Introduction

Ion transport in polymers is an important concept for improving solid state energy storage devices like batteries. Polymerized ionic liquids (PILs), which feature bulky ionic liquid groups tethered to polymer backbones, represent a promising advance in the field of ion conducting polymers1,2,3,4,5,6,7. PILs exhibit a unique combination of chemical and physical properties including high single ion conductivity, thermal and electrochemical stability, and selectivity and mechanical tunability making them attractive in applications such as energy storage systems8,9,10, separation media11,12, catalysis13,14 and antimicrobial15,16. Various architectures including linear17, block18,19,20 and network21,22,23 PILs have been investigated to improve conductivity, stability and mechanical integrity. Among these, block copolymers (BCPs) are particularly promising due to their ability to create nanoscale morphologies, and many factors can significantly affect the ionic conductivity of PIL BCP systems, including molecular weight, neutral block chemistry, doping with other components, tortuosity, volume fraction (ɸPIL) of ionic block, and nanoconfinement24,25,26,27,28,29,30,31,32,33,34. Conductivity can increase, decrease, or be invariant in a PIL diblock relative to the corresponding homopolymer. BCPs allow for the separate design of each block to provide complementary functionality between ionic transport and structural properties. PIL BCPs demonstrate a range of phase separation strength depending on the non-ionic block, which impacts the degree of confinement and ordering of ion-conducting channels, which affect ion transport. While PIL BCPs have shown great promise, the use of polymers with more ordered backbone structures (such as peptides) compared to the current random coil configurations has not been explored.

In previous work, Chen et al. developed a series of cationic polypeptides (PTPXGs) with a fixed ionic monomer structure but different secondary structures, either helical or random coil depending on the chirality of the monomer. The tethered ion was ammonium with free bis (trifluoromethane sulfonimide) (TFSI) and peptides varied from 50–950 repeating units. The helical structure greatly enhanced the ionic conductivity and thermal stability of solvent-free polymer electrolytes compared to analogous random coils, and conductivity increased with helix length35. The critical role of helical backbone on charge transport introduces a distinct parameter to the design of polymer electrolytes, and has also been shown to improve electron transport either with attached stable radical groups36 or through tunneling in monolayers37. The role of helical structure in solvent-free PIL BCPs is unexplored, both in terms of self-assembly and conductivity, in contrast to the many works on random coil polymer backbones. Rigid rod chains exhibit distinct aggregation behavior and intramolecular interaction compared to coils, and the self-assembly of rod-coil BCPs deviates from that of conventional coil–coil systems38,39. The predicted critical value of (χN)c for rod–rod and coil–coil BCPs are 8.2 and 10.5 (where χ is the Flory–Huggins interaction parameter, N is the total degree of polymerization), respectively40. Rigid rod polymers can be tuned in terms of shapes and size and leads to the formation of three-dimensional assemblies41,42,43, such as superlattices and liquid crystal (LC) phases. Peptides derived from N-carboxyanhydride (NCA) ring-opening polymerizations (ROP) are ideal rod-like ionic blocks because they can be readily synthesized to high molecular weight compared to solid-phase synthesis approaches. The helical blocks can self-assemble in PIL block copolymers and offer distinct nanoscale morphologies compared to random coils. Incorporation of helical peptide backbones into neutral BCPs can promote ordered nanostructures within the BCP matrix and has been investigated for creating functional materials with supramolecular architectures44,45,46,47,48,49. Ionic peptides have the potential to substantially impact ion transport and self-assembly, and PIL BCPs with helical ionic blocks provide an additional design parameter for polymer electrolytes.

In this study, helical–helical polypeptide PIL BCPs (PPIL BCPs) composed of an ionic poly(γ-3-triazolepropanyl-L-glutamate) (PTPLG) and non-ionic poly-(γ-benzyl-L-glutamate) (PBLG) block were synthesized. The role of helical structure on self-assembly and ion conduction was studied by varying the ionic block degree of polymerization (N = 15, 30, and 75) while keeping the non-ionic PBLG block length constant (N = 30). Strong microphase separation slightly lowered the glass transition temperature (Tg) of BCPs with higher PIL composition, although only one glass transition was observed due to the proximity of the two block Tgs. Hot-pressed films exhibited similar lamellar structure (LAM) with varying domain sizes (d) and degrees of order, as determined by small-angle X-ray scattering (SAXS). A more than 1.5 order of magnitude higher Tg-normalized and ɸPIL-normalized ionic conductivity was observed in PBLG30-b-PTPLG75 with a highly ordered bilayer LAM phase compared to less ordered structures, highlighting the importance of confinement, helix length, and nanoscale ionic domain connectivity. Furthermore, a morphology factor of \(f\) > 0.8 in PBLG30-b-PTPLG75 (above the theoretical value of 2/3 for lamellae) was achieved throughout the entire temperature window, indicating enhanced ion transport channels. Interestingly, these channels become more effective at low temperatures, as evidenced by \(f\) values greater than 1 below 65 °C. This study provides creative insights into the critical role of helical conformation in both self-assembly behavior and ionic conductivity of peptide electrolytes, providing directions for future study on structure–property relationships and the design of functional assemblies with well-defined size for next-generation stable and high-performance electrolytes.

Results and discussion

Construction of helical PPIL BCPs

The PPIL BCPs were derived from precursors with PBLG and poly(γ-(3-chloropropanyl)-L-glutamate) (PCIPLG) blocks synthesized using sequential ROPs. A series of PBLG30-b-PCIPLGns BCPs was prepared starting with a non-ionic helical PBLG-NH2 macroinitiator prepared by ROP of γ-benzyl-L-glutamate NCA (BLG-NCA, Supplementary Fig. 1) using n-hexylamine as initiator in N,N-dimethylformamide. Next, PBLG-NH2-initiated ROP of γ-(3-chloropropanyl)-L-glutamate NCA (ClPLG-NCA) in dichloromethane/phosphate buffer emulsion system (VDCM/VpH 7 buffer = 95%/5%) ([NCA]0 = 50 mM, Fig. 1) was performed. The N of PBLG-NH2 was fixed at 30 to ensure the formation of a stable α helix (Supplementary Fig. 2)50. Successful BCP precursor synthesis was confirmed by size exclusion chromatography (SEC, Fig. 2a) and 1H NMR to determine molecular weight (Mn). Attenuated total reflectance-Fourier transform infrared spectra (ATR-FTIR) was used to confirm the presence of helical conformation in the solid state. Controlled NCA-ROPs led to BCP precursors with well-defined length and low dispersity (Đ < 1.1) (Supplementary Table 1). The Mn of PBLG-NH2 (Mn = 6.83 kDa) and PBLG30-b-PCIPLGns (Mn = 9.12, 13.53 and 22.57 kDa) as determined by SEC aligned well with the theoretical Mn and those calculated from 1H NMR end group analysis (Fig. 2b). The n values and mole fraction (mol %) of PBLG30-b-PCIPLGns were calculated from the integral ratio of the proton on the central carbon of each block (4.95 and 5.05 ppm, respectively) in 1H NMR spectra (Supplementary Figs. 2–5). Azidation, copper-catalyzed azide−alkyne cycloaddition (CuAAC) click reactions and salt metatheses were employed to produce the final PPIL BCPs named PBLG30-b-PTPLGns, where positively charged ammonium groups are fixed on the end of side chain while bis(trifluoromethane) sulfonamide (TFSI) counterions remain mobile within the ionic block (Fig. 1). The final composition (n and mol %) after functionalization was confirmed by 1H NMR to be consistent with that of the corresponding precursors (Supplementary Fig. 6). Three PPILs BCPs (PBLG30-b-PTPLG10, PBLG30-b-PTPLG30 and PBLG30-b-PTPLG75) with fixed non-ionic block length (N = 30) but varying ionic block length (N = 10, 30, and 75, respectively) were obtained, as well as two homopolymers, PBLG30 and PTPLG50. All polymers were dried in a vacuum oven and stored in a glove box before use.

Fig. 1: Synthesis of helical PPIL BCPs by ring opening polymerization.
figure 1

Synthesis route for PBLG30-b-PTPLGns PPIL BCPs includes NCA formation, ring opening polymerization, chain extension, azidation, click chemistry, and anion exchange.

Full size image
Fig. 2: Characterization of molecular weight, helical secondary structure, and glass transition temperatures.
figure 2

a Normalized SEC-LS traces of α-helical macroinitiator PBLG30-NH2 (gray) produced by hexylamine-initiated ROP of BLG-NCA in DMF and BCP precursors PBLG30-b-PCIPLGns synthesized by PBLG30-NH2-initiated ROP of CIPLG-NCA in emulsion system (VDCM/VpH=7 phosphate buffer = 95% / 5%) with different feeding ratios [M]0/[I]0. b Comparison of the expected MWs, obtained MWs and Đ of PBLG30-b-PCIPLGns. c ATR-FTIR spectra indicating the characteristic amide I peak. d Second heat cycle DSC curves of PBLG30-b-PTPLGns and homopolymers.

Full size image

Conformation and thermal properties of PPIL-BCPs

The secondary structures of the PPIL BCPs in the solid-state were evaluated by ATR-FTIR and circular dichroism (Fig. 2c, Supplementary Fig. 7). As anticipated, all block copolypeptides exhibited a right-handed α-helical conformation, as a result of the strong α-helix formation tendency of each block within the appropriate N range (N ≥ 19 is required for the formation of α-helical PBLG50). The characteristic absorbance peaks of amide I at 1651 cm−1 and amide II at 1548 cm−1 are observed for all BCPs. The absence of peaks corresponding to the random coil at 1660 cm−1 indicates that both blocks predominantly adopt an α-helical conformation in PPIL BCPs. Furthermore, the helical secondary structures displayed stability across the whole temperature range investigated (25 to 140 °C), as evidenced by the invariance of absorption peak of amide I in temperature-dependent FTIR spectra (Supplementary Fig. 8).

Thermal gravimetric analysis and differential scanning calorimetry (DSC) were used for the evaluation of their thermal properties. All PPIL BCPs showed similar degradation temperature (Td, determined as 5% mass loss, Supplementary Fig. 9, Supplementary Table 2). Figure 2d shows the glass transition temperatures (Tgs) of three PPIL BCPs in comparison to the two homopolymers (Supplementary Fig. 10). Only a single Tg was observed in all BCPs, likely due to the relatively small Tg difference (10 K) between the two blocks (Tg, PBLG = 288 K and Tg, PTPLG = 298 K) and compatible nature of peptide backbones. A single Tg does not imply the absence of microphase separation that typically manifests as two distinct Tgs corresponding to the ionic block and non-ionic block, respectively51, and clear microphase separation was confirmed by SAXS. The Tg of the helical PPIL BCPs only slightly decreased with significantly increasing PIL composition. Specifically, for PBLG30-b-PTPLG30 and PBLG30-b-PTPLG75 the Tg decreased negligibly from 295 to 294 K as the PIL composition increases from 50 to 71 mol %. It was observed that PBLG30-NH2 exhibited the broadest glass transition, and as the PIL mol % increased the glass transition narrowed, reaching the smallest breadth in homo-PTPLG50. The narrowing of Tg is also consistent with strong microphase separation in BCPs.

Morphology of helical PPIL BCPs

SAXS was used to explore the self-assembly behavior of helical–helical PPIL BCPs. All samples were hot-pressed into self-standing films at 80 °C with 1.5 tons for 5 min and then annealed at 140 °C for 4 h to enhance the ordering and form stable morphologies (Supplementary Fig. 11). PBLG30-b-PTPLGns all showed well-defined structures (Fig. 3a). Strong microphase separation with long range periodicity was confirmed by multiple reflections. All BCPs exhibited LAM structures, though significant differences in domain spacings (d, the center-to-center distance between adjacent non-ionic or ionic lamellae) and the degree of order varied with PIL mol%. At a fixed N of the PBLG block, the increase of PIL block length led to larger d and more ordered LAM morphologies as expected for increased in N and χN. Notably, PBLG30-b-PTPLG75 displayed up to seven clear reflections at q*, 2q*, 3q* 4q*, 5q*, 6q*, and 7q* (with q* as the primary peak position, q* = 0.019 Å−1), indicating a highly ordered LAM morphology with d = 33 nm (calculated by d = 2π/q*). The LAM structures formed by rod–rod BCP self-assembly can be further classified as single-layer, interdigitated, or bilayer structures, depending on the specific packing arrangement of the helices43,52. To better understand the packing of each helical block within microphase-separated layers, the 18/5 helix model (α-helical conformation of 18 residues in 5 turns) with a helical pitch of 0.54 nm was applied53, where the helical backbone length Lhelix (nm) = N × 0.15 nm. For PBLG30-b-PTPLG75 where d ≈ 2 L, the theoretical calculated length of L = 15.75 nm (LPBLG = 4.5 nm and LPTPLG = 11.25 nm) was compared to the measured d = 33 nm from SAXS corresponding to a bilayer lamella structure with both blocks arranged in a head-to-head bilayer conformation (Fig. 3c). PPIL BCPs with lower PIL mol % exhibited poorly ordered LAM structure likely due to the disrupted balance between the molecular packing of two helical blocks with different cylindrical size, as indicated by the coexistence of two distinct interdigitated LAM with different domain spacings (Supplementary Fig. 12). In PBLG30-b-PTPLG30 (L = 9 nm), one set of scattering peaks at q*, 2q*, and 4q* (q* = 0.049 Å−1) corresponds to a partially interdigitated LAM morphology with d = 12.8 nm. A second set of peaks located between q*–2q* and 2q*–4q* (indicated by arrows) with a ratio of 1:2 (q ~ 0.06 Å−1) suggests another interdigitated LAM structure with d ≈ 10.4 nm, where both helical blocks packed in an interdigitated manner. Similarly, PBLG30-b-PTPLG10 (L = 6 nm) with the shortest PIL block also showed a mixed LAM structure evidenced by two sets of scattering peaks. The dominant LAM structure had a d spacing of 9.5 nm, characterized by peaks at q*, 2q*, and 3q* (q* = 0.066 Å−1). Likely, the PBLG30 blocks were organized into a bilayer structure while the PTPLG10 blocks adopt an interdigitated packing. The shoulder peak (arrow, q = 0.092 Å−1, d = 6.8 nm) suggests the coexistence of another phase with interdigitated packing in both the PBLG and PTPLG domains. The coexistence of two interdigitated LAM morphologies reflects the overall less ordered structure in PBLG30-b-PTPLG30 and PBLG30-b-PTPLG10 compared to PBLG30-b-PTPLG75. Previous studies on rod–rod BCP systems have attributed limited overall LAM order to rod diameter disparity53 or packing periodicity mismatch54, with defects mitigated by increasing volume fraction of large diameter blocks or incorporating β-sheet conformations that are more compatible with LAM packing55. The transition from weakly ordered to highly defined LAM structures in our system aligns with the former explanation, as the ɸPIL of the larger diameter block PIL (Table 1) increased from 43.2% in PBLG30-b-PTPLG10 to 85.1% in PBLG30-b-PTPLG75. Additionally, Banno et al. observed that head-to-head bilayer packing collapses when the overall helical length increased from ~17 to 34 nm in their helical rod−rod polyisocyanide BCPs43. Thus, a specific length of PTPLG block is required in our PPIL BCPs to achieve a highly ordered bilayer structure.

Fig. 3: Microphase separation of helical PPIL BCPs films.
figure 3

a SAXS and b WAXS profiles of PBLG30-b-PTPLGns annealed hot-pressed films, compared to two homopolymers. c Schematic representation of head-to-head bilayer lamella structure of PBLG30-b-PTPLG75. Top: the theoretical polymer length calculated based on a helical pitch of 0.54 nm. Middle: overall block copolymer structure. Bottom: the packing scheme of two blocks in their respective domains. PBLG and PTPLG are indicated by green and red colors, respectively.

Full size image
Table 1 PBLG30-b-PTPLGn physical properties and VFT fit parameters for temperature-dependent ionic conductivities
Full size table

Wide-angle X-ray scattering (WAXS) was employed to investigate both the peptide secondary structure and confinement effects for the ionic groups in bulk films. As shown in Fig. 3b, the characteristic reflections of PTPLG helix–helix correlations were absent, likely resulting from the ionic interactions of ammonium-TFSI ion pairs and complex side-chain packing that disrupt interhelix interactions. The steric effects of bulky ionic groups and anion-anion aggregation weaken the long-range packing of the PTPLG helices. However, reflections with ratios of 1: √3: √4, relative to the primary peak q* = 0.47 Å−1 in a PBLG film indicates 2D hexagonal packing of cylinders composed of 18/5 helices with a cylinder spacing of 13.4 Å. This primary peak was present in all BCPs, suggesting that the PBLG segments were organized in a similar way in the BCPs, which provided the driving force for segregating the PTPLG blocks into LAM structures. The absence of long-range orders in PTPLG likely leads to a more disordered structure within its lamellar domains. The diffraction peak with a spacing of 5.2 Å corresponds to the layer line spacing of the PBLG helix56. Additionally, two other scattering peaks were observed. The highest peak (q1) was assigned as the amorphous halo corresponding to the backbone correlations, while the intermediate peak (q2) reflects the correlation of the TFSI anion57,58. The non-ionic PBLG showed the broadest amorphous peak at q1 ~ 1.44 Å–1 while ionic PTPLG exhibited the narrowest peak with a significant shift to q1 ~ 1.35 Å–1. The intensity of q2 peaks increased with PIL mol% (Supplementary Fig. 13), indicating that higher PIL compositions lead to stronger ion–ion correlations. The q2 peak of PBLG30-b-PTPLG75 (q2 = 0.97 Å–1) was slightly shifted to a larger q value compared to homo-PTPLG (q2 = 0.94 Å–1), possibly due to the highly ordered bilayer LAM structure, which causes TFSI anions to pack more efficiently within the ionic domain. This increased intensity and q2 value enhanced the ionic conductivity as discussed later.

Helical polypeptides can form a range of LC structures59. Polarized optical microscopy (POM) revealed birefringent textures in PBLG30 and PBLG30-b-PTPLGn hot-pressed films (same samples used in SAXS/WAXS measurements), indicating the presence of LC-like ordering (Fig. 4 and Supplementary Fig. 14)49. All PPIL BCPs were heated on the POM to 150 °C, more than 120 °C above their Tgs, and retained the iridescence textures (Supplementary Fig. 15). For the ionic PTPLG block, no iridescence was observed for N = 50 and a longer helical backbone was required to induce ordering which was confined to the surface of hot-pressed films35. The incorporation of a PBLG block enhances the LC-like ordering of PTPLG in BCPs.

Fig. 4: LC-like ordering of PPIL BCPs hot-pressed films.
figure 4

Polarized optical micrographs of helical-helical PBLG30-b-PTPLGn hot-pressed films annealed at 140 °C.

Full size image

Ionic conductivity of PPIL BCPs

Figure 5a presents the temperature-dependent ionic conductivity of PBLG30-b-PTPLGns with varying PIL composition compared to that of homo-PTPLG50 and the non-ionic precursor (Supplementary Fig. 16) over a temperature range of 25 to 140 °C, with a cooling rate of 5 °C/h. A non-Arrhenius conductivity was observed as expected for polymers with tethered ionic groups7,57,60,61,62,63. Consistent with previous studies on PIL BCPs where the ionic conductivity typically increases with higher PIL mol %30,31, PBLG30-b-PTPLG75 with the highest PIL composition (71 mol %) exhibited the highest ionic conductivity. This conductivity is comparable to that of homo-PTPLG, 1–2 orders of magnitude higher than that of PBLG30-b-PTPLG30 (50 mol %), and nearly 2 orders of magnitude higher than PBLG30-b-PTPLG10 (25 mol %). The ionic conductivity of PBLG30-b-PTPLG30 and PBLG30-b-PTPLG10 converge at higher temperatures, while PBLG30-b-PTPLG30 and PBLG30-b-PTPLG75 exhibit similar behavior at lower temperatures. This variation in temperature dependence of ionic conductivity with different PIL mol % is attributed to confinement effects within the block copolymer domains, which influences the morphology of ion aggregates, thereby affecting conductivity64. This correlation is supported by the changes in the position and intensity of the q2 peaks observed in WAXS (Supplementary Fig. 13).

Fig. 5: Higher PIL composition promotes conductivity.
figure 5

a Temperature-dependent ionic conductivity of PBLG30-b-PTPLGns, compared with homopolymer PTPLG50. Solid lines are VFT fits. b Tg-normalized ionic conductivity.

Full size image

To investigate the temperature-related ion conduction behavior of our PPIL BCPs, ionic conductivities at Tg + 10 K and above were fit with the VFT equation (solid line in Fig. 5a):

$${{\rm{\sigma }}}\left({{\rm{T}}}\right)={\sigma }_{{\infty }}\exp \left(-\frac{D{T}_{0}}{T-{T}_{0}}\right)$$
(1)

where \({\sigma }_{{\infty }}\), D, and T0 are the limiting high temperature ionic conductivity before degradation, the strength parameter, and the Vogel temperature where the ionic conductivity hypothetically diverges to zero. Fit parameters are summarized in Table 1. The hypothetical high temperature conductivity (σ∞) of PBLG30-b-PTPLG75 was observed to be more than 20-fold than that of the PBLG30-b-PTPLG10 and PBLG30-b-PTPLG30, corresponding to an increase of the overall ionic content. Notably, all T0 values are consistently lower than their Tgs by around 45 K, which is similar to other PIL systems particularly those with rigid backbones63. The strength parameters (D) is also one factor relating to decoupling of conductivity from segmental dynamics, with lower D generally corresponding to higher conductivities at a given distance from Tg65. All PPIL BCPs exhibited similar D values, with a slight decrease in D with increasing PIL content. To gain further insight into the role of segmental dynamics, Tg-normalized ionic conductivity was also plotted and shown in Fig. 5b. PBLG30-b-PTPLG75 maintained overlap with homo-PTPLG despite a lower overall ionic content. This BCP also showed an ~1.5–2 order of magnitude higher conductivity than PBLG30-b-PTPLG10 and PBLG30-b-PTPLG30. The Tg-normalized ionic conductivity difference between PBLG30-b-PTPLG10 and PBLG30-b-PTPLG30 diminished at higher temperature and merge in a mol%-normalized conductivity indicating the ionic content is controlling this regime (Supplementary Fig. 17). Despite all PPIL BCPs displaying similar Tg - T0 and D values, significant differences in conductivities were observed and attributed to the distinct nanostructures and degree of order. An independent set of all PPIL BCPs were prepared and displayed reproducible conductivity (Supplementary Fig. 18 and Supplementary Table 3).

Interestingly, the ionic conductivity at fixed temperature increases non-linearly with increasing PIL mol %. Specifically, a more than 1.5 order of magnitude increase in ionic conductivity was achieved across the entire temperature range when the PIL composition increased from 50 to 71 mol %. More specifically, the ionic conductivity increased from 0.02 mS/cm to 0.71 mS/cm at 408 K. While increasing PIL content from 25 to 50 mol % led to a difference of 2 to 13-fold with decreasing temperature, with the largest conductivity enhancement of 2.7 × 10−4 to 3.7 × 10−3 µS/cm at 318 K. No discernible conductivity increase can be observed between 71 and 100 PIL mol %. This non-linear correlation between conductivity and composition of ionic block has been noted in other PIL BCP systems as well30,31 and attributed to morphological effects. The ionic conductivity was significantly increased when the morphology changes from 1-D hexagonally packed cylinders to a lamellar structure that can provide 2D pathways to facilitate ion transport. Conductivity was further increased in morphologies with percolated 3D pathways.

The PPIL BCPs were further analyzed using the morphology factor (f) to normalize the ion conductivity (Fig. 6):

$$f=\frac{\sigma }{{{{\phi }}_{c}\sigma }_{c}}$$
(2)

where \(\sigma\) is the measured ionic conductivity of PIL BCPs, ɸc, and \({\sigma }_{c}\) are the volume faction and intrinsic ionic conductivity of the pure PIL block, respectively. Typically, theoretical \(f\) values (\({f}_{{ideal}}\)) for PIL BCP solid electrolytes are 1/3, 2/3, and 1, corresponding to randomly oriented domains of 1-D hexagonally packed cylinders, 2D lamella, and 3D networks, respectively. However, experimental values often fall short of these values30,66,67 and are generally attributed factors like grain boundaries, grain orientations or interfacial dead zones68,69. Some experimental values exceed \({f}_{{ideal}}\), such as with added nanoparticles70 or using electrostatic interaction33, and can also exceed 1 for some BCP compositions64.

Fig. 6: Highly ordered bilayer lamellar structure enables higher morphology factor.
figure 6

Morphology factor f of PBLG30-b-PTPLGns with different ionic block composition as a function of temperature.

Full size image

Figure 6 illustrates the morphology factor for PPIL BCPs with PIL mol% = 25, 50, and 71, corresponding to ɸPIL = 43.2, 69.5, and 85.1, respectively (Table 1). Density of 1.28 and 1.51 g/cm3 for PBLG (from the literature52) and PTPLG (determined by pycnometer), respectively, were used to calculate the volume fractions. The \(f\) values of PBLG30-b-PTPLG10 ranges from 0.005 to 0.03 and PBLG30-b-PTPLG30 ranges from 0.03 to 0.05 with temperature, all significantly lower than fLAM, ideal = 2/3. This is likely due to the grain boundaries, coexistence of two LAM phases and poorly-ordered LAM structures as evidenced by the shoulder peak in SAXS. Interestingly, PBLG30-b-PTPLG75 gave an unexpectedly high \(f\) values (\(f\) > 0.8 across the entire temperature range), surpassing even \({f}_{{LAM},{ideal}}\), which was attributed to the combination of higher PIL composition (71 mol %) and the formation of highly ordered bilayer LAM structures. The highly ordered ionic domains and closely packed TFSI anions in this structure create a favorable environment for ion transport with a continuous pathway, resulting in high conductivity and morphology factor. This enhancement in conductivity may be due to the alignment of peptides and the macrodipoles along the interfaces, which can enhance the local dielectric constant leading to greater ion pair dissociation and increased TFSI mobility. The helices will also have a consistent spacing of tethered cations away from the interface which can facilitate TFSI motion between helices. Additionally, the high morphology factors may also be a result of some orientation induced during processing. Future work quantifying how processing or controlled shear alignment can orient the diblocks would be helpful to understand the upper limits of conductivity in peptide–peptide systems. When normalized by ionic content, this BCP is more conductive than the corresponding homopolymer, indicating more favorable ion transport within the nanostructures. The increase in N is also a factor that enhances the strength of microphase separation and increases the normalized conductivity27. Moreover, \(f\) values substantially increase as the temperature decreasing, reaching 1.3 at 40 °C. This indicates that the bilayer LAM structure enhances ion transport at lower temperatures compared to the homopolymer, offering creative insights for the design of future high-performance polymer electrolytes.

Solid-state helical–helical polypeptide polymerized ionic liquid block copolymers were investigated to determine the role of secondary structure on assembly and conductivity. PBLG30-b-PTPLGns with fixed non-ionic PBLG lengths but varying ionic PTPLG length were synthesized and analyzed. The helical conformation and thermal properties were characterized by ATR-FTIR and DSC, respectively. SAXS analysis confirmed strong microphase separation in these PPIL BCPs, which self-assembled into LAM structures with varying d and degrees of order, where higher PIL mol % (and largest N) led to larger d and a higher degree of order. PBLG30-b-PTPLG75 with the longest PIL block formed a highly ordered bilayer LAM structure and exhibited a ~ 1.5–2 order of magnitude higher conductivity than PBLG30-b-PTPLG30 and PBLG30-b-PTPLG10 with poorly order LAM structure. More than 20-fold increase in σ∞ of PBLG30-b-PTPLG75 indicates the enhancement of conductivity will continue to grow upon heating up to the point of degradation. These results demonstrate the critical role of highly ordered nanoscale ionic domains and connectivity in enhancing ion transport. This is further supported by f > 0.8 of PBLG30-b-PTPLG75, which exceeds the ideal value for LAM phases (\({f}_{{LAM},{ideal}}\) = 2/3), while PBLG30-b-PTPLG30 and PBLG30-b-PTPLG10 have \(f\) < 0.05. These insights into the structure-property relationships of helical PPIL BCPs provide a direction for optimizing ion transport and tailoring the properties of block copolymer systems using helical backbone polymers as next-generation stable, high-performance ion-conducting electrolytes. For applications in energy storage, ions, such as Li or OH will be important and understanding how other charges are impacted by the macrodipole of a helix, block copolymer assembly, and solvent will be critical for advancing peptide electrolytes.

Methods

All detailed experimental procedures and characterization of compounds can be found in the Supplementary Information.

Synthesis of poly(γ-benzyl-L-glutamate)-b-poly(γ-(3-chloropropanyl)-L-glutamate) (PBLG30-b-PCIPLGns)

Block copolypeptides PBLG30-b-PCIPLGns were synthesized with PBLG30-NH2 as initiator in an emulsion system (VDCM/VpH=7 phosphate buffer = 95 %/5%) using 18-Crown-6 as catalyst. Such as for PBLG30-b-PCIPLG30, 4 mM 18-Crown-6 DCM solution (12 μL, 0.048 mmol, 2 eq), pH = 7 phosphate buffer (0.66 mL), and 4 mM PBLG30-NH2 DCM solution (6 mL, 0.024 mmol, 1 eq) were mixed together, in which 100 mM ClPLG-NCA DCM solution (7.2 mL, 0.72 mmol, 30 eq) was added and then vigorously vortexed. The mixture was allowed to stir at room temperature for 24 h. After reaction, most of the water was removed via extraction from reaction mixture, that was then dried by using anhydrous sodium sulfate. After filtration, the organic phase was concentrated to 1–2 mL and then precipitated in cold hexane/ether (v/v = 1/1) twice to yield white solid (270 mg, 88%).

Morphology characterization

SAXS and WAXS experiments were conducted simultaneously using the 16ID-LiX Beamline at the National Synchrotron Light Source II where is located at the Brookhaven National Laboratory (Upton, New York). The X-ray energy was 13.5 keV. The intensity is expressed as a function of scattering vector, q defined as \(\frac{4{{\rm{\pi }}}}{\lambda }\sin \frac{\theta }{2}\), where θ is the scattering angle and λ is the wavelength. The data covers a q range from 0.005 to 2.5 Å−1. All samples were hot-pressed into self-standing films at 80 °C with 1.5 tons for 5 min and then annealed at 140 °C for 4 h.

Ionic conductivity measurement

Electrochemical impedance spectra were recorded on a SP-300 Potentiostat with an intermediate temperature system. All samples’ preparations were conducted in a glovebox. Pre-dried polymer was placed in a hole (with a diameter 5 mm) of a Kapton spacer between two stainless-steel electrodes (with a thickness 0.5 mm), heated to 80 °C, and then pressed into a film and held for 2 h. The hot-pressed polymer film thickness was determined by the average of three measurements with a micrometer. The cell was hermetically sealed by a crimper in the glovebox using a typical coin cell assembly. The coin cell was first heated to 140 °C and equilibrated at least 4 h. The ionic conductivity over time was collected from high temperature to low temperature with a 5 °C temperature interval and a 60-min equilibration time was applied before each measurement. The applied voltage was 100 mV, and the frequency range was 106 to 0.1 Hz. The impedance data was processed to make a plot of real (σ’) and imaginary (σ”) conductivities versus frequency, and the ionic conductivity of polymer electrolyte was taken as the real conductivity at the tan(δ) = σ’/ σ” maximum. Each polymer electrolyte was measured with two independent samples.