Rapid formation and real-time observation of micron-sized conjugated nanofibers with tunable lengths and widths in 20 minutes by living crystallization-driven self-assembly†

Preparing well-defined semiconducting nanostructures from conjugated polymers is of paramount interest for organic optoelectronic devices. Several studies have demonstrated excellent structural and size control from block copolymers (BCPs) containing non-conjugated blocks via crystallization-driven self-assembly (CDSA); however, the precise control of their size and shape remains a challenge due to their poor solubility, causing rapid and uncontrolled aggregation. This study presents a new type of fully conjugated BCP comprising two polyacetylene derivatives termed poly(cyclopentenylene-vinylene) to prepare semiconducting 1D nanofibers. Interestingly, the widths of nanofibers were tuned from 12 to 32 nm based on the contour lengths of their crystalline core blocks. Their lengths could also be controlled from 48 nm to 4.7 μm using the living CDSA. Monitoring of the growth kinetics of the living CDSA revealed the formation of micron-sized 1D nanofibers in less than 20 min. The rapid CDSA enabled us to watch real-time growth using confocal fluorescence microscopy.

All images were obtained on tapping mode using non-contact mode tips from Nanoworld (Pointprobe ® tip, NCHR type) with a spring constant of 42 N m -1 and a tip radius of ≤ 8 nm.
For each sample, length, area, aspect ratio, and angle distributions of nanoparticles were calculated by measuring over 50 samples of randomly picked nanoparticles using Gatan Digital Micrograph software (TEM imaging). Values of the number-average (Xn), weight-average (Xw), and standard deviation () of nanosheets were calculated as follows where N is the sample size.

S4
B. Polymerization procedure A 4 mL sized screw-cap vial with septum was flame dried and charged with a monomer and a magnetic bar. The vial was purged with Ar(g) four times, and degassed anhydrous THF was added ([M1]0 = 0.5 M, or [M3]0 = 0.5 M). After the Arpurged Grubbs third-generation catalyst in the other 4 mL vial was dissolved in THF, the solution was rapidly injected to the monomer solution at 0 °C under vigorous stirring. After the complete conversion of M1 to P1, or M3 to P3, the second monomer (M2) was added ([M2]0 = 0.1 M) to the vial at 0 °C.

1) Quenching and purification:
The reaction was quenched by excess ethyl vinyl ether (EVE) after the desired reaction time and precipitated in methanol at room temperature. The obtained purple solid was filtered and dried in vacuo. The conversion of monomer was calculated from the 1 H NMR spectra of the crude mixture. 1-3 2) NMR analysis of crude mixture for calculating the conversion of monomers (Table 1): After quenched by excess amounts of EVE, one 10 L aliquots of the crude mixture were dried in vacuo and diluted in CDCl3 for calculating conversion (the remaining reaction solution was precipitated in MeOH to obtain the isolated polymer powders). The monomer conversion was determined by using the reference peak in 1 H NMR of the crude solution (before removing unreacted monomers), and the polymer peak overlapping with the same of monomer was taken as a theoretical value, and the other remaining monomer peak which was not overlapped with its polymer peak was calculated.

3) INCP experiments (Fig. S7 and S20):
In the case of in situ TEM sampling during the polymerization process, the 20 L aliquots (THF solution) were taken out from the solution at different times using microsyringes, and diluted with THF or chloroform to 1 g/L after quenching by EVE. DLS, UV-vis analyses and TEM samplings were conducted with the in situ 1 g/L solutions. All experiments from The solutions of the purified BCPs in chloroform or DCM (1 g/L, 1 mL in 4 mL vial) was sealed with Teflon lined cap. In some cases, they were aged in the fume hood at 25 °C for a day to form 1D nanofibers.
■ Seeded growth experiment The unimer solutions, BCPs in chloroform (10 g/L) in general, were added to the seed solution (prepared by sonication) with various unimer-to-seed ratios (in chloroform or DCM, 0.1 g/L).
■ Self-seeding experiment The seed solutions of BCPs in chloroform or DCM (0.1 g/L, 1 mL in 4 mL vial) was sealed with Teflon lined cap and was heated in vial heating block at various temperatures (40 °C -61 °C) for 30 min. Then, the heated BCPs were aged in the fume hood at 25 °C.      at 20 °C and 50 °C at 0.1 g/L in chloroform. At each temperature, the solutions were heated at 50 °C for 30 min, and the Dh of the P150-b-P233 dissapeared and that of P150-b-P244 decreased as less than 100 nm after heating; however, even after heating, the longer BCPs still showed large Dh values. (b) 1 H NMR spectra of P150-b-P2n (n = 33-66) before and after heating up to 47 °C in chloroform-d for 30 min to compare the relative percentage of P2% at different temperatures. For the NMR analysis, 1 g/L deuterated solutions were prepared and analyzed with 2 sec relaxation time and 128 scan numbers. The relative P2% (from Fig. S2) was close to 100% in P150-b-P233 and P150-b-P244, and longer BCPs showed only 27.8% and 18.4% integration then expected values. Fortunately, more quantitative analysis was possible for P150-b-P233 and P150-b-P244 due to better solubility of P2 at the higher temperature. For the analysis, the BCPs were prepared after purification.    Fig. S11 (a) UV-vis absorbance spectra of P150 and P150-b-P2n (after purification) after 1 day aging in 0.05 g/L chloroform, and (b) their optical bandgaps (Eg) from the spectra in (a). All Eg values were in the semiconductor range from 0.25 to 2.5 eV. (We aged 1 g/L chloroform solutions for 1 day then diluted from 1 g/L to 0.05 g/L to analyze by UV-vis absorbance spectroscopy.)          Fig. 3f and S18c. Error bars in (d) indicate the standard deviation ().  Table S1.  Fig. S18c). The contour lengths of P2n from the calculation and average widths were well matched as we expected.    (d) TEM images and contour length histograms of 1D nanofibers prepared by CDSA process using unimerto-seed ratio (U/S ratio) from 1 to 2, 3, 5, and 10 with seed micelles from Fig. S26 at 25 °C. The number in parentheses is "the average Ln and its length dispersity." Fig. S28 (a) A plot of Ln values vs. time for assembly monitored for 1 day with U/S ratio = 5) and (b) TEM images in the growth kinetic studies. In the initial chloroform solution, a P150-b-P222 followed the typical seeded growth where only seed-to-unimer assembly occurred. In 100 min-200 min time range, the growth seemed to be stagnated, as if seeded growth was complete; however, the length increased again, indicating that end-to-end coupling (as the second growth) occurred. Presumably, the dynamic exchange from the end of the 1D nanofibers to unimers promotes the end-to-end coupling process. (c) TEM images of very long 1D nanofibers without added unimers and (d) with U/S ratio = 1 to seedmicelles. (e) TEM images of the 1D nanofibers by seeded growth using U/S ratio 5 at -20 °C and 40 °C. Lowering the aging temperature (temp.) further promoted all assemblies (view of thermodynamic perspective), resulting in longer 1D nanofibers after 1 hour (a living CDSA failed). Contrary to this, at higher temp (40 °C ), unimers cannot assemble to the seed due to higher kinetic energy. (f) Schematic illustration of the growing process of 1D nanofibers from P150-b-P222 in chloroform.   In detail, we first failed to get the narrow Lw/Ln when the unimer was added at once (U/S ratio > 5) because the growth process was quite fast due to the high crystallinity. To solve this limitation, we injected the unimers in two portions with 30 min interval and succeeded in living seeded growth up to Ln = 2.2 m with Lw/Ln = 1.11 (5 + 5 eq). Dh values of the1D nanofibers gradually increased, and Lns of them linearly increased according to the U/S ratios. For the longer Ln than the theoretically predicted one, we speculate that when adding a heated unimer solution to seed solutions, an increase in temperature and a decrease in concentration can cause the partial dissolution of the seeds and formation of longer 1D nanofibers. In detail, we first failed to get the narrow Lw/Ln when the unimer was added at once (U/S ratio > 5) because the growth process was quite fast due to the high crystallinity. To solve this limitation, we injected the unimers in two portions with 30 min interval and succeeded in living seeded growth up to Ln = 4.7 m with Lw/Ln = 1.04 (5 + 5 eq). Dh values of the1D nanofibers gradually increased, and Lns of them linearly increased according to the U/S ratios. For the longer Ln than the theoretically predicted one, we speculate that when adding a heated unimer solution to seed solutions, an increase in temperature and a decrease in concentration can cause the partial dissolution of the seeds and formation of longer 1D nanofibers.    Fig. 2g and 2h. Tables of (f) average width (Wn) before and after staining and (g) average height (Hn) of the 1D nanofibers.

S28
The Wn values were well-matched with the trend of widths of P150-b-P2n correlating with DP of P2. Fig. S37 The CDSA of P1100-b-P255 in chloroform via the self-seeding method. (a) DLS profiles, (b) UV-vis absorbance spectra, (c) TEM images, and contour length histograms of 1D nanofibers from P1100-b-P255 prepared by CDSA process from seed-micelles (Ln = 68.7 nm, Lw/Ln = 1.18, sonication on time 30 secs) in 0.1 g/L chloroform. We applied the selfseeding method to control the length of their 1D nanofibers because we failed to prepare the unimer solutions of P150b-P2n and P1100-b-P2n due to their high crystallinity. Also, as Fig. S36a, b showed, the low-height part of 1D nanofibers of P1100-b-P2n formed low-height seeds after sonication. This height deviation can be solved by self-seeding method, where seeds with low-height were preferentially melted due to low-crystallinity.       First, the seed micelles (B block, with Ln of 169 nm (Lw/Ln = 1.10)) were prepared by seeded growth from the initial seeds using U/S ratio of 5 at 10 °C, then the second unimer with U/S ratio of 5 was added. (An optimal condition for B block changed to 0.02 g/L due to reducing the selfnucleation of the second unimer (P325-b-P222)). (d) Contour length histograms of the B block of BCM1, and the BCM1.

Equation S1
These two functions were used to interpret the kinetic data for the seeded growth of P150-b-P2n. 7

Fig. S46
Fittings of the kinetic data were performed using Origin (OriginLab, Northampton, MA) software. Size of seed was held at the value from TEM images for all data sets. During a CDSA process of P150-b-P233, the micelle length at each time point was used to weight the fits (with an instrumental error). The kinetic values were obtained when the fits from Equation S1 converged. We applied two different equations to the kinetic data (of U/S ratio = 5 here); (a) The first-order kinetic function and (b) the stretched exponential function in Equation S1. The stretched exponential function gave higher R 2 values. An average 'b' value, which related to the "self-assembly of polymer chains" was obtained as 0.54. (c) Example data sets for the two fitting methods. Note: k' in the stretched exponential function model was reported in sec -1 . (d) The raw data of these fitting results which were reported in Fig. 8b. 7

Table S2
Data summary of kinetic studies on seeded growth of 1D nanofibers from P150-b-P233 in chloroform over 13 hours with various U/S ratios 2, 3, 5, and 10.

Descriptions for the videos
Video S1: This movie described the real-time growth of 1D nanofibers from P150-b-P233. The optimized conditions for observing real-time elongation using a laser scanning confocal microscope (LSCM) were slightly different from those of living CDSA by using TEM imaging. Since it is difficult to focus on free-floating seeds, 0.01 g/L seed solution was scattered on the slide glass to obtained attached one to the surface. After adding unimer solution with U/S ratio of 30, slide glass was sealed with a cover-glass in short time, and a real-time movie could be obtained without solvent drying. Even considering the 120 nm lateral resolution of the LSCM, after 2 minutes, the length of the attached seeds became longer, with a controlled growth rate. However, since real-time monitoring only reflects the growth on the surface, it was difficult to directly compare the results of real-time monitoring by LSCM with growth kinetics studies through TEM imaging.
Video S2: A gray-colored converted version of the Video S1 showing more clear growth than Video S2. Video S3: A movie of the dynamic movements of 1D nanofibers from P150-b-P222 in 0.05 g/L chloroform. * The original video rate was 10 fps (10 frames per second). The number at the top left of each video was the number of frames, which allowed us to check the actual time.