The D-band is positioned at 1340.50 cm⁻¹, while the G-band is placed at 1583.00 cm⁻¹. A small I_D/I_G peak value (i.e., 0.94) suggests a high purity while a high I_G/I_D peak value (i.e., 1.06) shows a high degree of wall graphitization and the type of CNTs is multiwalled CNTs [8]_ENREF_15. The produced CNTs displayed a 2D band at 2890 cm⁻¹, and the I_2D/I_G ratio was 0.2, confirming the presence of multi-layer G sheets (Fig. 3a) [17,37].

TEM analysis for the prepared CNTs showed short bundles of CNTs bundles with diameter range ~7.5–13.5 and length range ~103.8–252.2 nm (Fig. 3b,c).

Fig. 4 reveals FT-IR spectrum of CNTs’ peaks which centered at 3437, 2932, 1634, 1427, 1217 and 1058 cm⁻¹ which attributed to stretching vibration modes of OH, CH, C=O, O-C=O, C–O and C–O–C groups, respectively [17]. While, R1 shows peaks at 3299, 2925, 1631, 1548, 1448 and 1027 cm⁻¹ which attributed to stretching of OH, CH, amide I of Gel (i.e., C=O & C–N), amide II of gelatin (N-H bending), C–O of cellulosic paper and C–O–C groups of cellulosic paper, respectively. R2 and R3 show peaks between 3278–3292, 2919–2921, 1627–1629, 1542–1546, 1434–1446 and 1025–1027 cm⁻¹ related to stretching of OH, CH, C=O of CNTs & amide I of gelatin (i.e., C=O & C–N), O–C=O of CNTs and amide II of gelatin (N–H bending), C–O and C–O–C groups, respectively [25,43]. H-bonded OH stretching vibration of R1 appeared at 3288 cm⁻¹ was shifted to lower frequency at 3284 and 3274 cm⁻¹ for R2 and R3, respectively. Intermolecular H-bonding may be the cause of the shift in the absorbance value of the OH stretching vibration [44]. Decreasing C=O group intensity in R2 and R3 compared to CNTs confirmed the successful overlapping between amide of gelatin and carboxylate groups of CNTs [36]. The MHBS of CNTs, R1, R2 and R3 are summarized in Table 2. The C=O peaks for CNTs and R1 at 1634 and 1631 was shifted to lower wave numbers at 1629 and 1627 cm⁻¹ for R2 and R3, respectively. This may be due to the interaction between C=O of CNTs with the amide or OH of WPP and Gel [45]. The ash (%) of R3 is the lowest one compared to others. This may be due to the high content of CNTs.

The prepared CNTs are composed of few CNTs and graphene (G) nano-sheets. The G nano-sheets have an amorphous peak at 2θ = 9.3 and 30.2° related to (001) and (002) diffraction peaks (Fig. 5a) [17]. The R1 exhibits four obvious diffraction peaks at 2θ = 11.55, 22.69, 29.40 and 35.83°, related to the (1–10), (110), (200), and (004) planes of cellulose. The triple helix of gelatin’s triple peak is based on its inter-helix distance and the broad diffraction peak at 2θ = 22.69° is indicative of the amorphous portion of gelatin (Fig. 5b). When gelatin is attached to WPP, the diffraction peaks of gelatin might coincide with those of cellulose, but the diffraction patterns do not show any of the cellulose peaks [21].

The d-spacing of R1 (3.91 nm) is increased compared to R2 and R3 (3.06 and 3.00 nm). The increased d-spacing in R1 was caused by the free, unmodified functional groups inserting themselves between the cellulose layers in the WPP. In contrast, the free functional groups in R2 and R3 were connected or linked to the incorporated CNTs [44]. While the D value reaches its peak in pure CNTs, it decreases as the CNT content in R3 increases. The variations in σ and ε exhibit contrasting trends. The combination of the lowest D value, highest σ, and highest ε in R3 (containing 0.02 g CNTs) can likely be attributed to the significant variability in d-spacing [46]. The Cr. l (%) for CNTs, R1, R2 and R3 were 14.63%, 14.83%, 18.75% and 31.32%, respectively (Table 3). The Cr. l (%) content of the modified sheets (R2 and R3) was notably higher than that of R1 (Fig. 5b–d). This increase can likely be attributed to the elevated concentration of CNTs, which influenced R1. The IR spectra and XRD Cr. 1 (%) values align with these findings.

Fig. 6 shows the surface morphology of CNTs, R1, R2 and R3, and how the modification process affected on R1 surface morphology. SEM analysis revealed that CNTs formed short, bundled structures. EDX analysis identified impurities including iron, cobalt, nickel, and molybdenum (Table 4). R1 exhibited a smooth surface without CNT coating, as depicted in Fig. 6b. In contrast, Fig. 6c,d shows CNTs covering the surfaces of R2 and R3 sheets, resulting in a rougher texture. The porosity of the films, calculated from SEM images and presented in Table 5, increased with higher CNT content. R2, with the highest CNTs concentration, displayed the highest porosity (69.10%) compared to R1 (68.80%) and R3 (70.44%). This increase in porosity can be attributed to the higher oxygen content associated with the increased CNTs.

TGA/DTG analysis at 1000°C showed that R1, R2, and R3 retained 48.58%, 38.90%, and 46.94% of their original mass, respectively, indicating the presence of non-volatile components. The R1’s decomposition curve revealed two decomposition steps, whereas R2 and R3’s decomposition curves revealed three steps. Thus, it can be concluded that the decomposition mechanisms of the untreated R1 and treated R2 & R3 sheets are different. The data from TGA/DTG, such as temperature at maximum rate of weight loss (T_max) and char yields at 300°C (C_Y300) and 700°C (C_Y700), are summarized in Table 6.

The thermal degradation processes of R1 could be separated into two key reaction phases. The 1st weight loss occurred between 41.84°C–212.41°C, with a maximum of 92.14°C, and mass loss (ML%) of 6.73% was probably brought on by the loss of moisture content [26,36]. The 2nd main decomposition step was between 210.86°C–984.31°C, with a maximum of 346.01°C, ∑E 12.15 kJ/mole and ML was 44.69%, ascribed to the depolymerizations and carbonization (Fig. 7a) [8,23]. The R2 and R3 have 1st decomposition steps at 41.94–150.04 and 41.88°C–205.83°C, with a maximum of 96.02°C and 97.63°C and ML was 4.51% and 7.95%, respectively, and ascribed to the moisture content. With an increase of the CNTs concentration, the initial thermal decomposition temperature of R2 and R3 increased compared to R1 due to the improved thermal stability by CNTs addition. The 2nd and 3rd main decomposition steps occur at temperatures of 185.36–465.63 & 207.66°C–483.35°C and 641.15–762.06 & 609.85°C–721.55°C, respectively [17,44].

Once the WPP was treated with CNTs, the C_Y300 enriched by 103.17% and 107.90% for R2 and R3, respectively. Likewise, CY700 enriched by 105.32% and 105.38% for R2 and R3, respectively. From the results, treatment with CNTs decreased the amount of volatile products, formed during pyrolysis and diminished ML due to carbon combustion of R2 and R3 sheets compared to R1. Consequently, we speculated that the treatment with CNTs supports the production of carbonic char and retarded the degradation of cellulosic backbone of the recycled paper [17,38]. The main stage of R3 has a higher ∑E value (33.64 kJ mol⁻¹) compared to the ∑E values for the main stage of R1 (12.15 kJ mol⁻¹) and R2 (27.09 kJ mol⁻¹), which suggests that R2 and R3 has undergone thermal stabilization compared to R1 [38,45]. This is a result of the blends’ heterogeneity, which prevents heat energy from moving easily through the material. This demonstrates that the blends R2 and R3 have greater thermal stability than the neat R1 [46,47]. As a result, the heat reaction’s course was affected by the CNTs present in the cellulose backbone in the recycled paper, which is what caused the E_a to vary. The information provided by MHBS and E_a agreed with each other. However, there was no consistency with the LO values; this may be because of the kind and strength of the created H–bonding [8]. This agree with DFT calculations from the calculations of the E_g is lower in the case of WPP/CNTs/Gel (i.e., 0.078 eV). A lowered E_g promotes the formation of a char layer. Char is a carbonaceous material that forms when organic matter decomposes under high temperatures. Char layers can act as a barrier to heat and oxygen, helping to prevent the spread of flames and reduce the rate of combustion. which in turn proves the strong chemical reaction between WPP, gelatin and CNTs.

According to Table 6, the values for R1, R2, and R3’s ∆S are negative, indicating that the system degradation won’t happen naturally [17,36]. Additionally, as the reaction progresses, the value of ∆H for the R3 falls from between 8.71 and 18.20 kJ mol⁻¹ at the beginning of the reaction to between 2.42 kJ mol⁻¹ at the end, indicating a relatively low energy need (i.e., E_a was 11.68, 23.35 kJ mol⁻¹ for the 1st and 2nd stages then decreased to 10.29 kJ mol⁻¹ for the final stage). For R1, it was found that with lowering n (0.5) value, the Δ H was decreased (7.00 kJ mol⁻¹). This suggests that the pure R1 may decay with only a small amount of energy [17].

The noticeable drop in Δ H for the pure R1 can be attributed to the existence of free hydroxyl and carboxyl terminal groups, which are regarded as excellent leaving groups. The R1’s bifunctionality (i.e., the existence of both COOH and OH groups), which has a high possibility of crosslinking, also leads to a drop in Δ H. for the gelatin-and CNT–containing WPP. The Δ G of R1, R2, and R3 are 286.38, 486.33, and 530.40 kJ mol⁻¹. The non-spontaneous reaction is indicated by the highest Δ G value, which is found in R3. The Δ G states that R3 requires more external heat input than R1 and R2 (Fig. 7b).

The YM, EB, and TS are shown in Table 7 at maximum R1, R2, and R3 load. The results show that at 10% and 20% CNT loading, the addition of CNTs enhanced the maximal TS (by 16.93% and 283%) for R2 and R3. More pores produced at the surface and pore deficiencies throughout the thickness may be the cause of the significant rise in TS for R3 (it is also proved by SEM analysis). The substantial standard deviation values at 20% CNT loading (R3) can be attributed to the films’ high porosity [17]. The addition of CNTs, however, clearly reduced the YM% of R1, R2, and R3, particularly at a 20% loading of CNT (R3), where the reduction in YM% was 33% and 71% for R2 and R3, respectively (Fig. 8). As CNT loading was increased to higher levels, the YM values declined. This might be because larger CNTs concentrations result in enhanced stiffness properties. The decreasing YM of R3 provided evidence that CNTs were being used for nano-reinforcement [48,49].

As CNT loadings were increased to greater levels, the EB fell; for R3, it was roughly 33%, while it was still higher than that of the pristine R1. The EB of R2 increased up to 73%. The improvement in the films’ mechanical properties suggests that the recycled paper matrix and CNTs surfaces are highly compatible. Data from TS, E_a, MHBS, and CrI% were all in agreement. This shows how the stiffness of the sheets formed is impacted by CNTs. When comparing the increase in TS and EB reached in the current work to the results previously reported, the advantage of using CNTs isolated from MB by direct hydrothermal treatment is obvious.

The flame retardancy test under continues exposure to the flame is performed on the R1, R2, and R3 sheets. An illustration of the burning process is shown in Fig. 9. All paper sheets of uniform thickness are simultaneously ignited by a gas flame while resting on a metal support. The R1 sheet burns quickly, totally converting to ash in 6 s, whereas the R2 sheet takes 8 s to burn before changing to charcoal, demonstrating how the CNTs improve the sheet’s flame retardancy (Fig. 9). On the other hand, the R3 sheet’s burn time was extended up to 60 s. because it was the last to ignite. This is because R3 cannot thermally decompose in the presence of CNTs. It is predicted that the flame retardancy of the nanocomposites is significantly influenced by the CNTs network topologies on the WPP. The flame retardancy is appealing for use in industrial applications including packaging.

To assess the flame resistance of materials, the vertical burning test (UL94) and limited oxygen index (LOI) are widely used. The vertical burning test measures how easily a material ignites and how quickly flames spread. A rating of 0 indicates the highest level of flame resistance, while a rating of 5 suggests the lowest. Ratings between 1 and 5 mean that the material may not fully ignite, even if flames spread upward. The LOI determines the minimum oxygen concentration required for a material to burn in a mixture of oxygen and nitrogen.

The CNTs with varying concentrations, 0–0.02 gm (wt/wt) was used to enhance the thermal proofing properties of R2 and R3 sheets compared to R1. The vertical burning and LOI values of R1, R2 and R3 are presented in Table 8. When R3 exposed to flames, a protective carbon layer forms on the surface. This layer, strengthened by the presence of CNTs, acts as a shield, hindering the spread of flames and improving the material’s fire resistance [32].

From LOI results, we can observe that the LOI values for the treated recycled paper sheets increase with an increasing CNTs content from 0 to 0.02 gm (wt/wt). We can conclude that when the CNTs content increase, the flame retardancy increase. And since it is known that in order to achieve a level of flameproofing, the LOI value has to exceed 21 vol. % [21]. Accordingly, the larger LOI value for R3 corresponds to the better flame retardancy performance, whereas R1 and R2 sheets are combustible. The burning process of R1, R2 and R3 sheets at a 3 s ignition period in ambient air is shown in Fig. 10.