Direct Measurement of the Local Glass Transition in Self-Assembled Copolymers with Nanometer Resolution.

Nanoscale compositional heterogeneity in block copolymers can impart synergistic property combinations, such as stiffness and toughness. However, until now, there has been no experimental method to locally probe the dynamics at a specific location within these structured materials. Here, this was achieved by incorporating pyrene-bearing monomers at specific locations along the polymer chain, allowing the labeled monomers' local environment to be interrogated via fluorescence. In lamellar-forming poly(butyl methacrylate-b-methyl methacrylate) diblock copolymers, a strong gradient in glass transition temperature, Tg, of the higher-Tg block, 42 K over 4 nm, was mapped with nanometer resolution. These measurements also revealed a strongly asymmetric influence of the domain interface on Tg, with a much smaller dynamic gradient being observed for the lower-Tg block.

Introduction

Block copolymers, which self-assemble into nanodomain structures due to the incompatibility of chemically dissimilar monomer segments, have generated intense scientific interest and are used in a myriad of important technologies.[1] In such systems, a large fraction of the polymer segments can lie within a few nanometers of an internal interface, within a region where the dynamics and mechanical properties can be strongly modified from their bulk values.[2−5] The molecular dynamics at these soft internal interfaces can modulate the performance characteristics of copolymers and enable them to escape traditional material property trade-offs, such as those between stiffness and toughness,[6] that set limits on the performance of homogeneous polymers. Block copolymers, therefore, offer material solutions to address pressing societal challenges, including the upgrading of mixed plastic waste to a tough material,[7] thermoplastic elastomers with self-healing capability,[8] and nanostructured polymer solar cells[9,10] for more efficient harvesting of solar energy.

The ability to characterize the dynamics near the internal interfaces within block copolymers could enable the rational design of polymers with prescribed interfacial properties for next-generation applications. In addition, these same insights would advance our fundamental understanding of the complex ways in which interfaces and nanoscale confinement can influence the dynamics of polymers in technologically important macroscopic materials, i.e., block copolymers. At macroscopic or bulk length scales, the dynamic response is a composite of contributions from the dissimilar domains, their interdependence, and the presence of the internal interfaces. Decoupling these different contributions requires the ability to independently measure the dynamics of each block over length scales ranging from that of a few segments to that of the confining length scale or domain period where interfacial effects would be observed.

However, despite decades of both experimental and theoretical progress in understanding the thermodynamic nature of the interface within block copolymers, for example, the composition profile,[11,12] the field still lacks a predictive understanding of how dynamics are perturbed at the interface. An enduring barrier impeding a complete mapping of dynamics at and near interfaces within block copolymers is an inability to directly probe, at the nanoscale and with high resolution, interfacial phenomena. The challenge is especially acute because the length scale of block copolymer self-assembly is of O(10 nm). Overcoming this challenge holds promise for a conceptual leap in our understanding of how dynamics are altered at the interface, and beyond, in block copolymers.

Here, direct characterization of interfacial dynamics, as quantified by the glass transition temperature (Tg), across the domain period of lamella-forming diblock copolymers of poly(butyl methacrylate-b-methyl methacrylate), PBMA–PMMA, is presented. The direct and location-specific measurement of Tg in diblock copolymers is enabled by the precise placement of a fluorescent pyrene-containing monomer along the chain, at defined positions along either the PBMA or PMMA block, via anionic polymerization. This permits control of the spatial position of the fluorescent label when the copolymer self-assembles into a nanostructured material. Sparse labeling of the copolymer (<0.5 mol % at any position along the chain), combined with the high sensitivity of fluorescence to sense the glass transition, enables a nanometer-accuracy spatial resolution of Tg at prescribed distances away from the domain interface.

A gradient in PMMA-block Tg of 42 K is found over a length scale of ∼4 nm, thus revealing the extreme case of dynamic heterogeneity in copolymers. The gradient can be understood by considering the local composition experienced by each block within the domain structure, crucially combined with nanoscale confinement effects on Tg. The role of the relative softness of the confining and confined blocks is demonstrated by a comparison of the observed Tg to values calculated based the Lodge–McLeish (LM) model of self-concentration.

Results and Discussion

The ability of fluorescence to sense the Tg in the present polymers is corroborated by performing two sets of validation experiments. In the first set of experiments, Tg of PMMA homopolymers is measured as a function of the number-average molecular weight (M) via both fluorescence and differential scanning calorimetry (DSC). Both measurement techniques should reflect the strong M-dependence of Tg below a critical value.[13] To measure Tg by fluorescence, the temperature dependence of the fluorescence intensity of PMMA labeled with pyrene randomly along the chain (PMMA-py) was monitored (see Figure S3b). The inset of Figure a shows the fluorescence spectra for PMMA-py (M = 139 kg/mol), excited at 347 ± 0.5 nm at temperatures of 433 (solid line) and 333 K (dashed line), above and below the bulk Tg, respectively. A decrease in temperature yields an increase in fluorescence intensity due to reduced thermal energy and densification of the surrounding nanoscale medium.[14] Both effects reduce the rate of nonradiative decay of the excited-state pyrene fluorophore, a pathway competing with fluorescence during relaxation to the electronic ground state. We exploited the strong sensitivity of pyrene fluorescence intensity to the nanoscale medium to measure Tg. Figure a plots the integrated fluorescence intensity normalized to that at 433 K, as a function of temperature for two PMMA-py polymers with different values of M. The intersection of linear fits to the data at high and low temperatures provides an accurate measure of Tg.[15−17]

Figure 1

(a) Temperature dependence of the integrated fluorescence emission intensity of pyrene-labeled PMMA homopolymers (PMMA-py) of two different number-average molecular weights (M) and the corresponding linear fits in the glassy and rubbery regions. Inset, fluorescence emission spectra of PMMA-py (M = 139 kg/mol) at temperatures above and below Tg. The labeled monomer structure is also shown. (b) The molecular weight dependence of Tg for PMMA homopolymers, as determined by fluorescence or differential scanning calorimetry (DSC). Inset, gel permeation chromatography (GPC) traces of the series of PMMA-py homopolymers. (c) Temperature dependence of the fluorescence intensity of homogeneous PBMA–PMMA-py (M = 16 ± 1 kg/mol) diblocks with different compositions, where pyrene labels are attached randomly along the PMMA (blue) block. (d) Tg as a function of PMMA volume fraction in homogeneous PBMA–PMMA-py diblock copolymers. Inset, schematic of self-concentration of PMMA monomer units (blue) in a volume (gray) defined by the length of a Kuhn segment.

To confirm this assertion, the M-dependence of Tg for PMMA-py, as measured by fluorescence, was compared with the onset Tg as measured by DSC in Figure b. Over an M range of 8–139 kg/mol, the fluorescence-determined Tg is on average 4 K lower than the DSC onset Tg. At the lowest value of M, the trend is reversed, with a 12 K difference between the two techniques. Despite this difference, both techniques display the strong M-dependence of Tg expected for PMMA.[18] The sensitivity of fluorescence to the Tg of a pyrene-labeled PBMA homopolymer (PBMA-py) was also confirmed (see Figure S3a). The consistency of the trends combined with prior reports[19−21] confirmed the ability of the fluorescence method to sense Tg in bulk homopolymers.

In a second set of experiments, the dynamics of a single block in the diblock copolymer as perceived by fluorescence were characterized by selectively labeling that block. The Tg in a set of homogeneous diblock copolymers of PBMA–PMMA, wherein the PBMA and PMMA blocks are intimately mixed, with varying PMMA volume fraction (ϕPMMA), was measured by both fluorescence and DSC. The segregation strength (χN) for all of the homogeneous diblock copolymers, relative to χN for a hypothetical symmetric copolymer (ϕPMMA = 0.5) at the order–disorder transition ((χN)ODT), was estimated as χN/(χN)ODT = 0.7, where χ represents the Flory interaction parameter and N the total degree of polymerization (see the SI for the estimation of χN). The pyrene-labeled monomer was randomly (statistically uniformly) incorporated throughout the PMMA block of each diblock copolymer (PBMA–PMMA-py). Figure c plots the normalized integrated intensity vs temperature and the linear fits, which identify Tg for homogeneous PBMA–PMMA-py diblock copolymers with two different values of ϕPMMA. The composition dependence of Tg as determined by fluorescence (squares) or the DSC onset (triangles) is plotted for all samples investigated in Figure d. Both values systematically decrease as the content of PMMA within the copolymer is reduced. The Tg(ϕ) data obtained by DSC could be satisfactorily represented by the well-known Fox eq :[22]where Tg,PBMA = 296 K and Tg,PMMA = 388 K (at M = 22 kg/mol) are the DSC onset values measured on the homopolymers. In sharp contrast, Tg(ϕ) data obtained by fluorimetry could not be fit to eq with the Tg values determined by fluorimetry for the homopolymers of the two blocks. Instead, as illustrated in Figure d, the data could be well-fit to the LM model of self-concentration.[23] Self-concentration is a consequence of the chain connectivity of monomer units in a homogeneous polymer mixture. It may be expressed as the local volume fraction (ϕs) occupied by segments of the same chain in a polymer blend, or the same block, in a homogeneous block copolymer. The dynamics of the mixture are defined within a nanoscale volume, where chemically identical segments exhibit an effective concentration (ϕeff) greater than the bulk, eq :

The effective Tg of a component in the mixture is then determined by evaluating eq at ϕeff. The LM model provided an excellent fit with ϕs,PMMA = 0.38, where Tg for PBMA and PMMA were 289 and 385 K, as determined by fluorescence on the component homopolymers, see the SI for the estimation of ϕs. Therefore, fluorescence senses the effective Tg of a labeled block within a homogeneous block copolymer.

Figure a schematically shows the chain architecture for a series of 12 near-symmetric PBMA–PMMA diblock copolymers (M = 54 ± 9 kg/mol (±1 standard deviation of all 12, details in Table S1), dispersity (Đ) ≤ 1.1, and χN/(χN)ODT = 2.4), in which pyrene was attached at a specific location along either polymer block by employing sequential anionic polymerization. PBMA–PMMA diblock copolymers were prepared in which pyrene was located along the chain at positions varying from the block junction (J) to the chain end (E), or randomly within a particular block (uniformly labeled, U), as illustrated in Figure a. The spatial control of the label in the self-assembled copolymer is schematically shown in Figure b, where the label is placed at the end of the PMMA block. In all cases the local pyrene fraction within labeled sections of the copolymer was less than 0.5 mol %; the balance of the monomer units in the labeled section were either PMMA (left column in Figure a) or PBMA (right column in Figure a). These sets of selectively labeled copolymers allowed for the direct mapping of the gradient in dynamics in the self-assembled, nanostructured polymer.

Figure 2

Fluorescence characterization of self-assembled PBMA–PMMA diblock copolymers. (a) Schematic of the selective labeling (green) of a lamella-forming PBMA–PMMA (red-blue) diblock copolymer. (b) Schematic of a self-assembled lamellar PBMA–PMMA diblock copolymer, where the label is placed at the end (E) of the PMMA block. (c) Temperature dependence of the integrated fluorescence emission intensity and the corresponding linear fits in the glassy and rubbery regions of junction- and end-labeled PBMA–PMMA (χN/(χN)ODT = 2.4), where the pyrene labels are attached to the PMMA block. (d) Analogous to (c) with pyrene labels attached to the PBMA block.

Figure c plots representative normalized integrated fluorescence intensity vs temperature data for the PBMA–PMMA copolymer, in which the pyrene label was attached either on the PMMA side of the block junction (filled squares) or at the end of the PMMA block (open squares). There is a strong location dependence of Tg along the PMMA block: for the polymer labeled at the chain end, Tg,E = 364 K, while for the polymer labeled at the block junction, Tg,J = 322 K, representing a 42 K range in local Tg along the PMMA block. Conversely, Figure d shows representative fluorescence data in which pyrene was attached on the PBMA side of the block junction (filled circles) or at the end of the PBMA block (open circles). For the PBMA-labeled polymer, Tg,E = 295 K and Tg,J = 303 K, representing only an 8 K range in local Tg along the PBMA block. As a benchmark, recall that the PMMA and PBMA homopolymers have a 96 K difference in bulk Tg.

A powerful feature of fluorescence as an approach to measure Tg is that the reporter dye labels can be placed at prescribed locations along the chain, as schematically illustrated in Figure a. This provides a unique means to probe the full gradient of glass-to-rubber transition temperatures within the self-assembled diblock copolymer. However, the distribution of distances of the fluorescent label from the block interface must be evaluated. Here, self-consistent field theory (SCFT) calculations, correcting for fluctuations via renormalized one-loop theory, were used to accomplish this task (see the SI for details). The spatial distribution of monomer segments at selected positions along the chain was calculated using open source code created by Arora and co-workers.[24]Figure shows the calculated composition profiles for the pyrene-bearing segments across the domain period of a symmetric diblock copolymer with χN/(χN)ODT = 2.4, where d is the domain period and x is the distance along the lamellar normal, starting from the center of the PBMA-rich domain (x/d = 0, 1). A value of d = 27 nm was determined for a PBMA–PMMA diblock copolymer with M = 47 kg/mol by small-angle X-ray scattering (SAXS) (see Figure S6d). Vertical dashed lines in Figure represent the thickness of the interface (t = 3.8 nm) determined by SAXS.

Figure 3

Composition profile of labeled segments across the domain period (d) of a symmetric diblock copolymer where χN/(χN)ODT = 2.4, for the five different label positions schematized in Figure a. Profiles have been smeared with a displacement σ = 0.029 (see the SI for details). Dashed vertical lines demarcate the width of the interface as determined by SAXS. Profiles are normalized to equal area, with the highest value of labeled segment density (ϕA) set to unity.

The calculated composition profiles highlight that the location of the pyrene labels within the self-assembled nanodomain structure can be controlled and tuned. For instance, placement of the pyrene labels at the PMMA chain end resulted in a composition profile (red curve) whose maximum is in the center of the PMMA-rich domain. Conversely, placement of the pyrene labels at the block junction revealed a composition profile (blue curve) whose maximum is at the block junction, i.e., the center of the polymer–polymer interface. Collectively, the Tg determined for selectively labeled segments, the evaluation of the segment distribution, and the measurement of the domain period allowed us to map Tg across the domain structure of the self-assembled diblock copolymer with nanometer accuracy.

Figure plots the local Tg within the diblock copolymer, as a function of average distance from the nearest PBMA–PMMA interface, z (z/d = 0, 0.25), over one-half of the domain period. The PBMA–PMMA block junction is identified as z = 0 nm. The local Tg was measured for five label locations along each block, as highlighted in Figure a. As illustrated in Figure , the pyrene labels have some distribution in space (typically a few nm), and the measured Tg reflects an average over this distribution. For Figure , the position (z) of the label from the interface, corresponding to the measured Tg for each label location along the chain, was taken to be the concentration-average position[25] of the respective labeled segment composition profile (Figure , see Figure S10 for additional discussion). Remarkably, the 42 K variation in Tg noted above occurs across only a 4 nm distance. While classical measurements, such as DSC or broadband dielectric spectroscopy, have revealed heterogeneous dynamics within the nanodomain structure,[26−28] the present fluorescence measurements are the first to quantify the spatial variation and gradients governing said dynamics in block copolymers.

Figure 4

Experimental Tg as a function of average position across the domain period of a lamellar PBMA–PMMA diblock copolymer, where χN/(χN)ODT = 2.4: PBMA (red) or PMMA (blue) segments. The black symbols correspond to the calculated Tg of PBMA (left) or PMMA (right) segments, using the Fox equation and accounting for self-concentration effects (ϕs,PMMA = 0.38 and ϕs,PBMA = 0.5).

In the case of the rubbery-glassy PBMA–PMMA diblock copolymer, a highly asymmetric gradient in Tg about the interface was observed. Within the PBMA domain, Tg,E = 295 K at the chain end, i.e., z = −4 nm. This value was also consistent with the Tg measured by fluorescence for the uniformly labeled PBMA block, Tg,U = 296 K. Across the PBMA domain, an 8 K range in PBMA Tg was observed. In contrast, within the PMMA domain, Tg,E = 364 K at the chain end, i.e., z = +4 nm, which is also consistent with the Tg measured for the uniformly labeled PMMA block, Tg,U = 362 K. However, across the PMMA domain, a 42 K range in PMMA Tg was observed. Thus, the magnitude of the perturbation of the dynamics by the interface was much greater in the glassy domain than in the rubbery domain. This is qualitatively consistent with the fluorescence results of Baglay and Roth[29] on multilayer films, wherein a greater Tg perturbation was observed for a thin film sandwiched between rubbery layers than between glassy layers.

There are several key observations from the measurements presented in Figures and 4: (i) within each self-assembled nanodomain there exists a strongly location-dependent Tg, (ii) the ΔTg (Tg,E – Tg,J) has a larger magnitude in the PMMA domain in comparison to the PBMA domain, and (iii) there is a 19 K difference in Tg,J depending on whether the pyrene label is located within the PMMA or PBMA block, adjacent to the block junction. These observations combine to highlight a complex interplay between interfacial and self-concentration effects on the dynamics of nanostructured polymers, as discussed below.

To better understand these observations, a location-dependent Tg was calculated based on the LM model, with ϕs,PMMA = 0.38 and ϕs,PBMA = 0.5, using the local composition (ϕ) calculated by SCFT at each position along the domain period. If the dynamics were dependent only on the local composition, and if there were no additional influences of the domain interfaces, then the LM model would be expected to accurately capture the spatial variation of Tg across the domain period. The ϕeff and Tg profiles were then computed via eqs and 2 (see Figure S8). The effective Tg experienced by a labeled segment was determined by linearly weighting the effective Tg by the labeled segment distribution[25] (Figure ) and is also plotted in Figure (black triangles); black connecting lines represent a guide to the eye. The validity of linearly weighting the effective Tg by the labeled segment distribution was confirmed by summing the labeled segment distributions with weights that yielded a close match to the composition profile within the PMMA block for the uniformly labeled PMMA block, and using those weighting factors to calculate Tg,U based on the measured Tg values corresponding to each segment distribution (Figure , see Figure S9 and Table S3). The Tg,U determined from this calculation was 359 K, in good agreement with the experimental Tg,U = 362 K of the diblock copolymer with a uniformly labeled PMMA block.

Self-concentration can qualitatively account for the dissimilar Tg values measured for the two junction-labeled polymers, depending on whether the label was incorporated on the PBMA or PMMA side of the junction, since each component is locally enriched in its own type of segment. Moreover, the more compact coils formed by PBMA (higher ϕs) led to a smaller range of Tg experienced by the PBMA block than the PMMA block. Self-concentration also accounts for the spatial variation of the local Tg in the PBMA block, as noted by the good agreement between the experimental and calculated values. In contrast, for the PMMA block, the experimentally determined Tg was systematically lower than the value predicted by the LM model, even in the domain center. While the LM model considers self-concentration effects on Tg, it does not account for changes in Tg that could occur due to nanoscale confinement. We therefore attribute the local Tg profile within the PMMA domain to segmental mixing modulated by self-concentration, crucially combined with the presence of interfaces, which lead to nanoscale confinement and a reduction in Tg within the thin block copolymer domains.

It is well-known that in states of soft confinement, where interfacial effects become important, PMMA exhibits a suppression in Tg relative to the bulk.[30,31] In considering the effects observed here, the significantly lower Tg within the PMMA domain can be explained via soft confinement by the PBMA block which imparts additional mobility, originating at the domain interface, to the PMMA blocks. As such, the bulk value of Tg for PMMA is not recovered even in the center of the PMMA domain due to interface-induced gradients in Tg, which may propagate over distances of tens[14,32] or even hundreds[29,33] of nanometers, i.e., much greater than the domain size. The agreement between Tg as determined by fluorescence and the LM model in the PBMA-rich domain can similarly be explained within the context of nanoscale confinement. The glassy PMMA domain, which confines PBMA, acts as a solid substrate with no attractive interactions and represents the case of hard confinement. Under this circumstance, bulk values of Tg are expected and observed, in agreement with prior examples of polymers subject to hard confinement.[34−36]

The influence of interblock segregation strength on the local dynamics was assessed by a comparison of the local Tg of the PMMA block at weak (M = 26 ± 2 kg/mol, Đ ≤ 1.06, χN/(χN)ODT = 1.2) and intermediate (M = 54 ± 9 kg/mol, Đ ≤ 1.1, χN/(χN)ODT = 2.4) segregation strengths. Figure shows the local Tg for the PMMA block, as determined by fluorimetry and as calculated by the LM model, at weak and intermediate segregation strengths vs the average position of a labeled monomer segment. The pyrene-bearing monomer was placed at the same fractional distance, e.g., J + 50%, for both sets of diblock copolymers. Although the segment density profiles are significantly broader at weak segregation (Figure S11) vs intermediate segregation (Figure ), the local Tg as calculated by the LM model, averaged over the segment density distribution, is quite similar for the two segregation strengths at any value of z. At the block interface, in both weak and intermediate segregation, the PMMA blocks exhibit roughly a 30 K depression in Tg relative to the calculated value. In the center of the domain the difference in Tg between the experimental and calculated values, though smaller than that at the interface (∼10 K vs ∼30 K), persists for both segregation strengths.

Figure 5

Tg as a function of distance from the interface as measured by fluorimetry (squares) or calculated via the LM model (triangles) at weak (open symbols) and intermediate (closed symbols) segregation strengths.

Conclusions

This study has demonstrated the utility of fluorescence spectroscopy to characterize the glass transition in multicomponent polymers over different length scales where segmental mixing, self-concentration, and interfacial effects act to perturb Tg. In homogeneous diblock copolymers, we characterized the dynamics of one type of block and demonstrated the presence of self-concentration effects, which are active over the distance of a few monomer units. In nanostructured diblock copolymers, both interfacial and self-concentration effects act to perturb Tg yielding an asymmetric Tg variation across the interface. The location-specific nature of fluorescence spectroscopy to characterize Tg was highlighted, as gradients in nanostructured polymers over length scales less than 5 nm were characterized with nanometer spatial resolution. Insights gained from the nanometer-scale measurements of Tg will inform the design of nanostructured polymers for emerging applications where control of interfacial dynamics has been shown to enhance performance, e.g., block copolymer electrolytes for solid-state batteries.[37,38]

Methods

Fluorescent Label and Polymer Synthesis

The fluorescent label, 1-pyrenylbutyl methacrylate, was synthesized via the condensation of methacryloyl chloride (Sigma-Aldrich) and 1-pyrenebutanol (Sigma-Aldrich).[14] A mixture (4:4:1 stoichiometric ratio) of triethyl amine, methacryloyl chloride, and 1-pyrenebutanol, respectively, in tetrahydrofuran (THF) was stirred under nitrogen at 195 K for 12 h. The crude product was dissolved in toluene, washed with an aqueous solution of sodium hydrogen carbonate (Fisher Scientific) to remove amine salts, and finally purified by recrystallization.

The synthesis of labeled homopolymers and diblock copolymers was achieved via anionic polymerization.[39] Monomers and solvent (THF) were rigorously treated to remove impurities, water, and oxygen. Butyl and methyl methacrylate monomers (Sigma-Aldrich) were purified by first removing most of the oxygen via freeze–pump–thaw (FPT) cycles. The monomer was then stirred over trioctylaluminum (Sigma-Aldrich), added under nitrogen flow, to react with any protic impurities.[40] Next, nitrogen was removed via FPT cycles and the monomer was short-path vacuum transferred to a storage vessel and kept in a glovebox (MBRAUN UNIlab, < 0.1 ppm of H2O and O2) freezer. All polymerizations were carried out in THF, delivered from an MBRAUN compact solvent purification system, in the coldwell of the glovebox, which was cooled by an external dry ice–isopropanol bath at 195 K.

The polymerization conditions described below apply to the synthesis of both labeled homopolymers and diblock copolymers. A glass reactor rinsed with sec-butyllithium (s-BuLi, Sigma-Aldrich) was filled with clean THF (20:1 solvent to monomer volume ratio) and cooled to 195 K. Lithium chloride (LiCl, Sigma-Aldrich) and diphenylethylene (DPE, Sigma-Aldrich) were added to the reactor in a 10:1 LiCl:s-BuLi mole ratio and 3:1 DPE:s-BuLi mole ratio. LiCl was added to minimize attack on the C=O bond.[38] DPE was added to form a sterically hindered initiator with s-BuLi. Prior to starting polymerization, the reactor containing THF, LiCl, and DPE was titrated with s-BuLi until a red color persisted. Next, a predetermined amount, based on the target molecular weight and polymer batch size, of s-BuLi was added to the reactor. For the synthesis of labeled homopolymers, a mixture of labeled and unlabeled monomer was added to the reactor in a dropwise manner over a period of 1–2 min, allowed to react for 10 min, and then terminated by the addition of methanol, which capped the chain with a proton. A reaction time of 10 min was well in excess of the time required for essentially complete conversion (>99%) of the monomer.[41]

The synthesis of diblock copolymers of PBMA–PMMA is analogous to that described above for PMMA homopolymer; the PBMA block was polymerized first. The type and number of necessary monomer charges depend on the desired location of the label in the product, but in all cases, 10 min was allowed for the propagation after each monomer charge. For diblock copolymers labeled at specific positions along the chain, only 1% of the respective block was labeled. The label was added at trace levels, i.e., < 0.5 mol % of any monomer charge. This translates to a typical value of three labeled monomers per chain for the uniform labeling case; for the copolymers labeled at specific positions along the chain, this translates to approximately one label per 30 chains. All polymers were recovered postsynthesis by precipitation into methanol and then drying in a vacuum oven.

Molecular Characterization

The polymer molecular weights and dispersities (Đ) were characterized by gel permeation chromatography (GPC). The GPC system employed a model 515 pump (Waters) delivering THF mobile phase at 1 mL/min, two PLgel Mixed-C 30 cm columns (Agilent) operated at 308 K, a miniDAWN TREOS light scattering (LS) detector (Wyatt Technologies, 658 nm, room temperature), an Optilab T-rEX differential refractive index (DRI) detector (Wyatt Technologies, 658 nm, 298 K), and a Model 2487 Dual-Wavelength UV–visible absorbance detector (Waters). Polymer dispersity was measured using the DRI signal, with the elution times calibrated with narrow-distribution polystyrene standards. The true weight-average molecular weight of homopolymers of PMMA and PBMA was characterized by LS with specific refractive index increments (dn/dc) measured independently on an Optilab rEX differential refractometer (Wyatt Technologies): 0.0818 and 0.0763 mL/g for PMMA and PBMA in THF at 298 K and λ = 658 nm, respectively. For diblock copolymers, the true weight-average molecular weight was characterized by LS using a weight-fraction-weighted dn/dc.[42] In all cases the weight-average molecular weight from LS was divided by Đ from the DRI signal to yield the number-average molecular weight M, reported herein. The composition of each diblock copolymer was determined by proton nuclear magnetic resonance (1H NMR) in chloroform-d using a Bruker AVANCE III spectrometer operating at 500 MHz. The relative intensity of the resonances corresponding to the O–CH3 and O–CH2– protons of PMMA and PBMA, located at δ = 3.6 ppm and δ = 3.93 ppm, respectively, defined the diblock copolymer composition. The purity of 1-pyrenylbutyl methacrylate monomer was also characterized by 1H NMR.

Morphological Characterization

Small-angle X-ray scattering (SAXS) patterns were collected in transmission using nickel-filtered Cu Kα radiation from a PANalytical PW3830 generator with a PANalytical C-Tech long fine focus tube, a compact Kratky camera (Anton-Paar), and a BRAUN OED-50 M one-dimensional position-sensitive detector. Samples were mounted into a home-built hot stage, held in copper cells between mica windows. Data were corrected for empty beam scattering, detector sensitivity, positional linearity, sample thickness, and transmittance and were desmeared for slit length.[43] Absolute scattering intensities (I/IeV) were obtained based on a polyethylene standard and plotted as a function of the magnitude of the momentum transfer vector, q = (4π/λ)sinθ, where θ is half the scattering angle.

Tg Measurement

The bulk Tg values of homopolymers of PBMA and PMMA and diblock copolymers of PBMA–PMMA were measured using differential scanning calorimetry (TA Instruments Q2000, second heating at a heating rate of 2 K/min) calibrated with sapphire and indium standards. A typical run employed a 7 mg polymer sample. All reported Tg,Bulk values correspond to the transition onset defined as the intersection of the glassy line and transition line of a calorimetric thermogram (see Figure S5). The fluorimetric Tg was characterized on spin-coated or dropcast films deposited onto a silica substrate (VG-9 glass, Schott, North America) at ∼10 μm thicknesses. The films were annealed at Tg,Bulk + 30 K for 12 h. The fluorescence emission intensity was measured using a Fluorolog-3 spectrofluorimeter (Horiba Scientific). A typical measurement consists of rapidly heating a film to Tg,Bulk + 20 K (±10 K), maintaining an isotherm for 20 min to remove the processing history of the film, then collecting the steady-state fluorescence emission spectrum at 5 K intervals at a 1 or 2 K/min cooling rate. The pyrene labels were excited at 347 nm with a 1 nm bandpass. The emission spectra were collected over the range of 350–500 nm.