Introduction
PCW are the world's most abundant type of biomaterials and have applications ranging from dietary fibre and structuring the components in foods, raw materials for bio refining and as a source of Polymers for things such as gum and stabilisers. It also has an important role to play from a drug delivery perspective in that it can be the precursor to the formation of novel forms of biocapsules to used for a plethora of issues ranging from delivering insulin to type 1 diabetes patients to delivering cytotoxic drugs to patients who have cancer. These biocapsules would be superior to more traditional forms of drug delivery methods in that they would have far less contamination level and exhibit very low degree of stomach degradation due to the "recalcitrant nature"(Atalla) of PCW while at the same time being vulnerable to degradation in the small intestitine which is precisely the position at which the biocapsule should be degraded for maximum absorption. Studying the structural and mechanical properties of PCW is key to being able to utilise PCW for any of these functions. PCW are composed of Cellulose microfibrils which consists of unbranched, unsubstituted (1,4)-B-D-glucan chains that are able to form microfibrills. These exist in several different levels of crystallinity and length depending on site and function of cell. Furthermore there is difference of crystallinity in the individual cell where Cellulose is broken into layers of crystalline, paraccystaline and amorphous Cellulose. PCW also contains Pectins which are wall polysaccharides that are solubilized by aqueous buffers and dilute acidic solutions and calcium chelators (Cosgrove 2015). Pectins are very important for PCW as they form hydrated gels that push microfibrils apart, easing their sideways slippage during cell growth, while also locking them in place when growth ceases (Lord 2002). Hemicelluloses are heteropolymers that are present with Cellulose in almost all plant cells, these are a wide variety of molecules which have a random amorphous structure with little strength and they could be solubilised by a strong alkali. These would include Xylan, Glucuronoxylan, Arabinoxylan, Glucomannan and Xyloglucan. Different levels of these polymers are found in different plant and in different locations around the plant (Carpita 1993).Further complexity is added by the fact that plant cells normally have two cell walls. A primary cell wall formed during cell expansion and a secondary cell wall that is deposited once cell expansion has ceased. Cellulose and Lignin levels are typically higher in secondary PCW than in primary PCW. Secondary cell walls provide additional protection to cells and rigidity and strength to the larger plant. These walls are constructed of layered sheaths of Cellulose microfibrils, wherein the fibres are in parallel within each layer. The inclusion of Lignin makes the secondary cell wall less flexible and less permeable to water than the primary cell wall (Evert 2013) .In addition to making the walls more resistant to degradation, the hydrophobic nature of Lignin within these tissues is essential for containing water within the vascular tissues that carry it throughout the plant. All these factors contribute to making PCW highly complex systems that are very hard to analyse the individual parts without damaging the structure and thus not understanding the structure properly. Thus, PCW mimics were made in the form of Bacterial Cellulose Hydrogels and were characterised by SSNMR (Suzuki 2001) . These hydrogels are highly hydrated systems in which cellulose fibrillar networks interact with interstitial water at different structural levels. Several bacterial species such as Komagataeibacter xylinus (formerly known as Gluconacetobacter xylinus)(Lin 2012), are able to synthesise cellulose hydrogels in the form of pellicles when inoculated into a culture medium rich in carbohydrates or polyols. The synthesised cellulose hydrogels present a high degree of purity and hydration (ca. 99 wt% H2O) and possess a complex structure in which cellulose is hierarchically assembled to form different structural features (Martínez-Sanz 2016)(Martínez-Sanz 2016a). The incorporation of PCW components into the bacterial culture Makes it possible to study the effects of individual hemcelluloses on the strucutre of cellulose, this is especially useful since it has been shown that the way hemecelluloses get incorrporated in the gel is in similar form to the way they get incorporated in the PCW. The incorporation of several PCW polysaccharides such as XG, mannans, AX, MLG and pectins using this approach is well reported in the literature [2,12,13,15e18,21e27]. From some of these studies it has been inferred that only XG and mannans are able to interfere with the cellulose crystallisation process, reducing the crystallinity index and promoting the formation of the cellulose Ib allomorph [14,15,17,18,23,26]. This indicates that a certain fraction of XG (or mannan) is able to interact directly with the individual cellulose microfibrils. In addition, a different fraction of XG, which interacts with the surface of cellulose ribbons, has been identified [14,15,17]. This fraction is thought to correspond to the thin XG cross-bridges detected in the microscopy images from composite hydrogels [14,15,17,26]. The aim of this work is to use SSNMR to obtain data about the rigdity,Crystalinity,Solavation and relaxation times of Cellulose Hydrogels and to identify the differences that happen when either Xyloglucan or Arabinoxylan is added to it and to use a model of I-S or I-I*-S to Interpret data obtained and to also to point out the limitation that CP might have to probing the structure of Bacterial cellulose and to present methods to tackle that problem.
Materials and Methods
13C H2O or D2O Bacterial Cellulose Gel was made by the method provided by the Literature (Martínez-Sanz 2016a). They were made from from Komagataeibacter xylinus which was fermended on Protiated and deuterated Carbon Sources to produce Protiated and Deuterated Cellulose hydrogels. The bacterial strainwas cultivated in Hestrin and Shramm (HS) medium with the addition of glucose as the carbon source. Incubation was performed at 30 °C for 72 h.
At the end of incubation, cellulose hydrogels were harvested and washed in ice cold milliQ water with gentle agitation (150 rpm), with frequent rinsing to wash off excess medium and dislodge loosely-associated bacterial cells, until the harvested hydrogels were white in appearance.
Solid state MAS 13C NMR. For all resonance spectroscopy measurements a Bruker 400 MHz spectrometer equipped with Topspin software (Version ?) and a ?.? mm Probe. The MAS spinning rate was 6 KHz. Adamantane was used as an external reference and the downfield peak as set to ?.? ppm. Experiments were done in two temperatures (25 °C and 5 °C ). These general parameters were used for all the NMR experiments. The other parameters (Contact time,Mixing time,D20,etc...) were then adjusted for each series of experiments as deemed appropriate
Relaxation time. Relaxation experiments conducted were done using CP Saturation recovery to obtain the spin-Lattice (T1) relaxation time under CP conditions and direct polarisation inversion recovery to obtain the T1 with no CP conditions. The Parameters for the experiment were: 128 NS, 8 DS, D1 10 s and the P15 for the CPsatrec is 2000 us.
CP Kinetics. Cross polarization is the transfer of magnetization from the abundant spin I to the rare spin S using dipolar couplings. This is done to enhance signal-to-noise of nuclei such as 13C (1.1% of all carbons). This is done by sending spin-lock pulses that are set to achieve the Hartmann-Hahn Matching Condition γ𝐵1(1𝐻)=γ𝐵1(𝑋). When the Hartmann-Hahn condition is met, the energy gaps between H and X spin states are equal. Under these conditions, there is a transfer of population from the abundant spin I to rare spin S by dipolar coupling. The time in which the Hartmann-Hahn condition is allowed to exist is called the contact time (P15). The parameters for the experiment were: Number of scans (NS) 512, Dummy scans (DS) 8, Recylce delay (d1) 10 s. The P15 was arranged in sequence from 10 us to 6 Ms with very short sequences early on (10,20,40,60,80,100) because most of the growth phase in CP happens then. This is done to obtain a profile of CP efficiency in regards to the Contact time permitted. This should allow us to comprehend the rigidity of the system. That is because CP relies on heteronuclear dipolar couplings between the source and target spins, its efficiency is uniquely dependent on the concentration, distribution and mobility of the nuclei involved. Therefore, the technique is remarkably sensitive to the composition, structure, properties and molecular dynamics in the solid(Kaflak 2008). Processing parameters made to be the same for all graphs in order make them consistent with each other and hence comparable.
CP/HPDEC normalisation. High powered decoupling (HPDEC) is applied to remove the broading effects of heteronuclear coupling.The application of high power irradiation at the frequency of the 1H resonance causes those spins to undergo repeated transitions at a rate determined by the amplitude of the irradiation.If the rate is rapid enough the desired nuclei can be observed without the broadening effects of dipolar coupling.HPDEC is normally useful in understanding the flexibility of the sample and in being able to quantitatively analyze it. However. We used HPDEC as a normalization against CP. This is so we can quantitate the proportion of Crystalline to amorphous regions for each gel in the sample.
Fig. 1
CP Pulse sequence
Water polarisation transfer (WPT). Water polarisation transfer is an NMR technique that allows to obtain the hydration profile of polymer fibres. It does so by using a T2 filter to remove the transverse magnetisation of the surface of the fibres and leaving water polarisation to then be transferred to the polymer by spin diffusion, chemical exchange and NOE. By representing the normalised growth of water polarisation against the √ mixing time, the hydration profile of each carbon is obtained. The parameters for the experiment were: NS was 256 scans, DS was8 scans, D1 was 10 s, P15 was 500 us. One spectrum was run with no T2 filter (D20) to be used a reference sample. One spectra was ran with a T2 filter but with no mixing time to check for effectiveness of T2 filter. The rest were subsequently had a D20 of 2 Ms and a series of D8 that ranged from 1Ms to 324 Ms. All WPT experiments were conducted in 293 K.
Fig. 2
WPT Pulse sequence
Fig. 3
Transfer of Water polarisation
Data Fitting. The Equations used to fit the CP experiments are ones that are used to understand the mobility of gel like systems that have a high degree of water in them(Pérez-Hernández 2012) . The equations used operate the I-I*-S model in which I* denotes protons in close proximity to the S Nucleus. Within this model, source spins, such as protons, do not have a common spin temperature. The I*-S spin pairs or clusters can be viewed as isolated in a solid and exchanging polarization in an oscillatory manner, damped by the subsequent spin-diffusion contact with the remaining I spins.(Kolodziejski 2010) This model is especially useful in solids where the system is anisotropic. The equation used for adjusting the experimental curves:
From this fit, it is possible to obtain the TCH values which represent the efficiency of magnetisation transfer and that can be used as a guide to describe the rigidity of the system (The higher the magnetisation transfer the more rigid a system is). The software used for the fitting of the graph was Origin.
Results
Investingation of rigid compoents by CP kinetics. the CP done on BC gels first allowed for a characterisation of them. it showed that there were C1 at 105 ppm along and it had separate peaks for different allomorphs of C1 (specifically C1 \alphaα and C1 \betaβ). It managed to isolate C3 (75 ppm) from C2,5 (71-73 ppm). It managed to distinguish the amorphous C4 and C6 (85 and 62 ppm respectively) from The crystalline C4 and C6 (89 and 65 ppm respectively).
Fig. 4
1H-13C CP-MAS NMR, Unlabelled BC gel, SSNMR 400 MHz, 6 KHz, 25 ° C
The CP kinetics result showed that the CP efficiency was very high for all the carbons, with them reaching the peak of magnetisation transfer around 1250-1000 us. The Carbon 6, both amorphous and crystalline, had especially high CP efficiency with maximum magnetisation reached at 250 us. The CP efficiency was relatively unchanged between the BC gel and the BC- hemicellulose composites. The change in temperature from 20 ° C to 5 ° C showed slight increase in CP efficiency and effect was relatively the same for all the hydrogels.
Graph 4: CP kinetics of C1 XG under 25 °C and 5 °C
CP/HPDEC Data Fitting. The results of comparing CP to HPDEC showed that there was a 14% reduction in Crystallinity in BC-XG compared to BC gel while the decrease was only 3% in crystallinity in BC-AX compared to BC gel. This level of Crystallinity is consistent with the current literature on this subject (Citation)
Fig. 5
Superimposition of HPDEC spectra of BC gel and BC-XG composite under 25 °C
BC BC-XG BC-AX
Degree of Crystallinity 77% 63% 74%
Investigation of hydration profile using WPT. The graphs obtained from WPT show that all the Carbons under BC-XG exhibited significantly more solvation than BC and BC-AX. BC-AX was slightly more solvated than BC. Also interestingly, BC-XG has a curve that exhibits first order kinetics as opposed to BC and BC-AX which seem to operate under zero order kinetics. Also BC-XG clearly operates on three phases. An early rapid phase in water polarisation, a more gradual increase around 100 (?) and a plateau after wards.
Relaxation time measurement. The T1 of Carbon for the BC gel under CP conditions was around 2.1 seconds and was homogenous for all carbons, same result with BC-XG. The T1 of Carbon of BC gel fell from around 2.1 s under 25 °C to around 1.4 s under conditions of 5 °C. When the T1 was measured under T1 IR, the results were far more heterogenous between carbon to carbon in each gel (C6’ had the lowest T1 for each gel). Also differences arose between each gel. BC-XG had, in the range of 2-4 seconds, higher T1 than BC gel. In comparison, BC-AX had negligible differences between it and BC gel.
Carbon BC BC-XG
C1 2.202 s 2.121 s
C4 2.175 s 2.163 s
C4’ 1.961 s 2.186 s
C3 2.159 s 2.125 s
C2,5 2.149 s 2.167 s
C6 2.101 s 2.116 s
C6’ 2.054 s 1.899 s
Fig. 6
Table of T1 CPsatrec of Carbon for BC and BC-XG under 25 °C
Carbon 25 °C 5 °C
C1 2.202 s 1.423 s
C4 2.175 s 1.485 s
C4’ 1.961 s 1.349 s
C3 2.159 s 1.395 s
C2,5 2.149 s 1.437 s
C6 2.101 s 1.426 s
C6’ 2.054 s 1.020 s
Fig. 7
Table of T1 CPsatrec of Carbon for BC under 25 °C and 5 °C
Carbon BC BC-XG BC-AX
C1 7.7 s 9.2 s 7.2 s
C4 8.5 s 10.4 s 8.0 s
C4’ 4.7 s 8.9 s 5.0 s
C3 5.3 s 6.3 s 5.2 s
C2,5 6.6 s 6.9 s 6.6 s
C6 7.3 s 9.2 s 7.6 s
C6’ 4.1 s 6.8 s 4.4 s
Fig. 8
Table of T1IR of Carbon for BC, BC-XG and BC-AX under 25 °C
Discussion
The relaxation times of the gels were roughly similar between gels under CP conditions. They were also similar between carbon to carbon as well. When relaxation time was measured under direct polarisation differences started to appear which showed that BC-XG had on average 2-4 s more T1 compared with BC and BC-AX. The experiment also showed that C6’ had the T1 on average in all three gels (that could be explained by low amount of data from C6’ as very little of it was detectable in T1IR). Also another important part of the experiment is that it enabled the identification at which point the BC carbons are situated at the T1 graph. T1 relaxation graph is approximately a parabola where on the left side of the graph the less molecular motion the higher T1 because the more rigid a molecule, the less favoured is the kinetic equilibrium between the excited state and relaxed state. Thus making for longer Relation time. There is a minima for molecules which are medium tumbling molecules and then on the right side of the curve the smaller and faster a molecule becomes the higher the T1 becomes because the smaller the molecules becomes, the easier it is for something to stop its motion. In a system where molecules are sufficiently small. The average of the motion starts to decrease hence increasing T1. To know on which side of the parabola BC gel is. The experimented was conducted in Both 25 °C and 5 °C. The experiments showed that as you move from 25 °C to 5 °C (hence increasing rigidity) that there is a decrease in T1. Meaning that that BC gel is situated on the right side of the T1 graph. That would indicate that when BC-XG has its T1 increased, it does so because of it being less rigid, freer and more disordered. However, Experiments of the same type need to be conducted on BC-XG and BC-AX just to make sure that the composites exhibit the same kind of behaviour as BC gel.
The CP kinetics for all three compounds have the same behaviour, meaning that all three have the same CP efficiency. What is puzzling about this sample is that our results suggest that the CP kinetics for all three samples exhibit very high CP efficiency. This is surprising given that the gels are 99.9 % water. However the rigidity of the system might explain why something that is 99.9% water has solid-like rheological properties. The decay of all three samples was slow meaning that proton spin diffusion was high. That is also consistent with the fact the gel is 99.9% water. Also what is really strange is that the rigidity of C6 (both amorphous and crystalline) are the highest for the Cellulose hydrogel which is odd since that is the carbon normally most mobile.. CP kinetics when done under lower temperature exhibits small change in rigidity (becoming a bit more rigid). Meaning that the compound is already quite rigid.
The problem with these findings is that our other experiments that we performed seem to have results that, at least superficially, contradict the CP kinetics findings. First of all, the AFM studies that were conducted on the BC-XG gel show that it is way softer and the fibres are 10 times thinner. Also the Water polarisation studies that BC-XG had a significantly higer hydration profile with different behaviour c\cite{Pongsawatmanit_2006}\cite{Wise_1955}ompared to BC and BC-AX (Altough BC-AX showed significant increase in hydration, the shape of the curve was the same with BC gel). Hydration profile done under WPT can be increased by two things. Chemical exchange, which is the same for all compounds, and Coherent spin diffusion: mediated by distance- and mobility-dependent 1H-1H dipolar couplings.There is no reason to suggest that there is a decrease in distance 1H-1H dipolar couplings (the proposed mechanisms of how Xylogl\cite{Uhlin_1995}can and Arabinoxylan interact with Cellulose don’t mention a change in distance between Cellulose and its surroundings Citations required). Therefore the only possible explanation is that the mobility got increased in BC-XG and BC-AX. Also the CPH/HPDEC normalised Data show that Xyloglucan had a significantly lower crystallinity compared with BC gel and that would also imply a lower rigidity. Now of Course there are other factors that influence CP efficiency apart from rigidity and P15 (spin-lock power levels, decoupling power and the rate of magic-angle spinning (MAS). However we made sure during the experiment that these were consistent throughout and the same for all the gels.
We propose, in this paper, two ways in which harmonise the findings from CP kinetics with all our other findings and the wider literature. First, we suggest a Model of how xyloglucan breaks down the interfibrillar connections, increasing amount of amorphous regions in the cellulose while also increasing the surface area. Thus increasing the amount of water bound to cellulose at the surface. Water might be closer to the surface of cellulose than it was when the cellulose was rigid because the hydration profile for xyloglucan got better. This model is based on the idea that the bundles normally found in cellulose due to the interfibrillar connections are broken hence there is decrease in rigidity of the macrostructure but not of the microscopic domains. Xyloglucan may break interfibrillar connections by preventing Hydrogen bonding between Cellulose microfibril. This may help each microfibril to slide during cell enlargement, because xyloglucans have no mutual affinity (13). This would also explain the increased flexibility found in BC-XG composite.
Fig. 9
Model of how Xyloglucan and Cellulose Microfibril interact (14)
Fig. 10
Model to explain how Xyloglucan composite can have higher hydration
Second part of the explanation is in terms of CP limitations. The limitation is that CP targets the rigid components of the culture but we know cellulose is multi phased system with plenty of paraccystaline, amorphous (with different meaning of amorphous) and crystalline. The fact that CP kinetics that we got are standard solid like behaviour even though the sample is 99.99% water points out that there are mobile components that are not being detected by our experiments. This would also explain the homogenising effect of CPsatrec and how if you take CP out of the equation then you observe the heterogeneity found inside the gel. It also explain why the HPDEC data gives so different results from CP and explains why when we normalise CP with HPDEC we get results of the gel’s crystallinity that are consistent with the larger literature. Now for a full understanding of the BC hydrogels, there needs to be methods of identifying the disordered regions of the Compound. We propose methods of circumnavigating the problems associated with CP. CPSP (Cross polarisation with a single pulse (SP)) was developed to to quantitatively analyse compositions of multiphase polymers with marked difference in phase mobility. The quantification of CP spectrum is done by measuring the enhancement factor of CP with a reciprocity relation, while the quantification of SP is derived directly from the Boltzmann distribution in thermo equilibrium. CPSP is done by using a conventional CP combined with a SP sequence. The initial 13C 90° pulse rotates the magnetization to the x-y plane in the rotating frame and the magnetization is then spin-locked by the subsequent 13C CP pulse. Here, the 13C spin-locking pulse plays two roles: one is to hold the magnetization so that it will not dephase during the contact time t and the other to transfer polarization from 1 H to 13C. (15)
Fig. 11
CPSP pulse sequence
Another experiment that is relevant is Cross Depolarisation experiment (CDP). This experiment is needed for full quantification of CPSP(16). This is a method where, instead of the polarized S magnetization in the standard CP-MASS experiment, the depolarized S magnetization is monitored. The pulse sequence utilized is shown below. It is a Conventional CP sequence with some modifications. Two additional intervals ti; and tc 2 are inserted between the usual polarization time tcI, and detection time td. This is very good at revealing the details of cross relaxation process. With CP it is difficult to know the real maximum polarized S signal and the exact time dependence of S signal to be determined by cross polarization alone, because the creation of S polarization depends not only on the cross relaxation process itself, but also on some other factors, such as T~~ of I spin, and Hartmann-Hahn matching condition. In the depolarization scheme, however, the finite initial S magnetization at the beginning of t, 2 will certainly reach zero at last and the turning point occurs during the first tens of usec. This experiment allows for the quantitative expression of 13C magnetisation as a function of depolarisation time using the following formula(17):
Msx is the polarized 13C magnetization for 13CH group, b is dipolar coupling between ' 13C and its directly bonded proton, R is the spin diffusion rate among protons, and t is the polarization time. Both CPSP and CDP should tackle two problems. First, they should be able to allow us to make quantitative judgements on CP experiments and second, to able to target the amorphous regions of BC gel and its composites.
Fig. 11
Pulse sequence utilized in the depolarization experiment, where tci is the polarization time for S spins, ti is the dephase time for I spins, and tc 2 is the depolarization time for S spins. The development of the 5 magnetization is schematized by a solid line.
Conclusion
• To conclude, SSNMR has provided lots of insights into the structure of BC Hydrogels. The implementation of CP kinetics and WPT along relaxation time experiments showed that these were gels that have a high degree of CP efficiency. That the CP efficiency is equivalent between Hydrogels was equivalent and that a decrease in temperature has minimal effect on the CP efficiency. The WPT show zero order kinetics for the graph of both arabinoxylan and Bacterial Cellulose but higher hydration but also a curve operating under 1st order kinetics for Cellulose-Xyloglucan composite. The T1 of Bacterial cellulose and Arabinoxylan composites are practically the same but the Xyloglucan composite has a difference between 2-4 s between its T1 and the T1 of BC gel. There is also difference in T1 between different carbons T1 suggesting that there are different levels of mobility between each carbon in each sample. All these insights into the Structure of BC Hydrogels was done using SSNMR Techniques. Nevertheless, SSNMR has important limitations that need to be considered. Namely, Poor acquisition time and reliance of CP for most of the experiments and that makes the results qualitative and they also target the rigid components of the system and has homogenising effect on the less ordered components of it. Making them appear the same. We propose methods of how to tackle these limitations through CPSP and CDP which will make Cross polarisation experiments quantitative and will allow for detection of disordered materials as well. With the goal of formulating a model that incorporates both the crystalline regions and the amorphous regions as well.
Acknowledgement
References
Sorted By Appearance
1. Rajai H. Atalla, John W. Brady, James F. Matthews, Shi-You Ding, Michael E. Himmel. Structures of Plant Cell Wall Celluloses. 188–212 In Biomass Recalcitrance. Blackwell Publishing Ltd. Link
2. Daniel J. Cosgrove. Plant cell wall extensibility: connecting plant cell growth with cell wall structure mechanics, and the action of wall-modifying enzymes. Journal of Experimental Botany 67, 463–476 Oxford University Press (OUP), 2015. Link
3. E. M. Lord, J.-C. Mollet. Plant cell adhesion: A bioassay facilitates discovery of the first pectin biosynthetic gene. Proceedings of the National Academy of Sciences 99, 15843–15845 Proceedings of the National Academy of Sciences, 2002. Link
4. Nicholas C. Carpita, David M. Gibeaut. Structural models of primary cell walls in flowering plants: consistency of molecular structure with the physical properties of the walls during growth. The Plant Journal 3, 1–30 Wiley-Blackwell, 1993. Link
5. Ray F. Evert, Susan E. Eichhorn. Plants and People. 501–523 In Raven Biology of Plants. Macmillan Learning, 2013. Link
6. Kiyoshi Suzuki, Takao Itoh. The changes in cell wall architecture during lignification of bamboo Phyllostachys aurea Carr.. Trees 15, 137–147 Springer Nature, 2001.Link
7. Fang Lin, Jeffrey M Catchmark. Analysis of Gluconacetobacter xylinus exopolysaccharides and its impacts on bacterial cellulose production and crystallization. In2012 Dallas Texas, July 29 - August 1, 2012. American Society of Agricultural and Biological Engineers, 2012. Link
8. Marta Martínez-Sanz, Michael J. Gidley, Elliot P. Gilbert. Hierarchical architecture of bacterial cellulose and composite plant cell wall polysaccharide hydrogels using small angle neutron scattering. Soft Matter 12, 1534–1549 Royal Society of Chemistry (RSC), 2016. Link
9. Marta Martínez-Sanz, Deirdre Mikkelsen, Bernadine Flanagan, Michael J. Gidley, Elliot P. Gilbert. Multi-scale model for the hierarchical architecture of native cellulose hydrogels. Carbohydrate Polymers 147, 542–555 Elsevier BV, 2016. Link
10. Natalia Pérez-Hernández, Trung Quan Luong, Martín Febles, Carlos Marco, Hans-Heinrich Limbach, Martina Havenith, Cirilo Pérez, María Victoria Roux, Ricardo Pérez, Julio D. Martín. The Mobility of Water Molecules through Hydrated Pores. The Journal of Physical Chemistry C 116, 9616–9630 American Chemical Society (ACS), 2012. Link
11. Waclaw Kolodziejski, Jacek Klinowski. ChemInform Abstract: Kinetics of Cross-Polarization in Solid-State NMR: A Guide for Chemists. ChemInform 33, no–no Wiley-Blackwell, 2010. Link
12. Agnieszka Kaflak, Waclaw Kolodziejski. Kinetics of1H \(\rightarrow\)31P NMR cross-polarization in bone apatite and its mineral standards. Magnetic Resonance in Chemistry 46, 335–341 Wiley-Blackwell, 2008.
13. Hayashi, T. , Marsden, M. P. F. , Delmer, D. P. 1 987. Pea xyloglucan and cellulose. V. Xyloglucan-cellulose interactions in vitro and in vivo. Plant Physioi. 83:384-89
14. Hayashi, T. (1989). Xyloglucans In The Primary Cell Wall. Annual Review of Plant Physiology and Plant Molecular Biology, 40(1), p.156.
15. Shu, J., Li, P., Chen, Q. and Zhang, S. (2010). Quantitative Measurement of Polymer Compositions by NMR Spectroscopy:Targeting Polymers with Marked Difference in Phase Mobility. Macromolecules, 43(21), pp.8993-8996.
16. Xiaoling, W., Shanmin, Z. and Xuewen, W. (1988). Two-stage feature of Hartmann-Hahn cross relaxation in magic-angle sample spinning. Physical Review B, 37(16), pp.9827-9829.
17. S. R. Hartmann and E. L. Hahn, Phys. Rev. 128, 2042 (1962)
18.