2025 Volume 8 Issue 4

Open Access

INEOS OPEN, 2025, 8 (4), 193–199  Journal of Nesmeyanov Institute of Organoelement Compounds of the Russian Academy of Sciences Download PDF DOI: 10.32931/io2449a

Thermally Induced Gelation of Water-Soluble Bottlebrush
Block Copolymers with Temperature-Responsive Blocks

A. Yu. Konyakhina,^a A. Z. Umarov,^b E. A. Nikitina,^b A. I. Buglakov,^b,c M. Rosenthal,^d
E. N. Subcheva,^a J. Garcia,^e F. Vashahi,^e S. S. Sheiko,^e and D. A. Ivanov*^b,f

^a Scientific Center for Genetics and Life Sciences, Sirius University of Science and Technology, Olimpiyskiy pr. 1, Sochi, 354340 Russia
^b Faculty of Chemistry, Lomonosov Moscow State University, Leninskie Gory 1, str. 3, Moscow, 119991 Russia
^c Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, ul. Vavilova 28, str. 1, Moscow, 119334 Russia
^d Department of Chemistry, KU Leuven, Celestijnenlaan 200F, Box 2404, B-3001 Leuven, Belgium
^e University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
^f Institut de Sciences des Matériaux de Mulhouse (IS2M)-CNRS UMR 7361, 15, rue Jean Starcky, F-68057, Mulhouse, France
 

Corresponding author: D. A. Ivanov, e-mail: dimitri.ivanov.2014@gmail.com
Received 8 November 2024; accepted 22 February 2025

Abstract

The gelation of triblock copolymers, comprising a poly(ethylene oxide) (PEO) bottlebrush central block and thermosensitive poly(N-isopropylacrylamide) (PNIPAM) end blocks, was investigated using synchrotron X-ray scattering, transmission electron and atomic force microscopy, and computer simulations. Above the PNIPAM lower critical solution temperature, an interference peak in X-ray scattering emerged that corresponds to the PNIPAM domains. Coarse-grained simulations supported these findings, showing scattering from phase-segregated terminal block domains and a distinct domain form-factor.

Key words: bottlebrush block copolymers, thermosensitive polymers, physical gels, tissue-mimetic mechanics.

 

Introduction

The unique structure and mechanical properties of bottlebrush polymers (BBs) has attracted significant interest over the past decade [1–3]. In particular, their strain-stiffening properties, stemming from steric repulsion between densely grafted side chains, make BBs especially promising for biomedical applications, where they closely mimic the mechanical behavior of soft biological tissues [4–7]. The capacity to control the self-assembly and mechanical behavior of these materials by adjusting key architectural parameters, such as side-chain length and grafting density, allows precise tuning of their mechanical properties, including Young's modulus at low deformations and strain-stiffening, or firmness at higher deformations [4, 8–10]. This architectural flexibility, independent of the polymer chemical structure, provides a unique platform for the rational design of biomaterials with tailored properties [11].

The incorporation of stimuli-responsive segments into the BB structure allows for the design of self-assembled structures that respond to external stimuli [12, 13]. Among these, poly(N-isopropylacrylamide) stands out due to its lower critical solution temperature (LCST), being close to human body temperature, which makes it a highly attractive candidate for biomedical applications.

In this paper, we report on structural analysis of the gelation process for a series of ABA triblock copolymers where A stands for the thermosensitive PNIPAM block and B denotes a bottlebrush block with water-soluble PEO side chains. A combination of experimental techniques including atomic force microscopy (AFM), transmission electron microscopy (TEM), and variable-temperature synchrotron ultrasmall- and small-angle X-ray scattering (USAXS-SAXS) was employed to explore the ABA structural organization on different length scales.

For more than two decades, AFM imaging of individual polymer chains has proven invaluable for verifying synthetic approaches, offering insights into branching topology and molecular conformation that are often beyond the reach of conventional characterization techniques [14–16]. Additionally, AFM imaging at molecular resolution yields quantitative data on essential molecular parameters, including contour length, curvature, and end-to-end distance, which are crucial for evaluating the configuration and flexibility of surface-adsorbed chains [9, 11, 13, 17]. In some instances, AFM has also enabled real-time observation of chain dynamics, shedding light on molecular ordering on substrates and structural organization within polymer films [18].

In previous studies, USAXS-SAXS techniques have been applied to investigate the structural evolution during gelation across multiple spatial scales [19–21]. The scattering data have typically been analyzed as a combination of fractal object scattering [19], Guinier scattering [19, 22], and the Ornstein–Zernike law with an associated correlation length [22]. However, to our knowledge, no X-ray scattering measurements on the gelation of bottlebrush copolymers have been reported to date.

To further clarify the molecular mechanisms underlying the gelation process, we employed coarse-grained computer simulations. This combined approach, integrating experimental techniques with modeling, enabled a comprehensive investigation of the phase transition in these complex hierarchical systems.

Results and discussion

The chemical structure of the linear ABA triblock copolymers under investigation is shown in Scheme 1. These thermo-responsive triblock copolymers have been shown to form hydrogels through the self-assembly of A blocks into spherical domains that serve as physical cross-links between strands of the central B block [15].

Scheme 1. Structure of the PNIPAM–PEO–PNIPAM copolymer, containing a central bottlebrush block
with PEO side chains and PNIPAM linear terminal blocks.

The molecular characteristics of the synthesized samples are summarized in Table 1. These samples contain BB backbones with varying degrees of polymerization (DP), specifically n_BB values of 292, 576, and 888. The volume fraction of the terminal PNIPAM blocks, φ_A, varies from 6% to 24%, while the side chains of the BB block maintain a constant DP of 19.

Table 1. Structural parameters of the PNIPAM–PEO–PNIPAM bottlebrush triblock copolymers

Sample	nsca	nBBb	nAc	φAd
PPP 292-6	19	292	81	0.06
PPP 292-12	19	292	174	0.12
PPP 292-24	19	292	342	0.22
PPP 576-6	19	576	1895	0.06
PPP 576-12	19	576	301	0.12
PPP 576-24	19	576	632	0.24
PPP 888-24	19	888	1144	0.24
a degree of polymerization (DP) of side chains in the bottlebrush block; b DP of the main bottlebrush backbone; c DP of the symmetric terminal PNIPAM blocks; d volume fraction of PNIPAM.

Atomic force microscopy

Figure 1 shows the AFM tapping mode height images of the individual PNIPAM–PEO–PNIPAM bottlebrush copolymers deposited on a mica substrate, illustrating the distinct topographic contrast between the substrate and the adsorbed polymer chains. The macromolecules display characteristic wormlike structures, further validated by tangent-tangent correlation function calculations along the polymer contour (not shown here). Notably, the coiled conformation of dispersed macromolecules (Fig. 1a) contrasts with the extended conformation of densely packed monolayers (Fig. 1b).

Figure 1. AFM tapping mode height images showing individual bottlebrush chains.
Polymer aqueous solution (0.01 g/L) was deposited on mica substrate: PPP 292-6 (a), PPP 292-24 (b).

The AFM molecular images can be quantified by measuring contour lengths (L_c) for a representative set of 200 molecules (Fig. 2). The terminal PNIPAM blocks are generally not visible in the AFM images, meaning that the observed contour length corresponds entirely to the bottlebrush block. The molecular length distributions (MLD) obtained for the ABA copolymers with a backbone degree of polymerization (n_BB) of 292 and A blocks of n_A = 81 and 342 (Figs. 2a,b) show that the number-average contour lengths of 91.5 nm and 114.4 nm, respectively, are significantly greater than that of the brush backbone (L_BB = 73 nm). In particular, the 292-24 MLD reveals a significant fraction of longer macromolecules with L_c exceeding 150 nm. This behavior suggests the ABA association, which is stronger for PPP 292-24 with larger n_A (inset). For the samples of the 292 series, the observed chains are predominantly dimers, formed through the aggregation of terminal PNIPAM blocks. Similar behavior of end-functionalized brush association has been observed in the previous studies [14, 23].

Figure 2. Normalized molecular length distributions (h(L_c)) for PPP 292-6 (a) and PPP 292-24 (b), the inset shows the differential
histogram h(292-24)–h(292-6); MLD for PPP 888-24 (c); the relationship between the measured values of L_c and n_BB (d).

In contrast, the ABA macromolecules with longer n_BB are likely to correspond to single chains. For instance, in the case of sample PPP 888-24, the number average L_c = 183 nm is reasonably close to L_BB = 222 nm.

The relationship between L_c and n_BB is illustrated in Fig. 2d. Here, the values of L_c for samples PPP 292-6 through PPP 292-24 were halved to account for the aggregation effects. A linear fit through the origin provides a slope of 0.2, indicating that the chains on mica are nearly fully extended (the theoretical maximum being 0.25).

Transmission electron microscopy

The morphological changes occurring during gelation were visualized using TEM. The image in Fig. 3a captures the room-temperature structure of PPP 576-12 in aqueous solution, where the wormlike bottlebrush blocks exhibit a morphology similar to that observed in AFM. Additionally, bright spots are visible at the chain ends that likely correspond to globular structures of the terminal PNIPAM blocks. The globule diameter of 4.3 ± 0.6 nm aligns closely with the calculated diameter (d_calc) of 4.6 nm, derived from n_A and the mass density of PNIPAM.

Figure 3. TEM images of PPP 576-12: the configuration in homogeneous aqueous solution (0.001 g/L) at room temperature, displaying individual wormlike bottlebrush blocks with PNIPAM globules visible as bright spots at the chain ends (a); the gel morphology at 45 °C, illustrating the formation of nodules linking multiple bottlebrush blocks (b).

The image in Fig. 3b provides a view of the gel structure formed in aqueous solution of PPP 576-12 at 45 °C. The gelation is accompanied by a microphase separation resulting in the formation of nodules of PNIPAM connecting multiple bottlebrush blocks. Such features visible in the TEM images have diameters ranging from approximately 20 to 50 nm and are described by the aggregation numbers on the order of 10–20. Although these morphologies may not quantitatively reflect the bulk structure, they illustrate the overall trend in gel network formation.

Ultrasmall- and small-angle synchrotron X-ray scattering

To gain deeper insight into the structural evolution during gelation, we conducted variable-temperature synchrotron X-ray scattering at ultrasmall and small angles. Aqueous solutions of the bottlebrush copolymers, sealed in capillaries, were subjected to thermal cycling between 15 and 45 °C, with scattering patterns recorded during both heating and cooling. To minimize X-ray beam damage, exposure time was reduced, and the capillary was continuously moved using a 2D mesh scan mode.

Figures 4a and 4b display selected scattering measurements for sample PPP 576-24 at a concentration of C_p = 20 wt %, recorded during heating and cooling, respectively. The curves reveal three key features: the "bottlebrush peak" at q₁, which corresponds to backbone-to-backbone correlation [12, 24–26]; the first form-factor minimum of the PNIPAM spheres at q₂; and the primary interference peak at q₃, corresponding to the distance between the PNIPAM spheres (Fig. 4a).

Figure 4. Variable-temperature USAXS-SAXS measurements of 20% aqueous solution of PPP 576-24
during heating (a) and cooling (b), conducted at a rate of 1.0 °C/min.

In the initially homogeneous solution at low temperatures, only the bottlebrush peak, with a d-spacing d₁ = 7.5 nm, is detectable. However, as gelation sets in, the main interference peak and form-factor emerge, where q₃ corresponds to d ≅ 130 nm, while the PNIPAM domain diameter ranges between 23 and 28 nm. With the progression of gelation, the intensity of the main interference peak markedly increases, indicating enhanced microphase separation. Additionally, subtle changes in the shape of the bottlebrush peak, including a skew on the low-q side, are observed. These changes are likely to result from water expulsion from the collapsing PNIPAM chains, leading to additional swelling within the bottlebrush matrix.

On cooling, the system restores the original homogeneous state with some hysteresis (Fig. 4b), which is typical for such transitions [12, 27].

The progression of gelation can be monitored by calculating the SAXS invariant, defined as the integral of Lorentz-corrected scattering intensity over the q-range from 0.02 to 0.2 nm^–1 (Fig. 5). The resulting curves display a sigmoidal shape with a slight hysteresis of a few degrees between the heating and cooling cycles. Notably, the samples with higher φ_A values exhibit transitions at lower temperatures, which is consistent with the previous observations [12].

Figure 5. Temperature dependence of the normalized SAXS invariant for 20% aqueous solutions of PPP 576-6, PPP 576-12, and PPP 576-24.

The size of the PNIPAM domains depends on both the polymer concentration C_p and temperature, with larger radii observed at higher C_p. The most pronounced gelation effect on the PNIPAM aggregation was observed for PPP 576-24 with φ_L = 0.24: the radius (R) increased significantly from approximately 16 to 29 nm (Fig. 6a).

Figure 6. Correlation between the PNIPAM sphere radius (R) and temperature for PPP 576-24 at different concentrations
in aqueous solution C_p (a); dependence of R on C_p for PPP 576-24 (b).

Further insights into the molecular mechanisms of the process of gelation can be grasped from computer simulations.

Simulations of the gelation process

Dissipative particle dynamics [28] simulations of the triblock copolymer solutions were conducted using a generic coarse-grained model of the A-BB-A polymer, comprising a hydrophilic bottlebrush central block and two thermo-responsive linear end blocks. The details of the model are provided in the Experiment and calculations section. The simulations explored the physical gelation process and analyzed the resulting gel morphologies, with a focus on the effects of the polymer concentration (C_p) and end-block degree of polymerization (n_A).

As the repulsion between the A block and solvent S increases, spherical aggregates of A blocks form, surrounded by bottlebrush segments. Figure 7 provides snapshots of the A-BB-A solution with n_A = 60 and high A block insolubility in S. At low polymer concentration (С_p = 0.01), distinct clusters of connected triblock copolymer micelles emerge, with connections between A aggregates established through BB blocks. An increase in the concentration of С_p to 0.05 leads to the formation of a percolated network, characteristic of a physical gel, where spherical aggregates of end blocks serve as cross-links joined by neighboring bottlebrush subchains. Notably, some BB blocks form long bridges between cross-links, while others form loops, with both end blocks of a single A-BB-A molecule associating within the same aggregate. At higher concentrations (C_p = 0.10, 0.20 and more), similar physical networks persist in the system.

Figure 7. Results of the coarse-grained molecular simulations: representative snapshots of the physical gels of the bottlebrush triblock copolymers with n_A = 60 at varying polymer volume fractions (C_p), the domains of A end block are shown in red, bottlebrush main chains in black, and bottlebrush side chains in blue (a); static structure factor of S_A end block at different repulsion parameters between the end-block beads and the solvent a_AS (b); dependences of aggregation number of the end blocks on C_p (c); cross-link density as a function of C_p for different degrees of end block polymerization n_A (d).

The process of A-BB-A gel formation from a homogeneous solution upon worsening solvent quality for the end blocks was monitored by calculating the static structure factor S_A(q/2π) for A beads. Figure 7b shows S_A(q/2π) for the A-BB-A copolymers with n_A = 60 and C_p = 0.10 across various repulsion levels (a_AS) between solvent S and A beads. For a_AS = 101 up to a_AS = 104, S_A corresponds to uniformly distributed A chains. At a_AS = 105, S_A shows a sharp intensity increase in the low-q region, with an interference peak near q/2π ~ 0.2, indicating a characteristic spacing between A block aggregates. However, identifying the cross-link form factor at higher q is challenging due to the broad, polydisperse size distribution. At high insolubility (a_AS = 107), a distinct form-factor minimum and maximum appear near q/2π ~ 0.2 with a slight increase in the interference peak intensity. Further increases in a_AS cause minimal changes in S_A. These simulated structure factors closely mirror the experimental USAXS-SAXS data (Fig. 4), except for a missing higher-q peak in the simulation data, which corresponds to unentangled bottlebrush block interference maximum and would be present in the structure factor of the main brush block (vide infra).

The simulations also provided insights into the morphology of the physical gels by analyzing the aggregation number N_agg and cross-link density ρ_cs = K_agg/V as functions of the polymer concentration, calculated across different end-block polymerization degrees n_A (Figs. 7c,d). For varying n_A, both dependences exhibit similar trends, with higher n_A leading to greater N_agg and lower ρ_cs. Notably, two regimes emerge for N_agg and ρ_cs as functions of the polymer concentration. At lower C_p (up to ~ 0.1), the micellar cross-link size increases gradually with C_p, as the new polymer content primarily contributes to an increase in A block cluster numbers. At C_p > 0.1, the micellar size escalates more dramatically, while cross-link densities stabilize, indicating a plateau. For lower concentrations, a plateau in the cross-link size might also occur, though testing this would require a larger system and increased computational resources.

The gelation features and morphological characteristics of the networks obtained from molecular simulations closely resemble those observed in physical gels of fully linear triblock copolymers with hydrophilic central blocks and hydrophobic end blocks [29]. However, a key structural difference emerges in the volume packing of the bottlebrush blocks. Due to the high excluded volume of densely grafted side chains, the main chains of the bottlebrush blocks remain disentangled and have increased rigidity compared to their linear counterparts.

Figure 8a displays the static structure factors S_C(q/2π) of the bottlebrush block main chains across various polymer concentrations in solution. Each S_C profile exhibits a peak at an intermediate q/2π, within the range of 0.1 to 0.25, depending on the polymer content C_p. Similar peaks in the main-chain structure factors have been observed in the previous studies [12, 24–26], corresponding to high-q SAXS peaks, and are associated with the characteristic backbone-to-backbone correlation. Notably, we found that the square of the corresponding d-spacing (d₁) varies linearly with the polymer concentration (Fig. 8b), following d₁² ~ 1/C_p. This result suggests that even in solution, the bottlebrush structures pack locally as cylinders with their main axes approximately aligned, imparting a weakly nematic-like order to the gel system. In contrast, for isotropic objects like spheres, the spacing would scale as d³ ~ 1/C_p. This distinct concentration dependence [30] is a remarkable characteristic of bottlebrush physical gels, positioning them as intermediate materials between isotropic systems and liquid crystals.

Figure 8. Static structure factors S_C(q/2π) for the bottlebrush main chains, obtained from coarse-grained simulations, showing the structural organization at different solvent conditions (a); dependence of squared d-spacing of the bottlebrush peak positions on the polymer volume fraction C_p, illustrating the relationship between the structural arrangement and concentration (b).

Experiment and calculations

Atomic force microscopy

Individual macromolecules were visualized using AFM. A sample (5 μL) of a 0.01 g/L polymer solution was applied to a freshly cleaved mica surface that had been pre-treated with plasma. The solution was immediately removed using an air stream to achieve uniform dispersion. AFM measurements were carried out with a Bruker JPK NanoWizard Ultraspeed 2 atomic force microscope in an intermittent contact mode, in air, at a scan rate of 1 Hz. Bruker RTESPA-300 cantilevers were utilized, featuring a resonance frequency of 200–400 kHz and a probe stiffness of 40 N/m. The standard image processing and contour length measurements were performed using the FemtoScan online software (Advanced Technologies Center, Russia) and then graphed with Igor Pro (WaveMetrics, Inc.).

Transmission electron microscopy

TEM was employed for visualization of individual macromolecules and studying the gelation process. An aliquot (10 μL) of a 0.001 g/L polymer solution was applied onto freshly glow-discharged copper grids (230 mesh, formvar-carbon coated) and incubated for 5 min either at room temperature or at 45 °C. The grids were then washed and stained with a drop of 0.2% aqueous uranyl acetate solution. The samples were observed using a Carl Zeiss CrossBeam 550 (Carl Zeiss Microscopy GmbH, Jena, Germany) instrument operated at an accelerating voltage of 80 kV.

Synchrotron small-angle X-ray scattering

The USAXS-SAXS experiments were carried out at the ID02 beamline of the European Synchrotron Radiation Facility (ESRF) in Grenoble (France). The measurements were conducted in transmission geometry using photon energy of 12 keV. The available q-values, with |q| = 4π sin(θ)/λ, where θ is the Bragg angle and λ is the wavelength, cover a range from 8.0 × 10^–2 nm^–1 to 1.5 nm^–1. A Pilatus 1M detector (169 mm × 179 mm active area) was employed for recording scattering intensity at a sample-to-detector distance of 9.0 m. A custom-built capillary stage with a Peltier element allowed for temperature variation between 15 and 55 °C.  The data correction, calibration, and integration were performed using the fast azimuthal integration Python library [31].

The diameter of the A block domains was determined from the SAXS curves by fitting them to the form-factor of spheres.

The analytical expression in equation 1 was convoluted with a Gaussian distribution to account for the scattering from a polydisperse population of spheres.

Coarse-grained simulations

We considered a generic model of an A-BB-A triblock copolymer with hydrophilic BB central block and thermo-responsive linear A end blocks. Main chain of the BB block consisted of n_C linearly connected C beads. Every C bead has a grafted chain of n_O O beads; thus, linear grafting density of the BB block was equal to 1. The end blocks consisted of bonded n_A A beads. Each bond between connected particles in the triblock was set to be identical for simplicity. A solvent was explicitly introduced into the model of the triblock copolymer solution via S bead. Each spherical bead had the diameter R = 1. All parameters are given in dimensionless units.

The solutions of the A-BB-A triblock copolymers were simulated in an NVT ensemble with explicit solvent using dissipative particle dynamics (DPD), which is capable of reproducing hydrodynamic interactions since local momentum is conserved within this methodology [28]. Here, the interactions between beads i and j are a sum of pairwise conservative (F^C_ij), random (F^R_ij) and dissipative (F^D_ij) forces with additional bond (F^B_ij), acting between every two connected particles. Conservative force accounts for both excluded volume interaction and miscibility between beads and is written as  , where r_ij is a vector between i and j particles and r_ij = |r_ij|. R is a cut-off distance and a_αβ is constant, which defines affinity between each particle type (α, β = A, C, O, S). , where k_b is the stiffness and l is the equilibrium bond length. The functional forms of random and dissipative forces, which together implement thermostating of the system, can be found elsewhere [28].

The solution of the A-BB-A triblock copolymers was modeled as a periodic cubic box of edge length L = 60, filled with macromolecules with a volume fraction C_p and the solvent with a fraction (1 – C_p) in random fashion with average numerical density of beads ρ = 3. The volume fraction of the polymer C_p varied between 0.01 and 0.40. The architectural parameters of the triblock copolymer were chosen as n_C = 60, n_O = 7. Three different polymerization degrees of linear end block were considered: n_A = 30, n_A = 45, and n_A = 60. The repulsion parameters between the beads of the same type were set as a_AA = a_CC = a_HH = a_SS = 100, the bonding parameters were set as k_b = 100, l = 0.7. This setup ensures the absence of unphysical crossing between macromolecules during the dynamical simulations with soft-core potentials, similar to the scheme, proposed by Nikunen et al. [32]. Within the polymer considered in the experimental section, the bottlebrush block consists of a hydrophobic main chain and hydrophilic side-chains; therefore, the interaction parameters were set as a_CS = 110, a_OS = 100, a_CH = 110, which ensures solubility of the overall BB block, but also accounts for probable conformational changes of the main-chain as was shown in Ref. [3]. The temperature-responsivity of linear PNIPAM with LCST behavior leads to a decrease in the solubility of linear end blocks upon temperature increase in water. This behavior was simulated via a gradual increase of the repulsion between A beads and solvent S beads and between A and O beads from a_AS = a_AO = 100 to a_AS = a_AO = 110. The interaction parameters were chosen to prevent the system from freezing and to satisfy the condition to reproduce suitable density fluctuation of water, for the details see Ref. [33].

The calculations were carried out using the Velocity-Verlet integration scheme with time step Δt = 0.04. Starting from a homogeneous solution of the A-BB-A triblock copolymer with a_AS = a_AO = 100, every 5·10⁶ Δt a_AS and a_AO were increased by an increment Δa = 1 until a_AS = a_AO = 110 were reached, followed by the relation run during 5·10⁶ Δt. The statistical data were collected over the last 2·10⁶ timesteps and within 3 independent ensemble runs.

For the obtained trajectory aggregation numbers of linear block micelles N_agg and cross-link density 𝜌_cs = K_agg/V were calculated, where K_agg is the number of the A block aggregates and V is the system volume, using standard cluster analysis with cut-off radius r_cut = 1.0 [34].

Additionally, the static structure factors , where q is the scattering vector, n is the number of particles, and r_j is the radius vector of j bead, were calculated for the main-chain of the BB blocks (C beads)—S_С(q/2𝜋)—and for the linear end block (A beads)—S_A(q/2𝜋).

Conclusions

The gelation process of a series of bottlebrush–linear block copolymers, featuring a central BB block with water-soluble PEO side chains and symmetric, thermosensitive PNIPAM terminal blocks, was investigated for the first time using synchrotron ultrasmall- and small-angle X-ray scattering and TEM. The initial characterization by AFM involved measuring the contour lengths of these copolymers on the mica substrate. The results showed that the bottlebrushes with the shortest backbone (n_BB = 292) demonstrated a tendency to form dimers through PNIPAM segment aggregation, which intensified with longer PNIPAM blocks. In contrast, the copolymers with larger values of n_BB exhibited no aggregation at room temperature. A linear correlation was observed between the contour length (L_c) and n_BB, with a slope of 0.2, approaching the theoretical fully extended chain conformation slope of 0.25.

At low temperatures, the initially homogeneous aqueous solutions of the copolymers display a single "bottlebrush peak", indicative of backbone-to-backbone correlations. Upon heating to the lower critical solution temperature of PNIPAM, an additional interference maximum appeared at q₃, corresponding to the average distance between PNIPAM nodules, alongside a form-factor for these domains. The diameter of the PNIPAM domain increases with temperature, suggesting progressing phase segregation within the system. Additionally, the aggregation number was observed to increase with the water concentration in the bottlebrush solutions.

The coarse-grained simulations effectively capture the key aspects of the structural evolution of bottlebrush copolymers during gelation. The simulations revealed an intense interference maximum attributed to scattering from phase-segregated domains of the terminal blocks, accompanied by a distinct form-factor for these domains. Additionally, the simulations showed an increase in the aggregation number within the terminal block domains as solvent quality worsens. Notably, the simulations also replicated the shift in the bottlebrush peak with the increasing polymer concentration, supporting the correlation of d₁² ~ 1/C_p, which implies that even in solution the bottlebrush segments tend to locally pack as aligned cylinders.

Acknowledgements

The authors are grateful to the European Synchrotron Radiation Facility for providing synchrotron radiation facilities and the personnel of the ID02 and BM26 beamlines for the excellent technical support. D.A.I., A.Z.U. and E.A.N. are grateful to the Russian Science Foundation (project no. 23-73-30005). The simulations and data processing were performed using the facilities of the Interlaboratory Computer Center at INEOS RAS supported by the Ministry of Science and Higher Education of the Russian Federation. The ABA synthesis was supported by the National Science Foundation (DMR 2403716 and DMR 2324167).

References

1. Z. Li, M. Tang, S. Liang, M. Zhang, G. M. Biesold, Y. He, S.-M. Hao, W. Choi, Y. Liu, J. Peng, Z. Lin, Prog. Polym. Sci., 2021, 116, 101387. DOI: 10.1016/j.progpolymsci.2021.101387
2. F. Uhlik, O. V. Rud, O. V. Borisov, E. B. Zhulina, Gels, 2022, 8, 793. DOI: 10.3390/gels8120793
3. C. J. Wang, F. Vashahi, I. Moutsios, A. Z. Umarov, G. G. Ageev, Z. Wang, D. A. Ivanov, A. V. Dobrynin, S. S. Sheiko, Adv. Funct. Mater., 2024, 34, 2410905. DOI: 10.1002/adfm.202410905
4. M. Vatankhah-Varnosfaderani, A. N. Keith, Y. Cong, H. Liang, M. Rosenthal, M. Sztucki, C. Clair, S. Magonov, D. A. Ivanov, A. V. Dobrynin, S. S. Sheiko, Science, 2018, 359, 1509–1513. DOI: 10.1126/science.aar5308
5. G. Xie, M. R. Martinez, M. Olszewski, S. S. Sheiko, K. Matyjaszewski, Biomacromolecules, 2019, 20, 27–54. DOI: 10.1021/acs.biomac.8b01171
6. M. Maw, E. Dashtimoghadam, A. N. Keith, B. J. Morgan, A. K. Tanas, E. Nikitina, D. A. Ivanov, M. Vatankhah-Varnosfaderani, A. V. Dobrynin, S. S. Sheiko, ACS Cent. Sci., 2023, 9, 197–205. DOI: 10.1021/acscentsci.2c01407
7. D. Zhang, E. Dashtimoghadam, F. Fahimipour, X. Hu, Q. Li, E. A. Bersenev, D. A. Ivanov, M. Vatankhah-Varnoosfaderani, S. S. Sheiko, Adv. Mater., 2020, 32, 2005314. DOI: 10.1002/adma.202005314
8. S. S. Sheiko, A. V. Dobrynin, Macromolecules, 2019, 52, 7531–7546. DOI: 10.1021/acs.macromol.9b01127
9. A. N. Keith, C. Clair, A. Lallam, E. A. Bersenev, D. A. Ivanov, Y. Tian, A. V. Dobrynin, S. S. Sheiko, Macromolecules, 2020, 53, 9306–9312. DOI: 10.1021/acs.macromol.0c01725
10. A. Z. Umarov, J. Collins, E. A. Nikitina, I. Moutsios, M. Rosenthal, A. V. Dobrynin, S. S. Sheiko, D. A. Ivanov, Macromol. Rapid Commun., 2024, 46, 2400569. DOI: 10.1002/marc.202400569
11. W. F. M. Daniel, J. Burdyńska, M. Vatankhah-Varnoosfaderani, K. Matyjaszewski, J. Paturej, M. Rubinstein, A. V. Dobrynin, S. S. Sheiko, Nat. Mater., 2016, 15, 183–189. DOI: 10.1038/NMAT4508
12. F. Vashahi, M. R. Martinez, E. Dashtimoghadam, F. Fahimipour, A. N. Keith, E. A. Bersenev, D. A. Ivanov, E. B. Zhulina, P. Popryadukhin, K. Matyjaszewski, M. Vatankhah-Varnosfaderani, S. S. Sheiko, Sci. Adv., 2022, 8, eabm2469. DOI: 10.1126/sciadv.abm2469
13. D. M. Henn, W. Fu, S. Mei, C. Y. Li, B. Zhao, Macromolecules, 2017, 50, 1645–1656. DOI: 10.1021/acs.macromol.7b00150
14. S. Y. Yu-Su, S. S. Sheiko, H.-y. Lee, W. Jakubowski, A. Nese, K. Matyjaszewski, D. Anokhin, D. A. Ivanov, Macromolecules, 2009, 42, 9008–9017. DOI: 10.1021/ma901432v
15. S. S. Sheiko, B. S. Sumerlin, K. Matyjaszewski, Prog. Polym. Sci., 2008, 33, 759–785. DOI: 10.1016/j.progpolymsci.2008.05.001
16. S. Y. Yu-Su, F. C. Sun, S. S. Sheiko, D. Konkolewicz, H.-i. Lee, K. Matyjaszewski, Macromolecules, 2011, 44, 5928–5936. DOI: 10.1021/ma200821a
17. S. S. Sheiko, M. Möller, Chem. Rev., 2001, 101, 4099–4124. DOI: 10.1021/cr990129v
18. H. Xu, S. S. Sheiko, D. Shirvanyants, M. Rubinstein, K. L. Beers, K. Matyjaszewski, Langmuir, 2006, 22, 1254–1259. DOI: 10.1021/la052163c
19. M. Chalal, F. Ehrburger-Dolle, I. Morfin, F. Bley, M.-R. Aguilar de Armas, M.-L. López Donaire, J. San Roman, N. Bölgen, E. Pişkin, O. Ziane, R. Casalegno, Macromolecules, 2010, 43, 2009–2017. DOI: 10.1021/ma902655h
20. I. W. Hamley, V. Castelletto, N. M. P. S. Ricardo, M. E. N. Pinho, C. Booth, D. Attwood, Z. Yang, Polym. Int., 2007, 56, 88–92. DOI: 10.1002/pi.2118
21. T. Nicolai, M. Pouzot, D. Durand, M. Weijers, R. W. Visschers, EPL, 2006, 73, 299. DOI: 10.1209/epl/i2005-10391-8
22. K. Aoki, A. Sugawara-Narutaki, R. Takahashi, J. Phys. Chem. Lett., 2023, 14, 10396−10401. DOI: 10.1021/acs.jpclett.3c02631
23. S. Qin, K. Matyjaszewski, H. Xu, S. S. Sheiko, Macromolecules, 2003, 36, 605–612. DOI: 10.1021/ma021472w
24. M. J. Rymaruk, C. T. O'Brien, C. György, B. Darmau, J. Jennings, O. O. Mykhaylyk, S. P. Armes, Angew. Chem., Int. Ed., 2021, 60, 12955–12963. DOI: 10.1002/anie.202101851
25. L. Liberman, M. L. Coughlin, S. Weigand, F. S. Bates, T. P. Lodge, Macromolecules, 2022, 55, 2821–2831. DOI: 10.1021/acs.macromol.2c00337
26. C. Clair, A. Lallam, M. Rosenthal, M. Sztucki, M. Vatankhah-Varnosfaderani, A. N. Keith, Y. Cong, H. Liang, A. V. Dobrynin, S. S. Sheiko, D. A. Ivanov, ACS Macro Lett., 2019, 8, 530–534. DOI: 10.1021/acsmacrolett.9b00106
27. K. Hamamura, K. Watanabe, Y. Sanada, F. Tanaka, Y. Katsumoto, Polymer, 2019, 161, 92–100. DOI: 10.1016/j.polymer.2018.12.001
28. R. D. Groot, P. B. Warren, J. Chem. Phys., 1997, 107, 4423–4435. DOI: 10.1063/1.474784
29. T. L. Chantawansri, T. W. Sirk, Y. R. Sliozberg, J. Chem. Phys., 2013, 138, 024908. DOI: 10.1063/1.4774373
30. J. M. Sarapas, T. B. Martin, A. Chremos, J. F. Douglas, K. L. Beers, PNAS, 2020, 117, 5168–5175. DOI: 10.1073/pnas.1916362117
31. G. Ashiotis, A. Deschildre, Z. Nawaz, J. P. Wright, D. Karkoulis, F. E. Picca, J. Kieffer, J. Appl. Crystallogr., 2015, 48, 510–519. DOI: 10.1107/S1600576715004306
32. P. Nikunen, I. Vattulainen, M. Karttunen, Phys. Rev. E, 2007, 75, 036713. DOI: 10.1103/PhysRevE.75.036713
33. R. L. Anderson, D. J. Bray, A. S. Ferrante, M. G. Noro, I. P. Stott, P. B. Warren, J. Chem. Phys., 2017, 147, 094503. DOI: 10.1063/1.4992111
34. M. P. Allen, D. J. Tildesley, Computer Simulation of Liquids, 2nd ed., Oxford Univ. Press, Oxford, 2017. DOI: 10.1093/oso/9780198803195.001.0001