Supporting Information

Transcription

1 Supporting Information Wiley-VCH Weinheim, Germany

2 Application of Solvent-Directed Assembly of Block Copolymers to the Synthesis of Nanostructured Low Dielectric Constant Materials Thomas M. Hermans, Jeongsoo Choi, Bas G.G. Lohmeijer, Geraud Dubois, Russell C. Pratt, Ho-Cheol Kim, Robert M. Waymouth, James L. Hedrick * [ ] These authors contributed equally to this work [*] T. M. Hermans, Dr. J. Choi, Dr. B. G. G. Lohmeijer, Dr. G. Dubois, Dr. R. C. Pratt, Dr. H.-C. Kim, Dr. J. L Hedrick IBM Almaden Research Center, 650 Harry Road, San Jose, CA (USA) Fax: (+1) hedrick@almaden.ibm.com Prof. R. M. Waymouth, Department of Chemistry, Stanford University, Stanford, CA (USA) 1

3 Experimental Section Materials Reagents were available commercially and used as received unless otherwise noted. Solvents were dried using activated alumina columns from Innovative Systems. rac-lactide (LA) was received from Purac and recrystallized from dry toluene 3 times prior to use. N,N-dimethylacrylamide (DMA) and (-)-sparteine were purified by stirring overnight over CaH 2 and subsequent distillation under reduced pressure. The hydroxy-functionalized alkoxyamine, 2,2,5-trimethyl-3-(4 -p-hydroxymethylphenylethoxy)-4-phenyl-3-azahexane and the thiourea catalyst were prepared according to literature procedures. [1] PDMA macroinitiators synthesized by nitroxide-mediated polymerization (NMP) were dried in a vacuum oven and further dried by co-evaporation with dry, distilled toluene 3 times before transferring to a glovebox for assembly of the ROP reaction. Synthesis of PDMA macroinitiators 2,2,5-Trimethyl-3-(4 -p-hydroxymethylphenylethoxy)-4-phenyl-3-azahexane (0.20 g, 0.6 mmol), TEMPO (4.2 mg, 0.03 mmol) and N,N-dimethylacrylamide (2.80 g, 28 mmol) were stirred at 125 C for 14 hours and subsequently cooled to room temperature. The reaction mixture was diluted with CH 2 Cl 2 and precipitated in diethyl ether (2 ) to give 2.45 g (81%) of hydroxyl-functionalized poly(n,ndimethylacrylamide) 2. 1 H-NMR (CDCl 3 ): d = (m, 9 H; H aromatic ), 4.81 (bs, 1 H; HC-ON), 4.62 (bs, 2 H, PhOCH 2 ),), (m, 380 H; N(CH 3 ) 2 PDMAA backbone, CH 2 PDMAA backbone, CH PDMAA backbone, ON-CH, CH 3 initiating fragment, CH 3 CHCH 3, C(CH 3 ) 3 ), CH 3 CHCH 3, CH 3 CHCH 3 ). GPC (RI): M n (PDI): 2800 g mol -1 (1.08); 4100 g mol -1 (1.07); 7000 g mol -1 (1.09); g mol -1 (1.16). Synthesis of PDMA-PLA block copolymers A solution of the hydroxyl-functionalized poly(n,n-dimethylacrylamide) (PDMA), thioureacatalyst (TU), (-)-sparteine (SP) and rac-lactide (LA) were stirred in dry methylene chloride for 3 hours. For 1 this required 540 mg (193 µmol) PDMA 25, 178 mg (481 µmol) TU, 53.6 mg (240 µmol) SP, 1350 mg (9.64 mmol) LA in 10 ml of CH 2 Cl 2 ; for 2 this required 500 mg (125 µmol) PDMA 40, 416 mg (1.13 mmol) TU, 53.6 mg (552 µmol) SP, 1620 mg (11.25 mmol) LA in 16 ml of CH 2 Cl 2 ; for 3 this required 250 mg (36 µmol) PDMA 70, 198 mg (536 µmol) TU, 60 mg (268 µmol) SP, 770 mg (5.36 mmol) LA in 7.5 ml of CH 2 Cl 2 ; for 4 this required 250 mg (13.3 µmol) PDMA 175, 172 mg (465 µmol) TU, 50 mg (225 µmol) SP, 670 mg (4.65 mmol) LA in 5 ml of CH 2 Cl 2. The resulting viscous solution was diluted with CH 2 Cl 2 and precipitated in pentane (2 ). The white precipitate was collected by filtration. Isolated yields were > 90% with full conversion of LA. 1 H-NMR (CDCl 3 ): d = (m, 9 H; H aromatic ), (m, 86 H; CH PLA, HC-ON, PhOCH 2 ), (m, 2 H; CHOH, CHOH) (m, 512 H; N(CH 3 ) 2 2

4 PDMAA, CH 2 PDMAA, CH PDMAA, CH 3 PLA, ON-CH, CH 3 initiating fragment, CH 3 CHCH 3, C(CH 3 ) 3 ), CH 3 CHCH 3, CH 3 CHCH 3 ). The block ratio was calculated accordingly (see Table S1 for further details). Table S1. PLA-PDMA block copolymers Entry Macroinitiator M n / g mol -1 (PDI) Block copolymer M n / g mol -1 (PDI) 1 PDMA 25 -OH 2800 (1.08) PDMA 25 -PLA (1.05) 2 PDMA 40 -OH 4100 (1.07) PDMA 40 -PLA (1.08) 3 PDMA 70 -OH 7000 (1.09) PDMA 70 -PLA (1.11) 4 PDMA 175 -OH (1.16) PDMA 175 -PLA (1.12) Figure S1: GPC trace of PLA-PDMA (Table S1, entry 3) after polymerization of PDMA block (black) and after subsequent polymerization of PLA block (red). Formulation and casting of MSSQ/block copolymer blends Thin films of MSSQ/block copolymers were prepared by mixing 0.3 g of block copolymer solution (10 wt. % PGPE or nbuac) with 0.2 g of MSSQ solution (10 wt. % PGPE or nbuac) and 10 mgs of triethylamine. Thin films ( nm) were obtained by spin casting the above solution on clean silicon wafers with 3000 rpm. The films were first maintained under nitrogen for 3 days at 50 o C, and then heated from 50 o C to 450 o C at 5 o C/min and held at 450 o C to 2 hours. 3

5 Measurements 1 H-NMR spectra were obtained on a Bruker Avance 400 instrument at 400 MHz. Gel permeation chromatography was performed in THF using a Waters chromatograph equipped with four 5 µm Waters columns (300 mm x 7.7 mm) connected in series with increasing pore size (10, 100, 1000, 10 5, 10 6 Å), a Waters 410 differential refractometer and a 996 photodiode array detector, and calibrated with polystyrene standards (750-2 x 10 6 g mol -1 ). Differential scanning calorimetry (DSC) was performed using a TA Differential Scanning Calorimeter 1000 that was calibrated using high purity indium at a heating rate of 10 C/min. Melting points were determined from the second scan following slow cooling (to remove the influence of thermal history) at a heating rate of 10 C/min. The micelle formation in a selective solvent was studied by the 1 H NMR core signal quenching experiment using deuterated methanol (methanol-d 4 ) as a model selective solvent. Figure S2a are 1 H- NMR spectra of PLA-PDMA (Table S1, entry 3) in methanol-d 4 at different temperatures, indicating that the signals associated with the PLA block are suppressed up to 60 o C. This is well represented in Figure S2b where the temperature dependence of 1 H-NMR peaks corresponding to each PDMA or PLA block of PLA-PDMA (3f) in methanol-d 4 Films thicknesses and refractive indices were measured using a filmetrics F20 Thin-Film Measurement System. Dielectric constants were determined using a capacitance bridge with an HP 4192A impedance analyzer using a metal insulator semi-conductor (M1S) structure. Specular x-ray reflectivity measurements were performed using a diffractometer (X Pert Pro MRD, Pananlytical) with ceramic x-ray tube (wavelength=0.154nm) and high resolution horizontal Goniometer (reproducibility +/ degrees). The x-ray beam was line focused by a multi-layer crystal and a four bounced Ge (220) crystal was used as a monochromator. The reflected beam was received into a proportional gas filled detector after bouncing Prefix triple axis optics module. The critical angles from the reflectivity data were obtained from the peak position of Iq 4 vs. q plots (q =(4p/λ)sinθ, where λ is wavelength, θ is the grazing incident angle of the X-ray beam). The small angle X-ray scattering (SAXS) experiments were performed at the LB-4-2 beamline at the Stanford Synchrotron Radiation Laboratory (SSRL). X-rays having energies of 8.98 kev, (wavelength of 1.381Å) were used with a gas chamber detector. The samples were positioned with the substrate normal coinciding with the incident beam. In order to reduce attenuation from the silicon, ~ mm films were processed on double-sided polished wafers having thicknesses of ~80 mm. Tapping mode atomic force microscopy (AFM, Dimension 3100, Digital Instruments) was used for surface characterization. Standard silicon cantilevers with resonance frequency about 300 khz (Veeco OTESPA) were used. SEM images were taken from the uncoated samples on a Hitachi S-4700 with accelerating voltages of 1 kv. The cross-sectional SEM images were taken from the samples containing fresh edges with tilting them in electron microscope. 4

6 Figure S2a: 1 H-NMR spectra of PLA-PDMA (Table S1, entry 3) in methanol-d 4 at different temperatures (indicated in C on the left side of each spectrum). -CH- of PLA PDMA -CH3 of PLA Arbitrairy peak area 1H-NMR [-] Temperature [ C] Figure S2b: Temperature dependence of 1 H-NMR spectra of PLA-PDMA (3f) in methanol-d 4 5

7 SAXS Analysis SAXS intensity I(q), can be expressed as ( ) ( ) I q = c n rfqr ( ) Sqr ( ) dr (1) 0 where c is a constant, n(r) is the pore size distribution function, F(qr) is the spherical form factor, and S(qr) is the structure factor for the monodisperse hard sphere model. [2-5] We have considered different types of size distribution functions and found that the SAXS profiles are best modeled using the Schultz distribution function: ( ) n r Z + 1 Z + 1 Z Z + 1 r exp r r0 r0 = Γ + ( Z 1) (2) where r is the radius, r 0 is the radius corresponding to the maximum of the distribution, Z is the width parameter, and Γ is the gamma function. As shown in Figure 2, the scattering intensities of both films are reasonably fitted with the hardsphere model and the Schultz size distribution function. However, compared to well fitted result for thin film cast from PGPE, the modeled fitting to the experimental profiles of thin film cast from BuOAc shows big deviation, which is often attributed by less defined spherical shape of scatterer. The sizes and size distributions of the spherical pores in thin films cast from PGPE solution of MSSQ/copolymer 1-4 are determined from the modeled fitting of SAXS profiles and shown in Figure 3. The structure factor, S(qr), in Equation (1) can be extracted from SAXS data by dividing the total scattering intensity by n(r)f(qr) which is estimated from model fitting. This procedure is valid with an assumption that particle size and shape, i.e., F(qr), are not affected by condensing in a confined space. Figure S3 shows the structure factor S(qr) extracted from SAXS data by dividing out the form factor with size distribution, n(r)f(qr), for two thin films. As shown in Figure S3, a series of higher order peaks for thin film cast from PGPE are observed, while they are smeared out in the S(qr) of thin film cast from BuOAc. It is noted that the positions of peaks in Figure S3(a) can be mapped with S(qr) of packed spheres in the manner of for example simple cubic (SC, 1: 2: 3:2: 5: 6: 8:3) or body-centered cubic (BCC, 1: 2: 3:2: 5: 6: 7: 8:3), which suggests relatively high degree of order in thin film cast from PGPE.. 6

8 (a) 2.0 S(qr) (cm -1 ) o q (A -1 ) (b) S(qr) (cm -1 ) o q (A -1 ) Figure S3. Structure factors, S(qr), extracted by dividing I(q) by n(r)f(qr) for thin films cast from (a) PGPE and (b) BuOAc. Arrows corresponds to the typical peak ratios (q/q *, where q * is the first order peak position) for sphere packing (1: 2: 3:2: 5: 6). 7

9 References [1] Bosman, A. W.; Vestberg, R.; Heumann, A.; Fréchet, J. M. J.; Hawker, C. J. J. Am. Chem. Soc. 2003, 125, 715. [2] J. S. Pedersen J. Appl. Cryst. 1994, 27, 595. [3] D. J. Kinning, E. L. Thomas Macromolecules 1984, 17, [4] E. Huang, M. F. Toney, W. Volksen, D. Mecerreyes, P. Brock, H. C. Kim, C. J. Hawker, J. L. Hedrick,; V. Y. Lee, T. Magbitang, R. D. Miller, L. B. Lurio Applied Physics Letters 2002, 81, [5] B. Lee, Y.-H. Park, Y.-T. Hwang, W. Oh, J. Yoon, M. Ree Nat. Mater. 2005, 4,