P 1 st exp 60mJ AZ AX 2110P VRC: DAE, flow 3L/min, 180 o C/2min Post VRC: none 2 nd exp 60mJ AZ AX 2110P 1 st exp 60mJ AZ AX 2110P VRC: DAE, flow 3L/min, 180 o C/2min Post VRC: none 2 nd exp 60mJ AZ AX 3110P 1 st exp 52mJ AZ AX 3110P VRC: DAE, flow 3L/min, 170 o C/2min Post VRC: none 2 nd exp 60mJ AZ AX 3110P page 14

15 Profile variations in dual imaging 1st exp 52mJ VRC NH H 2 N 2 Conditions DAE, flow 3L/min, 180 o C/2min Post VRC: 200 o C/60s, Soak ArF Thinner 2nd exp 60mJ Impact of second exposure on frozen image understood 1 st image 2 nd image Impact of 1. worn substrate 2. frozen image on 2 nd exposure Not well Understood page 15

16 Impact of 2 nd exposure on frozen image 1st exp AZ AX 2110P exposed lines-space 52mJ. VRC DAE, flow 2.5L/min, 180 o C/2min NH H 2 N 2 2nd exp AZ AX 2110Pblanket exposure. Dose is indicated in the x-axis. 71nm 59nm 47nm CD (nm) of line from 1st exp after 2nd blanket exp Dose (mj) for 2nd blanket exposure VRC Chemical Shrink 51nm Impact of 2nd exposure on frozen image understood page 16

17 Impact of worn out substrate on 2 nd exposure 1.) 1C5D 2.) 1 st exp 60mJ 1.) 1C5D 2.) VRC 3.) 1 st exp 60mJ 1.) 1C5D 2.) 1 st exp L/S100mJ 3.) VRC 4.) 2 nd exp 60mJ 1.) 1C5D 2.) 1 st exp L/S120mJ 3.) VRC 4.) 2 nd exp 64mJ 1.) 1C5D 1.) 1C5D 2.) 1 st exp Flood 100mJ 2.) 1 st exp Flood 100mJ 2.) VRC 2.) Solvent rinse 3.) 2 nd exp 45mJ 3.) 2 nd exp 45mJ Footing/scumming Footing/scumming Further evidence its not BARC interacting with freeze vapors Actual dose shining at the BARC interface w/ open frame flood exp is enormously higher Absorbance Impact of VRC on BARC interface 8min-0 2min-0 0 2min 8min Residual resist at the BARC interface can possibly interacts with the freeze vapors and creates the scumming in the 2 nd exposure Wavenumber (cm-1) page 17

18 Profile variations in dual imaging, Chemical considerations Chemically amplified resist are prone to BARC interfacial interactions which may be worse during the 2 nd exposure since its has been worn by the 1 st exp Possibly from; first photo may depleted BARC additives which served to assisting in the 1 st photo image BARC surface altered from the strong acid/peb processing in the first imaging step. Residual bonding of the 1st photo resist to the substrate BARC Frozen image 1 st image 2 nd image Bulk diffusion of resist additives into the substrate and frozen image diminish acid concentrations near the interfaces and leads to profile variations 2 nd resist Areas w/ acid in 2 nd resist Foots are created at the bottom corner of frozen resist where PAG/acid loses are the highest since there are 2 diffusion directions which contribute. Insufficient acid in interfacial regions does not allow for the proper level of deprotection of the resin which in turn creates the foot/scumming. Lose of PAG from 2 nd resist during SB Exposure Lose of Acid from 2 nd resist during PEB page 18

19 Summary of profile variations in dual imaging ptical considerations during 2nd exposure Refraction from non-plannar air/photoresist interface Very subtle optical differences at the resist/resist interface, bulk materials look similar. (VASE data suggests this is not important for VRC) Geometric considerations during 2nd imaging step Tops of 2 nd images are not rectangular especially when equal resist FT are used Confinement of acid diffusion is different in 2 nd exposure (area is defined by a non-planner top and inclusion of a non-imaging medium) Chemical considerations Frozen images acts as a PAG/acid sink to the 2 nd resist coating which leads to; encasing of the frozen pattern by the 2 nd resist resin. and dose change in second exposure profiles variation in the 2 nd image since acid depletion may not be uniform. Barc/ interface conditions is not the same after first exposure. page 19

20 Reverse tone trilayer Freeze w/ VRC Spin on glass in organic solvent Positive resist Thick organic UL Back etch (CF4/2) SG UL Resist plug removal and transfer to UL (2) reverse tone transfer of photoresist image into UL! page 20

21 Conclusions AZ has developed a track-compatible resist freeze process that exhibits high sensitivity and freezes in under 1 minute. Diamines appear to be the best freeze liquids: they predominately crosslink through the lactone functionality of the photoresist (found in practically all 193nm photoresists). Regenerative post VRC processing did so far not show any improvements. Some profile improvements were observed on SiN substrate. VRC freezing leads to bulk crosslinking which is superior to surface crosslinking making the image more resilient to SEM exposure. VRC has more benefits than just double imaging, one example is the inverse trilayer process. * n n * R H NH HN H R n * * n page 21