Functionalization of Multi-Walled Carbon Nanotubes (MWNTs) for Sulfonated Polyether Ether Ketone (SPEEK)/MWNT Composite Elaboration ^†

Abstract

In this study, we present the covalent functionalization of pristine multi-walled carbon nanotubes (P-MWNTs) with sulphonate poly(ether ether ketone) (SPEEK) chains, employing hexane diamine as an interlinking molecule. SPEEK-functionalized MWNTs were then used to create SPEEK-MWNT/SPEEK composites. We used FTIR spectroscopy to confirm the covalent attachment of the SPEEK chains to the MWNTs. XRD and TEM were used to characterize the morphology of the functionalized tubes and composites. We then evaluated the composite membranes for their structure and elastic modulus. Our results show that SPEEK-grafted MWNT composites with 2% and 5 wt% SPEEK-MWNTs exhibited a 7.1% and 16.1% improvement in Young’s modulus, respectively, compared to SPEEK. However, oxidized MWNT/SPEEK membranes exhibited slightly better improvement.

1. Introduction

Sulfonated poly(ether ether ketone) (SPEEK) has emerged as a leading hydrocarbon-based alternative to fluorinated proton exchange membranes due to its high thermal and mechanical stability, tunable degree of sulfonation, and relatively low cost [1]. However, excessive swelling at high sulfonation degrees compromises its dimensional stability and mechanical integrity. To overcome these drawbacks, the incorporation of nanofillers such as carbon nanotubes (CNTs) has gained substantial attention [2].

CNTs possess exceptional mechanical strength, high electrical conductivity, and high aspect ratios, enabling them to act as both reinforcing and transport-modulating agents within polymer matrices [3]. When judiciously dispersed in SPEEK, CNTs can enhance proton conductivity through the formation of interconnected hydrophilic pathways while simultaneously improving tensile strength and thermal stability [1]. Functionalization of CNTs with –COOH, –OH, or grafted organic chains has been particularly effective in improving dispersion and interfacial compatibility [3]. Recent studies have demonstrated that optimized CNT loadings (typically <1 wt%) significantly improve SPEEK membrane performance in electrochemical devices. For instance, CNT/SPEEK composites achieved nearly two-fold higher power density in microbial fuel cells (MFCs) [4] and improved coulombic efficiency and stability in vanadium redox flow batteries (VRFBs) [5]. These enhancements arise from synergistic effects between CNT-reinforced mechanical frameworks and proton-conducting channels formed along the SPEEK sulfonic acid network.

Despite these advances, challenges remain in achieving uniform CNT dispersion, preventing electronic leakage, and maintaining long-term stability under fuel cell operating conditions [1,2]. Current research thus focuses on advanced functionalization, scalable fabrication methods, and hybrid approaches combining CNTs with other nanofillers to optimize the trade-off between conductivity, selectivity, and mechanical robustness.

In this study, functionalized carbon MWCNTs were used as a filler material to create a SPEEK composite membrane. Both oxidized and SPEEK-functionalized MWNTs were incorporated into the SPEEK matrix at different weight percentages (2% and 5%) to examine the effects of the surface functionalization, as well as the filler wt%, on the Young’s modulus of the composite membranes.

2. Materials and Methods

MWNTs (Nanocyl, 4–18 nm) were oxidized in HNO₃ (70%) to generate carboxylated surfaces. SPEEK was synthesized by sulfonating PEEK in concentrated H₂SO₄ (98%_ (DS ≈ 70%). Functionalization proceeded via amidation of oxidized MWNTs with HDA in DMF and DCC, followed by coupling with SPEEK following the method described in [6]. Composite membranes were synthesized using the solution casting method described elsewhere [7].

Samples were characterized by FTIR (using a BRUKER IFS 66V spectrometer equipped with a N₂-cooled MCT (Mercury Cadmium Telluride) detector, L2C, Montpellier, France) and TEM (JEOL 1200 EX II, Lorraine University, Nancy, France). X-ray diffraction (XRD) experiments were conducted using an INEL CPS120 powder diffractometer equipped with a 120° curved detector, a Cu Kα x-ray source (λ = 1.5418 Å), and a germanium monochromator, with a step size of 0.03° (2θ) and a Q range from 0.14 to 7 Å⁻¹. All samples were finely ground, and the resulting curves were corrected for background signals from the glass and air. Film tensile properties were assessed using a LLOYD LR 10k Universal Testing Machine. Young’s Modulus values for each film were determined following ASTM D-638, with a crosshead speed of 2 mm/min.

3. Results and Discussion

3.1. FTIR Analysis

FTIR spectra recorded after each functionalization step (Figure 1) show systematic chemical changes consistent with covalent grafting. Oxidation of pristine MWNTs (P-MWNTs) produced a pronounced C=O stretching band at ~1733 cm⁻¹ and enhanced C–O features (1000–1200 cm⁻¹) (OX-MWNTs), confirming carboxylation of defect sites. Reaction with hexane diamine (HAD-MWNTs) introduced new absorptions in the C–N/amide region (~1490–1550 cm⁻¹) and attenuated the carbonyl band, consistent with amidation of surface carboxyls.

The band around 1620 cm⁻¹ (indicated by the arrow) in Figure 2a, is characteristic of the aromatic carbon backbone in MWNTs (analogous to the “G-band” in Raman spectra). This confirms the presence of conjugated sp² carbons from nanotubes.

Subsequent coupling with SPEEK (SPEEK-MWNTs sample) produced a marked decrease in the SPEEK –SO₃H bands (1320, 1082, 1034, and 930 cm⁻¹), indicating involvement of sulfonic groups and nitrogen functionalities in bond formation. The composite spectra differ significantly from the arithmetic sum of the parent MWNT and SPEEK spectra, which argues against mere physical mixing. These features, combined with slight shifts or reduced peak intensity in the sulfonic acid (–SO₃H, ~1030–1080 cm⁻¹) and ether (C–O–C, ~1200 cm⁻¹) regions, support covalent attachment or strong interfacial interactions between SPEEK chains and oxidized MWNTs.

3.2. TEM Analysis

Figure 3 shows TEM micrographs for before and after SPEEK functionalization. The P-MWNTs (left) are clean, smooth-walled, and distinct, with well-resolved hollow cores and graphitic fringes. The tubes appear loosely entangled but individually separated. After grafting with SPEEK chains, the nanotubes appear significantly darker, thicker, and more irregular in contour shape compared with the pristine sample. Amorphous coatings or sheath-like layers (2–5 nm thick) surrounding many tubes are visible as diffuse halos or uneven contrast along the tube lengths. Inter-tube junctions are bridged by amorphous material, suggesting the formation of a polymer network connecting multiple MWNTs. Additionally, the local contrast is non-uniform, and darker bundles and overlapping polymer clusters are visible, consistent with partial wrapping or deposition of a polymeric phase. Some agglomeration or entanglement is apparent, but tubes are embedded in an amorphous matrix rather than simply bundled together.

3.3. XRD Analysis

The XRD diagrams in Figure 4 illustrate the structural differences between non-covalently mixed MWNT/SPEEK composites (left) and covalently linked SPEEK-MWNT/SPEEK hybrids (right).

The MWNT/SPEEK composite (green curve) closely resembles a simple superposition of oxidized MWNTs (black) and pure SPEEK (pink) signals. The “Sum (MWNTs + SPEEK)” (orange) curve, normalized to the MWNT peak at 1.8 Å⁻¹ to match the maximum intensity of the composite, nearly overlaps the measured composite curve, indicating minimal structural interaction between the two components. The main MWNT (002) reflection at q ≈ 1.8 Å⁻¹ remains sharp and strong. The broad SPEEK halo persists at ~1.5 Å⁻¹, showing no significant shift or suppression. The preserved intensity and position of MWNT peaks suggest that the polymer does not intercalate or strain the nanotube walls. The broad SPEEK signal indicates unchanged amorphous character, meaning no induced crystallinity or ordered interphase develops. The non-covalent composite behaves as a physical mixture, with limited interfacial stress transfer and distinct structural domains of CNTs and polymer.

On the other hand, for the covalent attachment (right), the MWNT/SPEEK (black) pattern deviates substantially from the simple arithmetic sum (orange) of MWNT + SPEEK. A noticeable decrease in the MWNT (002) peak intensity at ~1.8 Å⁻¹ occurs, along with slight broadening, evidence of distortion in the graphitic planes. The “difference” curve (gray) highlights this intensity deficit relative to the simple sum. The SPEEK pattern (blue) still shows its broad amorphous halo, but minor shifts toward lower q (≈ 1.5 Å⁻¹) appear, consistent with expanded interchain spacing due to covalent bonding and strain. A faint low-q shoulder (<1 Å⁻¹) emerges, suggesting formation of larger-scale heterogeneities or interfacial domains. The chemical linkage (via amine bridging) partially disrupts the CNT graphitic lattice, decreasing order within the outer walls (reflected in peak broadening and reduced intensity).

3.4. Young’s Modulus Measurements

Both the oxidized MWNT-based fillers and the SPEEK-grafted MWNT fillers significantly stiffen SPEEK in the dry state. As shown in Figure 5, at 2 wt% the OX-MWNT/SPEEK composite exhibits a larger gain (+15.4%) than the covalently grafted SPEEK-MWNT/SPEEK (+7.1%), while at 5 wt%, both systems reach similar stiffnesses (+18.9% and +16.1%, respectively). Statistical comparison shows the 2 wt% OX-MWNT/SPEEK composite is significantly stiffer than the 2 wt% SPEEK-MWNT/SPEEK composite, whereas for the 5 wt% load, the difference is not significant. These results indicate a trade-off: covalent grafting improves interfacial bonding and promotes network formation at higher loadings but introduces a polymer-rich interphase and surface disorder that reduces reinforcement efficiency at low loadings. Optimizing graft density or linker stiffness could recover low loading performance while preserving interfacial stability.

Our finding that oxidized (non-grafted) MWNTs give a larger modulus increase at 2 wt% than covalently grafted SPEEK–MWNTs fits the general trade-off reported in the literature between (i) maintaining high intrinsic stiffness of the nanotube (minimizing lattice damage) and (ii) improving interfacial bonding/compatibility. Several recent reviews and experimental studies emphasize that covalent functionalization can (a) improve dispersion and interfacial adhesion but also (b) introduce defects or create a polymer-rich interphase that reduces the effective stiffness contribution per tube, particularly at low filler fractions where interphase volume is a larger fraction of the composite [1]. At higher loading (5 wt%), the two systems converge, which we interpret as the onset of a filler-network/percolation effect: once the volume fraction is sufficient, the sheer number of load-bearing nanotubes and inter-tube bridging (whether physical or covalent) compensates for interphase softness or tube damage, producing similar bulk stiffness for both treatments. Observations of load fraction thresholds and improved mechanical response when fillers are better distributed or aligned have been reported for SPEEK-based and other polymer/nanotube membranes [8].

4. Conclusions

This study successfully demonstrated the covalent functionalization of multi-walled carbon nanotubes (MWNTs) with sulfonated poly(ether ether ketone) (SPEEK) chains via hexamethylene diamine bridging. FTIR confirmed the formation of amide and sulfonamide linkages between oxidized MWNTs and SPEEK, while TEM revealed polymer coatings around nanotubes and inter-tube bridging indicative of a covalently connected hybrid network. XRD analysis showed a marked reduction in MWNT (002) peak intensity and broadening in covalently bonded composites, evidencing partial disruption of graphitic ordering and polymer-induced lattice strain. Mechanical testing showed that both oxidized and SPEEK-grafted MWNTs enhance the stiffness of SPEEK, with 2 wt% and 5 wt% oxidized MWNT/SPEEK composites showing +15.4% and +18.9% improvements in Young’s modulus, respectively, compared to +7.1% and +16.1% for the corresponding SPEEK-MWNT/SPEEK composites. These results highlight a trade-off between interfacial adhesion and nanotube structural preservation: covalent grafting improves compatibility and network formation but slightly reduces reinforcement efficiency at low loadings. Future work should focus on optimizing the linker chemistry and grafting density to balance stiffness enhancement with interfacial stability, as well as on correlating interphase structure with mechanical and transport performance in electrochemical operating environments.