Effect of the Surface modification of Cellulose nanofibers on the Mechanical Properties and Disintegrability of Specific PLA/Cellulose Composites

Chloroform of analytical grade was purchased from POCh, Poland as a solvent for the preparation of PLA solution and NFC suspensions.

Poly(lactic acid) (PLA) in the form of pellets was supplied by NatureWorks® LLC as PLA Ingeo™ 6201D polymer. Selected physicochemical properties of PLA 6201D according to the manufacturer’s data are summarized in Table 1 [32]. The chemical purity of this material was verified in a study by Jóźwicka et al. [33].

Cellulose nanofibres (NFC) were prepared at Łukasiewicz – Lodz Institute of Technology, Łódź, Poland. NFC was extracted from pulp derived from hemp stalks. The crude pulp was pretreated using cellulolytic enzymes (endoglucanases) to weaken interactions between individual microfibrils. Enzymatic treatment promotes the mechanical breakdown of pulp fibers, reducing the amount of energy needed for this process. Next, the aqueous suspension of the pulp was subjected to high mechanical shearing in a CAT, X1740 homogenizing system using a DK40 flow-through chamber equipped with a cooling system. In the last step, the NFC was isolated from the slurry by centrifugation (Hettich, Rotina 420) at 4500 rpm and partially concentrated on a rotary evaporator (Büchi, Rotavapor R-300).

Partially saponified fatty acid methyl esters of castor oil (ZS1) were supplied by the Łukasiewicz-Institute of Heavy Organic Synthesis, Kędzierzyn-Koźle, Poland. Castor oil is a mixture of (tri, di-, mono-) glycerides of fatty acids, i.e. ricinoleic acid (90%), oleic acid, linoleic acid and others. Methyl esters are usually obtained in the process of transesterification of oils with methyl alcohol in the presence of an alkaline catalyst. ZS1 belongs to the group of anionic surfactants (Figure 2).

Fig. 2.

Cetyltrimethylammonium bromide (CTAB) was purchased from J&K Scientific, USA. CTAB, a cationic surfactant, like others, forms micelles in aqueous solutions (Figure 3)

Fig. 3.

Sucrose palmitate RYOTO™ P-1670 was supplied by Mitsubishi Chemical Foods Co. Sucrose esters are non-ionic carbohydrate surfactants obtained by the esterification of sucrose with fatty acids. The sucrose moiety is polar and acts as the hydrophilic end of the molecule, while the long fatty acid chain is its lipophilic end. Due to this amphipathic property, sucrose esters act as universal emulsifiers (Figure 4).

Fig. 4.

Compost “Próchniaczek” used for the disintegrability test was supplied by the Waste Management Board in Lodz, Poland. It is obtained from the selective organic fraction of municipal waste as a result of the aerobic decomposition of biomass by the action of microorganisms.

The morphology of the modified cellulose particles was characterized using a FESEM scanning electron microscope (FEI, Quanta 200) at an electron beam accelerating voltage of 10 kV. Prior to SEM analysis, the samples were gold sputtered using a Q150R vacuum coater from Quorum Technologies.

The structural changes of NFC/surfactants and characteristics of the inert structure of polymeric composite foils were specified by X-ray diffraction (XRD) analysis. PXRD experiments were registered on a Panalytical Empyrean powder X-Ray diffractometer using copper radiation (λ = 1.5419 Å) and zero-background holders. The 2Θ range in each case was 5-45°.

The UV-vis transmittance of the PLA NFC composite films prepared was examined from 200 to 800 nm by a Thermo Spectronic, Helios γ spectrophotometer.

TGA and DSC analyses were performed to evaluate the thermal properties of PLA composite films using a Setaram, TG Labsys Evo 1150 instrument with “Calisto” AKTS control and processing software. Analysis parameters: atmosphere - oxygen-free nitrogen, flow 20 cm³/min; vessel - Pt 100 μl; heating rate – 10 K/min; temperature range - 20 – 200°C

Tensile tests of pure PLA and PLA NFC composite films were carried out under standard environmental conditions (temperature 20±2°C and relative humidity 65±4 %). The following parameters were estimated: tensile strength and elongation at break. The measurements were made using an Instron 5544 tensile tester. Three specimens were tested per each sample. Specimen measuring section length: 50 mm, stretching speed: 100 mm/min, sample width: 15 mm.

A disintegration test was carried out following the PN-EN ISO 20200:2007 standard by the accredited Laboratory of Biodegradation and Microbiological Research at Łukasiewicz – Łódź Institute of Technology, Poland.

NFC was mixed with three different compatibilizers at a ratio of 9:1 (w/w) based on dry weight. The functionalized cellulose materials (NFC/ZS1; NFC/CTAB; NFC/P1670) were then spray-dried (GEA, MOBILE MINOR®) to give powder forms that could be used in combination with hydrophobic polymer matrices.

In screw-capped bottles, appropriate amounts of each of the functionalized cellulose samples i.e. NFC/ZS1, NFC/CTAB, and NFC/P1670 was dispersed in chloroform using a handheld homogenizer (CAT, X 120). A 10 % solution of PLA in chloroform was prepared in another vessel. All samples were then sonicated to eliminate NFC or PLA aggregates and possibly air bubbles. In the next step, a proportional amount of PLA solution was added to each NFC suspension sample, then mixed on a shaker and again subjected to ultrasound. The final concentration of PLA in chloroform was 5 % and the concentration of modified NFC – 10 % in relation to the dry weight of PLA. The resulting compositions (45 g) were poured into shallow rectangular molds (16×16 cm) and air-dried in a fume hood for ca. 3-5 days. The formed PLA/NFC films were then placed in a desiccator under a vacuum for 48 hours to completely remove the organic solvent from the composite mass.

In the same way, reference samples were prepared in the form of PLA with NCF filling without a carrier of surfactants, marked as PLA NFC/0, and film made of pure PLA. The initially transparent layer of pure PLA became opaque during storage in a desiccator. For comparative tests, we also used a transparent PLA film, i.e. dried only in air at atmospheric pressure. The reference samples of pure PLA were named PLA crystalline and PLA amorphic, respectively, which corresponded to their crystalline structure, confirmed by PXRD analysis during further studies.

Decomposition studies of pure PLA films and PLA composites with cellulose modified with different types of surfactants obtained by casting the solution were conducted under simulated composting conditions with controlled temperature and humidity. The samples were weighed and transferred to reaction nets, which were placed in a research reactor filled with compost for a period of 14 days. After the allotted time, the samples were weighed and put back to continue the experiment for one week. The compost used in the biodegradation tests was characterized by moisture in the range of 40-65 % and biological activity (total number of microorganisms) not lower than 10E6 cfu/g. The reactor was placed in a climate chamber. Moisture loss was monitored and replenished daily.

A microscopic analysis of the morphology of NFC surfaces modified with various surfactants is shown in Figure 5. After spray drying, all NFC samples in the form of powder had a shape similar to a spherical one with a folded surface. According to the scale bar in the SEM images, the particle diameters were in the range of 0.85-10.6 μm; but the particle size of the fraction modified with the anionic surfactant ZS1 seemed slightly larger compared to the NFC particles modified with other reagents, i.e. cationic CTAB and non-ionic P1670.

Fig. 5.

PXRD diffraction was used to analyze changes in the crystal structure of cellulose fibers under the influence of the surfactants. In Figure 6, a set of PXRD patterns for unmodified and surfactant-modified NFC fibers can be seen with virtually no difference in the peak shape or intensity for individual samples. Observable peaks at 2θ = 15°, 17°, 22° and 34° correspond to (101), (101¯), (002) and (040) planes, respectively, and are typical of the cellulose I structure [34]. The PXRD result clearly indicates that surfactant molecules penetrating the cellulose fibers do not affect the crystalline domains of the cellulose structure.

Fig. 6.

In order to check whether NFC functionalizing substances have a protective effect on cellulose during NFC drying, preventing so-called keratosis, samples of the surfactant modified NFC powders were re-dispersed in water and air-dried. Keratosis is a physical process involving structural changes occurring as a result of the progress of degradation of the amorphous phase in cellulose fibers, e.g. as a result of spray drying. This may result in the shortening and weakening of mechanical properties of cellulose fibres. Figures 7 a-c show SEM images of re-dispersed surfactant modified NFC particles.

Fig. 7.

As can be seen from the SEM images, NFC/ZS1 and NFC/CTAB particles regained their fibrous structure after re-dispersing in water, although in the case of NFC/P1670 the structure recovery was the least visible.

Samples of 14x14 cm were cut out of the formed films in order to eliminate possible inhomogeneities and deformations of the foil related to the edge effect on the walls of the forming vessel. The thickness of each PLA film was about 0.1 mm. Pictures of the composite films and reference samples of pure PLA in an amorphous and crystalline form are shown in Figure 8.

Fig. 8.

Macroscopic optical evaluation of the composite film obtained showed transparency only for the pure PLA film formed at ambient temperature and pressure. The addition of CNF significantly reduced the light permeability of the composite membranes, which was consistent with the optical observations in both visible and UV wavelengths (see Figure 9). The pure PLA film held under a vacuum to completely remove the organic solvent from the polymer pulp; like the PLA NFC composites samples, it was also less transparent.

Fig. 9.

PLA/NFC composite and pure PLA films were also characterized by PXRD analysis. The diffractograms presented in Figure 10 indicate the crystalline structure of pure PLA film probably obtained by diffusion of chloroform from a transparent amorphous PLA film.

Fig. 10.

Pure PLA crystal as a reference sample was obtained under the conditions in which the composite films with cellulose additives were prepared. In the last stage, all foil samples were placed in a chamber with reduced pressure to get rid of chloroform residues. Solvent vapors escaping from the polymer matrix could favor the ordering of polymer chains through the nucleation and growth of crystals in the matrix. Interestingly, such a phenomenon occurred only for the foil not filled with cellulose particles. NFC-containing samples remained practically amorphous or at most semi-crystalline. In the diffraction pattern of the transparent PLA amorphic film, only a characteristic broad peak at 16° is observed, which may indicate semi-crystallization during processing. The difference between the diffraction patterns of both reference samples made of pure PLA is evident and can be helpful in the analysis of the PLA/NFC composite structure after forming the foil.

The addition of NFC fibers, both in the unchanged form and after modification with surfactants, causes slight changes in the diffraction pattern, i.e. the appearance of convexities on the spectral line at the angles 2Θ=17° and 23° in relation to the spectrum of the amorphous form of PLA. The highest proliferation of the peak at 2Θ=23° can be observed in the diffraction pattern of the PLA NFC/ZS1 sample. It is known from the literature that the addition of nanocellulose to the polymer matrix may increase PLA crystallinity as a result of the introduction of additional crystallization nuclei [26]. However, the changes observed in the spectrum may come from diffraction peaks of NFC crystals, largely masked by the wide diffraction peak of PLA.

Thermal properties of neat PLA and PLA NFC composite films were characterized by DSC and TG techniques. The full DSC curves of the first heating in the range of 20-200 °C and the corresponding TGA plots for all samples are shown in Figure 11.

Fig. 11.

The thermal parameters from the DSC analysis, such as the glass transition temperature (T_g), cold crystallization temperature (T_cc), melting temperature (T_m) and appropriate enthalpies ΔH_cc (cold crystallization enthalpy) and ΔH_m (melting enthalpy) are summarized in Table 2. TGA data are also shown in Table 2, indicating differences in temperature ranges (T₁- T_inf - T₂) and weight loss (Δm) between PLA and PLA NFC composites during the release of residual chloroform. Checchetto et al. confirmed that the desorption process of residual solvent from PLA films occurs in the temperature range from T_g to T_m (DSC), where TG evidences the first mass loss [35].

The DSC profile of the PLA crystalline sample is characterized by a strong endothermic peak associated with melting at T_m = 155°C, which indicates a highly crystalline structure, and the glass transition peak (C_p) is the smallest and at a slightly lower temperature (T_g) compared to other PLA films tested. According to the TG at 20-200°C, this sample is characterized by the lowest weight loss (Δm = -3.3 %), which proves the highest degree of solvent removal from the sample mass during film formation (conditioning in a vacuum desiccator).

The DSC pattern of the amorphous pure PLA film (PLA amorphic) confirmed its character, although the enthalpies of melting and cold crystallization are not completely balanced (ΔH_cc + ΔH_m = 0.3 J/g), which would indicate a low content of crystal structure. Traces of crystallinity in the sample are also recognized by a weaker glass transition stroke (change in heat capacity, ΔCp=0,54 J/g*K) compared to other amorphous samples. The PLA amorphic film is the only one of the samples tested to show two stages of weight loss during heating at 20-200°C (see TG plot), but the total weight loss (Δm = - 3.5 %) is not greater than for the others, although the cast PLA amorphic film was not placed in a vacuum chamber during post-processing. This result is probably due to two different types of interactions between chloroform molecules and PLA chains in the polymer matrix.

The amorphous structures of all composite films containing the NFC additive (PLA NFC/0, PLA NFC/ZS1, PLA NFC/CTAB, PLA NFC/P1670) were confirmed by the following features of DSC profiles: strong glass transitions, where the specific heat values (Cp) are ca. 0.6 J/g*K), as well as the balancing effects of cold crystallization and melting (ΔH_cc + ΔH_m = 0.0 J/g). The addition of unmodified NFC particles to the PLA matrix (see TG for PLA NFC/0 sample) shifts the desolvation temperature towards lower values as compared to the PLA crystalline sample prepared under the same conditions (vacuum post-processing). On the other hand, the release of residual solvent from PLA film with NFC filling modified with surfactants (PLA NFC/ZS1, PLA NFC/CTAB, PLA NFC/P1670) starts at higher temperatures and results in greater weight loss of samples during heating at 20-200°C.

According to TG, the PLA NFC/ZS1 composite foil shows the highest weight loss compared to the other foils tested (Δm = 5 %). This sample is characterized by the lowest glass transition temperature, which, combined with the largest weight loss, would indicate a higher content and weaker binding of the substance volatiles in the PLA matrix. Probably, the slightly larger particles of NFC/ZS1 compared to NFC/CTAB and NFC/1670 (see Figure 3, chapter 3.1) could have adsorbed more chloroform residues.

The PLA NFC/CTAB film shows amorphous properties with cold crystallization and enthalpy of fusion similar to those of the PLA NFC/ZS1 film, and the mass loss (TG) is the lowest in the group of composites with the addition of surfactants (Δm = -3.7 %). The PLA NFC/P1670 film loses Δm = 4.2 % of its weight during heating.

The above analysis of the thermal behaviour of all PLA films at 20-200 °C indicates that the addition of surfactant-modified cellulose nanofibers favours the retention of residual chloroform in the solvent-cast polymer film matrix and hinders its removal under vacuum conditions at ambient temperature.

Tensile properties, such as: the tensile strength and elongation at break, of pure PLA in a crystalline and amorphous form and of PLA NFC composite films were evaluated. The results are summarized in Figure 8.

Fig. 8.

The tensile strength of neat PLA specimens in crystalline and amorphous forms were 42,0±2,4 MPa, and 37,0±2,4 MPa, respectively. Xu Zhang and Xiao Chen obtained slightly lower values of this parameter for foils of the PLA 6000 series but formed from a molten state at various crystallization temperatures (max. 29,4 MPa; min. 12,4 MPa). The authors pointed out that the strength of the cast foils decreased with the increase of the process temperature, and the highest result was obtained for the amorphous form (after rapid cooling). The result was explained by the spherulitic nature of the PLA crystals in the polymer matrix. Many nuclei are formed at a lower temperature, but spherulites increase in size as the crystallization temperature increases. At high temperature, spherolites become larger and irregular, and crystalline and amorphous regions begin to form two separate phases, deteriorating the mechanical properties of polymeric films [36]. In our study, the crystal structure of neat PLA was more robust than the amorphous form in solvent-cast film samples, indicating stronger interactions between ordered crystalline domains.

All composite PLA NFC samples had a lower tensile strength than both pure PLA control samples, especially with respect to the reference sample of crystalline PLA prepared under the same conditions (vacuum post-processing). The tensile strengths measured are 32,8±4,9 MPa for PLA NFC/0 films with non-modified cellulose nanofibers and 3l.5±2,8 MPa, 31,0±2,3 MPa and 34,1±1,5 MPa for samples with surfactant-modified cellulose, i.e. PLA NFC/ZS1, PLA NFC/CTAB and PLANFC/P1670, respectively. These differences were small or non-existent taking into account measurement error. Hence, it is very likely that the cellulose nanofibers in the PLA matrix limited the free diffusion of chloroform between the polymer chains, which slowed down the process of film ordering and strengthening.

Chloroform residues interacting strongly with the polymer matrix (high desolvation temperatures by TGA analysis) can also play a role in both strengthening and plasticizing the film, as evidenced by measurements of the sample’s elongation at break.

The elongation at break for crystalline PLA reached as much as 63.2 %, at which the solvent release temperature was the highest among all films. For amorphous PLA film, this parameter was about 7.92 %; but this is still higher than for samples containing cellulose nanofibrils. It should be noted that this sample was prepared without vacuum treatment and contained residual chloroform with varying degrees of interaction with the PLA chains.

Analyzing the effect of cellulose nanofibers and surfactants on the mechanical properties of composite films, the effect of the content and interaction of solvent residue molecules in the PLA matrix is also clearly visible. The elongation at break for the PLA NFC/0 sample was 4.07 %, similar to that of PLA NFC/ZS1 film, with a value of 4.15 %.

To sum up, the addition of surfactants does not increase the support function of NFC in PLA foil, although CTAB and P1670 may have an indirect plasticizing effect on the composite obtained through stronger bonding of chloroform molecules. Such an effect was not observed in the case of NFC modification with the ZS1 surfactant.

Disintegrability is expressed as a percentage and calculated as the ratio of the mass of the sample after a specific incubation time in simulated aerobic composting conditions to the mass of the starting sample.

In our study, we used the procedure described by Luzi et al. [37]. The disintegration tests were carried out in a controlled environment, at a temperature of 58±2°C and humidity of 40 – 65 %.

Under the influence of moisture and heat, PLA decomposition begins with adsorption and then the diffusion of water into the material. At first, the structure of the polymeric materials loosens because water molecules penetrating the matrix weaken and/or interrupt non-covalent interactions between polymer chains. This step depends on the ability and speed of water adsorption on the polymer surface. In the next stage, the separated polymer chains are cleaved into smaller fragments and/or lactic acid monomers by hydrolysis of ester groups. Hydrolysis is an autocatalytic process because the unblocked carboxyl groups accelerate the decomposition of subsequent ester groups in PLA. This step may be related to the rate of water diffusion in the polymer matrix. Finally, microorganisms present in the soil decompose these polymer residues into carbon dioxide, water and humus.

The decomposition of pure PLA films and PLA composites with cellulose modified with different types of surfactants was analyzed by observing changes in the appearance and weight of the tested samples after 2 weeks and after 3 weeks of storage in soil. After this time, the samples were sieved, dried and weighed. The effects of the samples’ decomposition and percentage weights loss are shown in Figure 9.

Fig. 9.

After 3 weeks of storage in compost soil, the highest weight loss, almost 90 %, was recorded for the pure PLA sample in amorphous form, which was a predictable result compared to the crystalline form. Greater crystallinity gives the polymer a more ordered structure, which significantly limits the diffusion of water into the matrix and reduces the susceptibility to hydrolysis of polymer chains.

Meanwhile, all PLA composite films, both containing native cellulose nanofibers and modified with surfactants, turned out to be more resistant to decomposition in composting conditions. The lowest degree of sample decomposition after 3 weeks of being in the soil was recorded for the composite film with the unmodified cellulose composite, PLA NFC/0, and with cellulose modified with a non-ionic surfactant, PLA NFC/P1670. The best results in the disintegration of PLA/NFC/surfactant composites were obtained for samples containing ionic modifiers, such as cationic CTAB and anionic ZS1 surfactants. Research reports on the impact of cellulose fibers on biodegradation are very different and sometimes contradictory, as excellently summarized in a comprehensive review by Hubbe et al. [38].

The addition of cellulose to PLA generally hinders the diffusion of water into the polymer matrix. According to Jon Trifol et al. the presence of cellulose nanoparticles induces a crystalline structure that more effectively increases the barrier properties of the composite [39]. In our tests, cellulose did not have a significant effect on the crystallinity of the composites prepared. All PLA films with NFC and NFC/surfactant additives were amorphous (see chapter 3.2). In our opinion, unblocked and available – OH groups of cellulose chains can also effectively retain water and slow down its spread into the composite. The treatment of cellulose with a surfactant changes the nature of the surface of cellulose fibers, which promotes the rate of disintegration of the PLA/cellulose composite [37]. Additionally, the disintegrability increases with the value of the HLB surfactant parameter, which was also confirmed by our results [40].

It is worth nothing that the differences in the disintegration of all samples after the first 2 weeks in the soil were small, although the trend was consistent with predictions. A noticeable change in the rate of weight loss of the samples tested occurred in the following week, especially in the case of crystalline PLA films, as a result of autocatalyzed hydrolysis of ester bonds in PLA chains undisturbed by additives that can block access to free -OH and -COOH groups (see Figure10). The terminal OH group of the lactic acid oligomer was found to play an important role in the decomposition of the polymer matrix, both in alkaline and acidic media. As mentioned by de Jong et al. protection of these hydroxyl groups results in significantly delayed degradation [41].

Fig. 10.

It is known that PLA degradation occurs mainly in the amorphous part of the polymer material. The high optical purity of the polymer significantly hinders crystallization, therefore the lower the content of the d-lactide isomer, the greater the tendency to form and retain an amorphous phase.

Giełdowska et al. conducted research on the hydrolytic degradation of fibers made of various types of PLA available on the market (Ingeo series, 4060D, 2002D and 6201D, i.e. with different contents of the D-lactide isomer). The authors showed that the amount of d-lactide isomer in the fibrous material has a significant impact on degradation, with a low d-lactide content leading to faster irreversible changes [42]. In our study we used PLA Ingeo™ 6201D, which is characterized by a low share of the D-lactide isomer in the mass, approximately 1.4 %. For comparison, PLA Ingeo™ 4060D and PLA Ingeo™ 2002D contain 12 % and 2.5 % of the D-lactide isomer, respectively. It seems that the disintegration tendency of PLA 6000 series materials is mainly determined by this parameter, with almost 90 % degradation of the amorphous form of pure PLA. Nevertheless, the introduction of modified cellulose fibers into the polymer matrix, especially with ionic surfactants, supports the degradation of PLA NFC composites as compared to those containing cellulose fibers in their native form.