NMR characterization
The purified furcellaran fraction (Knutsen and Grasdalen 1987), analyzed with HSQC experiment (Fig. 1, left; Table S1 in supplementary information) gave only two anomeric signals. It means β-DA and κ-DA have identical chemical shifts at 5.10/97.10 ppm (1JCH1 165 Hz), and β-G and κ-G4S signals are overlapped at 4.64/104.92 ppm (1JCH1 168 Hz). Small differences were observed for two H-2 signals of κ-DA and β-DA at 4.14 and 4.09/72.47 ppm in COSY and HSQC spectra. Further assignment for both β-DA and κ-DA units was based on TOCSY and HSQC, e.g. H3/C3 signals were assigned at 4.55/81.67 ppm, as well as H3/C3 (3.89/82.46 ppm) of β-G and H3/C3 (4.01/82.46 ppm) of κ-G4S. The known data of β- and κ-carrageenan were in agreement with our data (see Table S1) and helped to assign the values of signals at H4/C4, H5/C5 and H6/C6 positions (van de Velde et al. 2004).
The three-bond spin-spin interaction across the glycosidic bond between H-1 of κ-DA (5.11 ppm) and C-3 of κ-G4S (80.74 ppm) was clearly detected in HMBC (Fig. 1, right; Table S1). The same type of interactions were detected between β-DA H-1 (5.10 ppm) and C-3 (82.51 ppm) of β-G and between H-4 of κ-G4S (4.63 ppm) and C-1 of κ-DA (105.05 ppm). Furthermore, the signal (4.63/105.05 ppm) confirms the connectivity of β-DA at H-4 with C-1 of β-G. The data from these HSQC and HMBC experiments differ up to two ppm to those presented in the recent review. This difference is likely due to different origin of the compound; presented alkali-modified furcellaran published by the authors from Estonia (Marangoni et al. 2021).
Two anomeric signals (5.30/94.41 and 4.67/104.55 ppm) were detected in HSQC spectrum of ι-carrageenan (Fig. 2, left). Both signals were assigned to ι-DA2S (5.30/94.41) and to ι-G4S (4.67/104.55 ppm) based on 13C-NMR data (van de Velde et al. 2004). The third signal (5.39/102.48 ppm), having 1JC1−H1 = 173.5 Hz, originated likely from an unknown α-linked glycan present in FL seaweeds, since our NMR data are different from those of the NMR of isolated starch (Knutsen and Grasdalen, 1987). The signals originating from the ring carbons of ι-carrageenan were assigned from HSQC, COSY and TOCSY spectra. The signal at 4.93/74.42 ppm belongs to ι-G4Srβ (Préchoux et al. 2014) and signals of α-glycan (αG) are: 3.67/80.00 ppm (2 αG), 3.97/76.26 ppm (3 αG), 3.85/74.10 ppm (4 αG), 3.67/74.47 ppm (5 αG) and 3.95, 4.00/63.68 ppm (6 αG). The peak corresponding to long-range spin-spin interaction across the inter-glycosidic linkage (4.67/80.09 ppm) between H-1 of ι-G4S and C-3 of ι-DA2S unit was observed in HMBC (Fig. 2, right). The other four signals originated from the intra-ring interactions in ι-carrageenan.
Four anomeric signals were detected in HSQC spectrum of κ-carrageenan (Fig. 3, left). Two overlapping signals at 5.11/97.27 ppm originated from β- and κ-DA units and 4.66/104.80 ppm from β-G and κ-G4S units similarly to furcellaran (see Table S1). However, there was no signal at 4.14/66.70 ppm related to non-sulfated G4. This indicates the absence of β-carrageenan in κ-carrageenan sample. In addition, there was also a signal at 5.30/94.46 ppm corresponding to ι-DA2S anomeric chemical shift (van de Velde et al. 2014). The low intensity of ι-DA2S anomeric signal indicate the minor content of ι-carrageenan in the κ-carrageenan sample. The fourth anomeric signal, which was already present in the spectrum of ι-carrageenan at 5.39/102.56 ppm, originated from α-glycan. The remaining signals were identical with those observed in HSQC spectrum of ι-carrageenan (Fig. 2, left). The peaks corresponding to inter-glycosidic correlations between κ-DA (H4) and κ-G4S (C1) (4.61/104.81 ppm) and κ-DA (H1) and κ-G4S (C3) (5.11/81.53 ppm) were observed in HMBC (Fig. 2, right). These and other signals in HMBC (intra-ring proton-carbon interactions), are typical for κ-carrageenan. The presence of other peaks with low intensities in HSQC spectrum could not be reliably assigned from obtained HMBC data either due to a low concentration or due to high molecular weight (short T2 values). The κ-carrageenan polysaccharide showed different HSQC spectrum (Fig. 3), compared to one observed for κ-carrageenan oligosaccharides (Yu at al. 2002). These results indicate the effect of high molecular weight of our samples on the shape of the NMR spectra.
The isolated FLW extract represents 44% yield from the Furcellaria lumbricalis specia. In this case following anomeric signals can be assigned in HSQC spectrum (Fig. 4, left; Table S1): λ-G2S (4.71/105.86), λ-D2S,6S (5.63/93.87 ppm), θ-G2S (5.39/102.55 ppm), θ-DA2S (5.22/98.50 ppm), ζ-D2S (5.23/94.77 ppm), ζ-G2S (4.71/105.84 ppm), µ-D6S and ν-D2S,6S (4.97/100.70 ppm), µ- and ν-G4S (4.49/106.18 ppm). These results are in agreement with previously published data (van de Velde at al. 2004). We note that these are solely sulphated units. This may be due to better water solubility of sulphated components (solubilized due to the ion-exchanging sulphate groups), as opposed to less sulphated carrageenans. Sulphur content of the FLW fraction (11.80 %) further supports this theory. Similarly to previous cases (ι- and κ-carrageenan), the signal at 5.39/102.55 ppm originated from α-glycan (Table S1 in supplementary information). The anomeric signal at 4.65/98.68 ppm was assigned to the κ-β-κ hybrid unit based on proton chemical shift at 4.67 ppm (G4Srβ-H1 or G4Snr-H1 (Pelosi at al. 2003). The remaining ring signals were assigned similarly to previous cases (see Table S1 in supplementary information). Additionally, the ζ-D2S unit showed identical anomeric signal as α-DA2S (5.23/94.77 ppm). Also observed was the (1–3)-α-G-G6S signal overlapping with θ-G2S signal (5.39/102.55 ppm). The signal of α-carrageenan (α-G) was overlapping with signals of λ-G2S and ζ-G2S (Table S1 in supplementary information). The signal at 3.90/102.54 ppm in HMBC originated from the intra-ring correlation between αG6S(H3) and αG6S(C1) and was overlapped with αG(H3)→αG(C1) correlation (3.92/102.60 ppm) (Fig. 4, right). Both of these observed correlations confirmed the presence of (1–3)-α-G-G6S glycan.
Used sonication extraction method resulted in acceptable yield of Gigartina water-soluble extract (GSUW, 42 %). The sample was extracted in the presence of NaN3 (1 %) to prevent fungi degradation. This polysaccharide fraction showed seven signals and was analysed by COSY (not shown) and HSQC experiment (Fig. 5, left and Table S1). The anomeric signal resonated at 5.39/102.54 ppm, H6/C6 signal at 3.68–3.90/63.40 ppm, H2/C2 signal at 3.66/74.39 ppm and H3/C3 signal at 3.97/76.17 ppm. Similar analysis using TOCSY (not shown) and HSQC enabled the assignment of H4/C4 at 3.84/74.11 ppm and H5/C5 at 3.65/74.39 ppm. The other H6/C6 signal was detected at 4.32–4.35/72.08 ppm and likely originated from the sulphated C6 group. The HMBC experiment (Fig. 5, right) allowed the assignment of the signal at 5.41/76.16 ppm which originated from inter-glycosidic interaction (H1→C3) in the non-sulphated (1–3)-α-glycan unit. The signals at 5.38/74.11 ppm represent the (H1→C5), interaction; further intra-ring interactions across two or three bonds were identified: 3.97/79.92 ppm (H3→C4), 3.97/74.33 ppm (H3→C2), 3.66/74.14 ppm, (H4→C5), 3.66/76.15 ppm (H4→C3), 102.50/3.65 ppm (H2→C1), and 3.66/63.34 ppm (H5→C6). The α-anomer was compatible with the value 1JC1−H1 = 173.5 Hz. This result is supported by the NMR data of (1–3)-β-glucan run in (CD3)SO with chemical shift at 86.30 ppm (C3; Pelosi et al. 2003). Observed value is ten ppm apart from our α-glycan C3 signal run in D2O. The water-solubility of our glycan with much higher molecular weight (see SEC-MALS section) indicates a dramatical difference in its physical properties. Relatively high value of the C-6 chemical shift (6 αG6S 72.08 ppm), along with relatively high sulphur content of the fraction (11.94 %) support the assumption that C-6 carbon of (1–3)-α-glycan is sulphated. Except for the furcellaran fraction, this polysaccharide was present in all the previous fractions. Based on 1H-NMR chemical shift of C6-sulphated signal is at 4.40 ppm observed for sulphated starch (Vega-Rios at al. 2014), we assume that our (1–3)-α-glycan is sulphated at C6 carbon (αG6S).
When GS was extracted using sonication under the carbonate-peroxide oxidative conditions (GSUCP), the fraction yield (63 %) increased in comparison to GSUW fractions. HSQC experiment (Fig. 6, left) enabled the assignment of the αG anomeric signals in (1–3)-α-glycan (5.40/102.50 ppm) as well as anomers in θ-carrageenan (5.36/102.75 ppm, G2S and 5.22/98.42, DA2S). Besides the anomeric and ring signals, we also observed three types of C6 groups at 3.83, 3.87/63.39 (αG6), 3.78, 3.99/68.56 (µ-D6S) and 3.73, 4.15/70.36 ppm (ν-D2S,6S) that are typical for the carrageenans. HMBC spectrum (Fig. 6, right) revealed two inter-ring correlations, particularly θ-G2S(H3)→θ-DA2S(C1) at 3.77/98.50 ppm and θ-G2S(H1)→θ-DA2S(C3) at 5.39/75.00 ppm, and two intra-ring correlations: αG(H2)→αG(C1) at 5.39/76.22 ppm and ν-G4S(H1)→ν-G4S(C2) at 4.98/72.57 ppm. These results suggest the GSUCP fraction contained (1–3)-α-G-G6S glycan, θ-carrageenan, and µ- and ν-carrageenans hybrids without ν- and µ-G4S HSQC signals at 4.49/106.18 ppm otherwise present in FLW fraction (Fig. 4). The GSUCP is a very mild extraction method with 184 kJ of total converted mechanical energy (see Methods) and produced the highest molecular weight fraction. This low-energy extraction process suggests the absence of the µ-D6S and ν-D2S,6S units is not related to the extraction-induced material degradation. Instead, we believe the reason for this is the presence of hybrids containing only the µ-D6S and ν-D2S,6S repeating units and not the µ-G4S and ν-G4S units typical for µ- and ν-carrageenans. This assumption is supported by a complete absence of anomeric signals at 4.49/106.18 ppm in Fig. 6, otherwise observed in FLW extract containing µ- and ν-carrageenans (Fig. 4). Also, there was no signal observed at 175 ppm in the HMBC spectrum of GSUCP extract, which further supports the absence of oxidation products.
The GSW extraction resulted in the highest yield (68 %). Besides the signals observed in the previous fractions (αG, αG6S, θ-carrageenan), the signals originating from α-, β-, ι-, κ- and ζ-carrageenan were also present (Fig. 7, left) in this fraction. The absence of µ- and ν-G4S anomeric signals indicate the presence of some hybrid, similarly to the case of GSUCP extract. HMBC spectrum (Fig. 7, right) revealed mainly inter-glycosidic bond correlations in θ- and κ-carrageenan indicating their predominance in the extract. The presence of β-carrageenan was confirmed by the two overlapping peaks present in HSQC spectrum at 5.10/97.10 and 5.11/97.70 ppm.
NMR results of all the seven studied samples indicate that GSW consisted of the most complicated mixture of α-, β-, κ-, ι-, θ-, ζ-carrageenans, (1–3)-α-glycan partially sulfated at C6, and µ-D6S and ν-D2S,6S hybrids. Analysis of GSUW and GSUCP extracts showed that the sonication extraction method was the most selective and sensitive. We note the GSUW method yielded only the (1–3)-α-glycan predominantly sulphated at C6. The GSUCP extraction enabled separation of only three hybrids (µ-D6S, ν-D2S,6S and (1–3)-α-glycan containing G6S groups) and resulting fraction lacked other known regular carrageenan structures. To the best of our knowledge, these three polysaccharides have not yet been observed in red algae seaweeds. No cellulose fragments could be detected in none of the seven analysed extracts. This is based on absence of known 13C NMR chemical shifts of solubilized cellulose fragments at 103.50 (C1) and 60.90 ppm (C6) in comparison to our data (Table S1, resp. Toffanin at al, 1994).
Sec-mals Analysis
SEC-MALS results are summarized in Fig. 8. It compares molecular weight distribution (MWD) and conformation plot Rg = ƒ(M). MALS detector is an elastic or total intensity light scattering which determines macromolecule size (radius of gyration Rg) from the angular variation of the scattered light and potentially it is able to determine size Rg when it assumes values greater than 10–20 nm depending from the specific conditions. Specifically, for the analyzed polysaccharides in the complex acidic aqueous solvent (0.1M acetate buffer pH 4.5) the Rg value is measured with sufficient accuracy in the range from ≈30 nm to ≈150 nm. The conformation plot Rg=K·Mα is the scaling law between macromolecules size (Rg) and molecular weight (M). Conformation plot is very important because is correlated to the macromolecules stiffness in solution and the thermodynamic strength of the solvent (good or poor or theta). Based on the slope of the conformation plot (α) there are these potential conformations: i) α > 0.5 up to 0.6 flexible random coil in good solvent; ii) α > 0.6 stiff, i.e. semi-rigid up to rigid rod; iii) α < 0.5 low to 0.33; α = 0.33 is the slope of a compact sphere. Also macromolecules structure (linear or branched particularly long chain branching) has meaningful influence on slope value (α) and curvature of the conformation plot.
Analyzing data shown in Fig. 8 indicates meaningful differences between various samples in both MWD and molecular conformation. Multimodal chromatogram (three peaks) and different polymer conformation (macromolecules stiffness) of many samples demonstrate that they are mixture of polysaccharides with different macromolecular structure. SEC-MALS results are an important confirmation of NMR results.
MWD of κ-, ι-carrageenan and furcellaran (Fig. 8, upper panel) shows three peaks (i.e. three components) with molecular weight ranging from very high to relatively low. On the contrary, MWD of GSUCP, GSUW, FLW, and GSW (Fig. 8, bottom panel), with exception of FLW sample shows a single main peak with molecular weight ranging from very high to high. FLW sample show also a minor low molecular weight peak (Mp=3.8 kg/mol). Summary of the SEC-MALS results (Mp and Amount) for κ-, ι-carrageenan and furcellaran is in Table 1.
Table 1
Summary SEC-MALS results of three samples.
Sample
|
Peak 1 (Main)
|
Peak 2
|
Peak 3
|
|
Mp (kg/mol)
|
Amount (%)
|
Mp (kg/mol)
|
Amount (%)
|
Mp (kg/mol)
|
Amount (%)
|
κ-carrageenan
|
173.8
|
77.8
|
35.0
|
14.9
|
14.0
|
7,3
|
ι-carrageenan
|
264.4
|
87.0
|
56.0
|
8.6
|
11.9
|
4.4
|
furcellaran
|
144.0
|
75.6
|
15.4
|
16.8
|
4,3
|
7.6
|
Average molecular weight results for κ-carrageenan are: Mn=66.4 kg/mol, Mw=143.3 kg/mol, polydispersity index Mw/Mn=2.2. Furcellaran assume these average values: Mn=29.8 kg/mol, Mw=153.5 kg/mol, Mw/Mn=5.2.; very broad polydispersity (Mw/Mn > 5) indicate presence of several fractions different in molecular weight and structure. The ι-carrageenan sample has higher molecular weight than κ-carrageenan and furcellaran: Mn=110.2 kg/mol; Mw=296.8 kg/mol; Mw/Mn=2.7.
Summary of the SEC-MALS results (Mp, Mw, Mw/Mn) of GSUCP, GSUW, FLW, and GSW extracts are in Table 2.
Table 2
Summary SEC-MALS results four extracts.
Sample
|
Mp
|
Mn
|
Mw
|
Mw/Mn
|
|
kg/mol
|
kg/mol
|
kg/mol
|
|
GSW
|
77.8
|
61.1
|
158.5
|
2.6
|
GSUW
|
564.6
|
305.5
|
479.8
|
1.6
|
GSUCP
|
994.0
|
430.5
|
1373.5
|
3.2
|
FLW
|
497.7
|
311.9
|
531.0
|
1.7
|
According to SEC-MALS results the peak molecular weight Mp decrease: GSUCP (994 kg/mol) > GSUW (565 kg/mol) > FLW (498 kg/mol) > ι-carrageenan (264 kg/mol) > κ-carrageenan (174 kg/mol) > furcellaran (144 kg/mol) > GSW (78 kg/mol). GSUCP fraction show the highest molecular weight and the biggest size of coil, than GSUW and FLW. The lowest SEC-MALS values were determined for the GSW extract. This resulted from use of sonication method in combination with carbonate/peroxide dissolution potential as opposed to just pure water reflux treatment.
Conformation of κ-carrageenan macromolecules, Fig. 8 upper panel, substantially correspond to flexible random coil (slope of conformation plot α = 0.54). Conformation of ι-carrageenan is much less flexible respect to κ-carrageenan up to nearly compact sphere. Conformation of furcellaran macromolecules is substantially flexible, only a little more compact respect to κ-carrageenan. In conformation analysis it should be considered that the presence of some extent of molecular aggregation in the aqueous acidic solvent (pH = 4.5) could influence results in particular for ι-carrageenan compact polysaccharide. Conformation of four extracts is quite different, varying between κ-carrageenan (flexible) and ι-carrageenan (compact).
Xrd Analysis
Symmetric XRD scans were measured on all samples to study the presence of cellulose; Fig. 9 presents the acquired data and shows very different diffraction pattern for each sample, however mostly amorphous character and virtually no cellulose content can be concluded in all cases. The latter was corroborated by our NMR results discussed in previous sections.
XRD trace of the GSUCP film (topmost profile in Fig. 9) showed a slight resemblance of ball-milled cotton cellulose II (Nam et al., 2016), i.e. with diffraction maxima \(110\) and \(020\) (2θ = ~ 21°) with very broad full width at half maximum (FWHM), however no signal corresponding to \(110\) reflection was observed. This means significant cellulose II content can be ruled out. Similarly, due to the lacking \(110\) and \(110\) reflections at corresponding 2θ positions, significant cellulose Iβ content in the GSUCP film was unlikely (Nam et al., 2016). We note this fraction was extracted with Na2CO3/H2O2 method at room temperature and SEC-MALS analysis confirmed the highest molecular weight of all samples. This indicates the observed XRD pattern corresponds to the presence of some non-cellulosic amorphous polysaccharide(s) with a broad diffraction peak centered at 2θ = ~ 21°. Similar results were obtained for the extract of Antarctic algae Cystosphaera jacquinottii (Paniz et al., 2020). Reported data on this specia confirmed 5% content of cellulose before extraction.
The shape of FLW XRD profile (second from top) showed an additional maximum centered at 2θ = 14°-16°. This could be related to the diffraction maxima of cellulose Iβ (diffraction indices \(110\) and\(110\)), however no signal was observed for the most intense peak corresponding to \(220\) reflection centered at 2θ = ~ 23° (French, 2014). Therefore, we believe this feature was related to some non-cellulose polysaccharide.
XRD profile of GSW film (third from top) was almost identical to that of GSUCP with the exception of a weak shoulder at 2θ ~ 6° unrelated to cellulose.
GSUW (fourth from top) profile showed similar shape to FLW (second from top), however somewhat lower intensity of the main band was observed; no traces of cellulose were concluded. This reflects the low extraction power of used ultrasound treatment, which was unable to extract cellulose from the substrate.
Similarly to previous samples, the three remaining XRD profiles, i.e. furcellaran (third from bottom), ι-carrageenan (second from bottom), and κ-carrageenan (first from bottom) showed diffraction maxima unrelated to neither of the two types of cellulose, confirming our NMR results.
Pf-qnm Analysis
In this section, we present surface morphology and surface mechanical properties of all seven films. RMS surface roughness and correlation lengths where applicable and reduced modulus, stiffness, and adhesion quantities were determined. Due to the large variations in spatially-resolved measured mechanical properties resulting, e.g. from AFM tip contact area variations close to the edges of surface features, all PF-QNM-derived quantitative properties, i.e. reduced modulus, stiffness, and adhesion were evaluated from reduced scan areas (100×100nm) selected from presented 1µm × 1µm and 2µm × 2µm scans. Several (3–5) reduced areas were chosen across each scan; Showed quantitative data represent mean values derived from all reduced areas in each sample. Bar charts in Fig. 14a – d show a comparison of determined values of RMS surface roughness, reduced modulus, stiffness and adhesion for all samples. Surface topography maps and spatially-resolved mechanical properties are showed and discussed in detail in following sections.
Topography
Figure 10a) - g) show representative surface topography scans of all studied samples. 1D line profiles corresponding to white horizontal lines are also shown. Values of RMS surface roughness and correlation lengths, where applicable, were determined from 2µm × 2µm topography scans. In case of grainy surface, correlation length represents the average diameter of particles observed in the surface topography scans. Determined surface roughness values were increasing in following order: FLW (2.4 nm) < GSUCP (3.6 nm) < κ-carrageenan (5.2 nm) < ι-carrageenan (5.5 nm) < GSW (8.0 nm) < GSUW (14.0 nm) < furcellaran (18.6 nm). Correlation lengths were determined only for samples, where grainy surface was observed, i.e. κ-carrageenan, ι-carrageenan, and furcellaran and their values are presented and discussed in the next section.
We also note that circular particles with mean diameter of ~ 80 nm and mean height of ~ 20 nm were observed scattered across the GSW sample surface (e.g. shown in the bottom of Fig. 10b) with concentration of ~ 2.3×108cm−2, which were likely extraction-related residues. Based on the values of gyration radii (Rg) determiend by SEC-MALS analysis we conclude that surface roughness and correlation lengths are related to polysaccharide coil dimensions.
Reduced Modulus
Size and distribution of macromolecular coils constituting the polysaccharide films can significantly affect their mechanical properties, therefore in this section we compare reduced moduli of our samples with the determined values of gyration radii (Rg) and correlation lengths (T). Correlation lengths and partially also surface roughness are related to polysaccharide coil dimensions expressed by gyration radii of the seaweed fractions. AFM-determined correlation lengths were found to be in good agreement with Rg values of polysaccharide coils obtained from SEC-MALS analysis.
Reduced modulus maps of all studied samples are shown in Fig. 11a) – g). Average moduli values extracted from reduced 100 × 100 nm areas in ascending order were: ι-carrageenan (24 ± 4.5 MPa) < GSW (378 ± 87 MPa) < κ-carrageenan (433 ± 186 MPa) < GSUCP (807 ± 360 MPa) < furcellaran (817 ± 224 MPa) < GSUW (822 ± 262 MPa) < FLW (1685 ± 575 MPa). Figure 12 (a) shows reduced modulus map of ι-carrageenan; similar to topography scan (Fig. 11 (a), coil-like structures packed close to each other were observed. According to SEC-MALS, for the main fraction with molecular weight at the peak Mp 264 kg/mol and the Rg values (50–80 nm) observe only in fraction of Mw from 264 to 1800 kg/mol, is in the range of calculated correlation length T = 57.7 ± 3.2 nm. Reduced modulus map of GSW film is shown in Fig. 12 (b). This film consisted of α-, β-, κ-, ι-, θ-, ζ-carrageenans, (1–3)-α-glycan partially sulfated at C6, and µ-D6S and ν-D2S,6S hybrid polysaccharides. This fraction has the smallest Mp at 7 kg/mol and we think it is the reason that no coil-like particles were observed in this case. According to SEC-MALS analysis the observed Rg values at 40–55 nm. For GSW Mp value is too small and consequently the Rg value is already too small to be considered as exact. For that reason we think that for such small molecular weigh the polysaccharide do not form the coil conformations. Value of GSW reduced modulus was more than one order of magnitude lower (24 MPa vs 378 MPa). The κ-carrageenan fraction consisted of κ-, β-, ι-carrageenan, and (1–3)-α-glycan partially sulfated at C6. Its correlation length T = 43.6 ± 2.7 nm and according to Mp 173 kg/mol and Rg = 40–150 nm is higher than for ι-carrageenan, although it is a component of κ-carrageenan fraction blend. In this case, coils were less closely packed when compared to those of ι-carrageenan. Reduced modulus of furcellaran increased to 817 MPa and correlation length was T = 85.3 ± 5.7 nm, which agreed with Rg = 40–100 nm measurable only from Mp 140 kg/mol. The GSUW film image showed most of the surface without coils. According to SEC-MALS analysis of (1–3)-α-glycan partially sulfated at C6, the coil dimensions Rg = 45–120 nm observed at 120–800 kg/mol. Reduced modulus of GSUCP composite film (Fig. 12f) was 807 MPa and Rg = 100–135 nm observed at range from Mw 700–6000 kg/mol. The FLW film showed the highest reduced modulus (1685 ± 575 MPa) in relation to Rg = 45–120 nm observed at Mw range 180–1300 kg/mol. The fluctuation of the parameters could be explained by broader distribution of the coil volumes due the Mw/Mn distribution.
Stiffness
Stiffness reflects the measure of resistance to deformation and factors mechanical properties of the material and its shape. According to average stiffness values extracted from 100×100nm reduced areas of PF-QNM-determined stiffness maps, these were increasing in following order: ι-carrageenan (0.8 ± 0.1 N/m) < κ-carrageenan (10.7 ± 3.2 N/m) < GSW (11.5 ± 1.7 N/m) < furcellaran (13.2 ± 2.4 N/m) < GSUCP (13.4 ± 3.8 N/m) < GSUW (14.4 ± 3.0 N/m) < FLW (23.1 ± 5.5 N/m). The order is similar as in case of reduced modulus, but with exchanged order for GSW and κ-carrageenan, and furcellaran and GSUCP. However, due to the measurement uncertainty, these differences were negligible. Compared to the majority of GSW surface, slightly lower stiffness was observed for sparsely distributed isolated circular grains (Fig. 12b), typically ~ 7.9 ± 1.5 N/m. Determined stiffness maps for all studied samples are shown in Fig. 12a – g).
Adhesion
Determined adhesion maps detail work of adhesion required for AFM tip to detach from the surface of the sample on each measured pixel. Capillary, electrostatic, or van der Waals forces are some of the typical sources of adhesion at the nanoscale. Determined adhesion maps for all samples are shown in Fig. 13a-g. Average adhesion forces extracted from 100×100nm areas for studied composite films were in following ascending order: κ-carrageenan (4.5 ± 1.7nN) < ι-carrageenan (4.7 ± 2.1nN) < furcellaran (7.3 ± 1.3nN) < GSUW (9.4 ± 1.0nN) < GSUCP (14.5 ± 1.3nN) < FLW (15.0 ± 3.1nN) < GSW (22.1 ± 2.8nN). The order is similar to reduced modulus and stiffness, with the difference that GSW showed the highest adhesion from all samples. Contrary to the majority of GSW surface due to smallest Mp, sparsely distributed isolated circular grains showed much lower average values of adhesion, typically ~ 6.5 ± 1.6nN. Largest variations in adhesion across the surface of calculated adhesion maps were observed for GSW, GSUW, and GSUCP films. This fractions have the biggest Mp values and various amounts of hybrids, which constituted these films. Large variations in topography in case of sample GSUW may also be partially responsible.
The values of composite films roughness, reduced modulus, stiffness and adhesion are on Fig. 14. According to AFM analysis using PF-QNM methodology there is relation between Rg values, corresponding Mp values and topography roughness, reduced moduli, stiffness and adhesion parameters. There is certain optimum between molecular weight and the coil dimensions of individual types of polysaccharides in the bland. The biggest surface roughness of furcellaran and GSW might be due to the second smallest Mp values in comparison to other fractions. The opposite situation occurred for reduced modulus of FLW fraction when the second biggest Mp was observed. Also for stiffness parameter the highest value was observed on FLW. For adhesion parameter the highest value observed on GSW fraction indicates the smallest Mp and Rg values are affecting the adhesion process the most.