The proportion of the crystalline to amorphous cellulose for the glucose carbon 1 to carbon 6 transfer appears to decrease more for the 2-minute milling period to 80.28 ± 1.36% and 74.65 ± 1.39%, respectively.The signal intensity for the crystalline cellulose and amorphous cellulose appear to be nearly equal for the glucose carbon 1 to carbon 6 after 15 minutes of milling.In the future, dephasing curves could be applied to specifically probe cellulose fiber substructures with the 2D CP-PDSD experiment similar to previous studies on the plant cell walls.Dephasing curves would give more specific mixing times for the experiment between 30 ms and 1500 ms mixing times of the 2D CP-PDSD experiments.However, for the purpose of exploring the recalcitrant implications of milling where cellulose fibrils are disrupted to accommodate particle sizes smaller than the fibril width 1 μm or length 100 μm, other experiments may be also explored.Future use of the 30 ms CP-PDSD experiment may be more suitable for cellulose morphology based on carbon 4 as sizes of cellulose fibrils are reduced and long range CP mixing times becoming more inefficient.After milling,vertical garden indoor system cellulose fibrils are still present in the morphology of the sample and there is not a reduction in the crystalline cellulose cross-peaks in the 30 ms 2D CP-PDSD.This means that the crystalline morphology of cellulose within the fibril is not directly converted to amorphous cellulose as suggested by analysis milling pure cellulose fibril.
The atomic structure of the cellulose fibrils appears to be maintained after milling based on the 30 ms CP-PDSD cellulose one carbon correlations, for example the crystalline cellulose to amorphous cellulose signal intensity ratio was maintained for carbons 4 and 6.Cellulose carbon 4 signals are particularly important because it can offer insight into the cellulose 1-O-4 polymer backbone of cellulose and chemical shifts which indicate the environment interior through exterior of the cellulose fibril.Signals in the 2D CP-rINADEQUATE arise from highly immobilized polymers in the secondary plant cell wall structural hemicellulose and cellulose.Interpretation of these signals is less straightforward due to resonance overlap for these carbohydrates, but the signals pertaining to the carbon 4 chemical shifts of glucose within cellulose are well resolved.The CP-rINADEQUATE shows directly bonded nuclei and is useful for identifying specific subtypes of prominent monomers within the secondary plant cell wall.The chemical shifts of the bonded nucleisum to another peak in the indirect dimension shown in the y-axis.The single quantum chemical shifts were assigned using previously established values of sorghum secondary plant cell wall polymers and were verified using cross peaks.Monomers were traced out using cross peaks because the y-axis double quantum coherence is assigned by a chemical shift from the sum of chemical shifts from two directly bonded 13C spins.Note, the monomers found by the rINADEQUATE are contained within the polymers within the secondary plant cell wall and shifted downfield compared to literature of dry solid monomers and monomers in water.Many of the unambiguous signals of cellulose all reduced in signal intensity : after 2 minutes of milling the signals reduced by 15% for all unambiguous amorphous cellulose peaks.
After 15 minutes of milling the crystalline cellulose signal associated with interior crystalline cellulose within the dehydrated space around 84 ppm of the cellulose fibril decreased nearly to the noise.The signal reduction of the crystalline and amorphous cellulose appeared to be equal suggesting crystalline cellulose did not convert to amorphous cellulose when milled, contrary to pure cellulose as found in Ling et al.2019.However, the remaining crystalline cellulose carbon 4 does not appear to significantly differ in the proportions of decreasing amorphous cellulose 4 for the interior and exterior positions, as noted in Gao et al.2020.In Figure 9, additional subscripts i to iii refer to the least to the most dehydrated cellulose i.e.the interior to exterior polymers in the cellulose fibers based on previous studies.Results of the current study show overall signal reduction of the amorphous cellulose after 2 minutes is only around 3% and by 75% after the 15 minutes of milling.Beyond the carbon 4 area of the CP-rINADEQUATE spectrum, cellulose signals often overlap with rigid hemicellulose.The total amount of cellulose estimated by the glucose carbon 1 peak at 105.2 ppm overlaps with many of the hemicellulose carbon 1 peaks including structurally relevant 2-fold xylanhemicellulose at 105.1 ppm.The carbon 1 of cellulose can be less quantitative in the CPrINADEQUATE compared to the CP-PDSD experiments discussed here.The slice from the carbon 1 region of the CP-rINADEQUATE supports an overall cellulose signal reduction in Figure 9G which is apparent in the overlay of the control, stems milled for 2 minutes, and stems milled for 15 minutes in Figure 9D.To summarize, there is an overall signal reduction and the proportion of amorphous and crystalline cellulose appear to be maintained after milling in the CP based experiments CP-rINADEQUATE, 30 ms and 1500 ms 2D CP-PDSD experiments.
Additional evaluation of the signals using normalized intensities of carbon 1 for the 30 ms CP-PDSD experiment could offer further statistical insight on cellulose within the sample based on changes in monomers, without handling as much signal overlap as the CP-rINADEQUATE.As the cellulose structure is further milled beyond the 1 μm particle size, the amount of information which will be available in the 1500 ms CP-PDSD experiment is expected to decrease and may be removed from the experiment set for more informative experiments.CP becomes less efficient for more liquid like samples and the 1500 ms CP-PDSD has a mixing time suited for distances potentially exceeding cellulose fibril lengths as they are broken down.Overall, the argument of increased recalcitrance due to conversion of crystalline to amorphous cellulose using milling is not supported by the CP-PDSD and CPrINADEQUATE spectra.However, understanding the available digestible cellulose may be more informative especially as hemicellulose and lignin could sit on the surface of cellulose fibrils and reduce available surface area for digestion.Signal intensities from arabinose all significantly decrease by at least 30% after ball-milling while 3-fold xylancarbon 4 and carbon 5 signal intensities increase after 2 minutes.For the stems milled for 15 minutes,mobile vertical grow racks the arabinose intensities decrease by >87%, but are still more prominent than cellulose.After 15 minutes of milling, xylan peaks remained consistently 10% above the relative intensity of other peaks in the CP-rINADEQUATE , while cellulose signals frequently dropped below 10% of the relative signal intensity compared to the control.Arabinose serves as a common substitution and a defining feature dictating some xylan interactions with cellulose and is known for facilitating cross linking between hemicellulose and lignin.The hemicellulose signals appear to remain intact in the 2D CP-rINADEQUATE, although reduce in signal intensity after 15 minutes of ball-milling, meaning that polymers maintain their rigid conformations while the cellulose signals area reduced in the spectra.Hemicellulose signals appear to be more preserved when contrasting overall spectra of the control to the stems milled for 15 minutes , supporting the structural importance of hemicellulose.As highlighted in Figure 9C, the arabinose substitutions on rigid hemicellulose signals persist with decreasing intensity in the spectra as presented in the 2D CP-rINADEQUATE overlay.Observation of xylan hemicellulose in the CP-rINADEQUATE has the same issue of signal overlap as cellulose and only a select number of unambiguous chemical shifts , but tracing the full monomers within the spectra helps with some deconvolution.Another note is the predominant xylancellulose interaction between the 3-fold xylan and the amorphous cellulose surfaces in sorghum: it is not necessarily expected to increase as amorphous cellulose did not appear to increase.The 2fXylan that interacts with crystalline cellulose surfaces exhibits heavy signal overlap and has a lower prominence in sorghum secondary plant cell walls.Findings from this study appear to confirm that the use of filtered DP-rINADEQUATE could assist in detecting semirigid hemicellulose structures relative to cellulose.It is inconclusive whether hemicellulose is creating recalcitrant interactions in the current 2D CP-rINADEQUATE, however remaining structural hemicellulose is more rigid than cellulose in the secondary plant cell wall.
The line shapes are generally narrower for the dynamic polymers captured in the 1H-13C 2D rINEPT experiment.To better understand how this is possible, note that spins have an equilibrium aligning their precession along the applied magnetic field and realign their precessions through longitudinal and transverse relaxation to the applied magnetic field after RF pulses.In solid-state NMR this relaxation of spins back to equilibrium is typically limited by T2 relaxation.In the experiment dynamic peaks have a greater signal intensity thanks to the long periods between pulses which captures the long relaxation dynamic spins in liquid-like samples and the signals arising from short T2 relaxations predominating rigid spins are not captured.So, some heterogeneous line broadening from rigid signals and dynamic signals overlap at similar chemical shifts in the rINEPT.And as the more dynamic signals are captured, narrower lines are also observed because the random motion of the spins allows some proton chemical shifts to be canceled out.Directly bonded 13C and 1H are probed through J-coupling delays between pulse excitations for each nucleusand Matlahov and van der Wel.The assignment of the 1H-13C 2D rINEPT was possible given the chemical shifts of polymers identified in the sorghum plant cell wall by Gao et al.2020and previous work examining pectic hemicelluloses in Wang et al.2014.It is worth noting that in this work, the arabinose at 108 ppm was confirmed using the control DP-rINADEQUATE experiment.Historically, the monomer chemical shifts are up field of the polymer chemical shifts, computational modeling in silico was used to confirm chemical shifts of xylan and its symmetry, and polymer chemical shifts tend to remain consistent between solid state NMR experiments even if line broadening needs to be considered.Some challenges are associated with identifying hemicellulose in the carbon 1 region of the 2D rINEPT as the peaks aside from Arabinose coincide with chemical shifts assigned for pectin and hemicellulose found in the primary plant cell wall, which is disputed in the secondary plant cell wall and it is challenging to distinctly separate polymers thus far between regions of the primary and secondary plant cell wall.For these reasons, full interpretation of the 2D-rINEPT beyond a fingerprint requires further research.Using general chemistry knowledge of 1H chemical shifts information the carbon 1 region was supported by the ether chemical shifts typical around 5 ppm consistent with sugar backbones.As seen in Figure S3 for the freeze-thaw experiment, the protons of the secondary plant cell wall range from aliphatic shifts to protons in conjugated aromatic systems in the 1H range of 3 ppm to 10 ppm for the plant cell wall polymer 13C chemical shift range of 170 to 60 ppm.Perhaps future work with 13C-13CrINADEQUATE or more dynamic J-coupled experiments could extract information on the number of sub environments monomers, polymers, and substitutions of hemicellulose.This information would be particularly useful in deconstruction to assess the changes in 1H chemical shifts after preprocessing methods such as oxidation and acid treatment used in plant cell wall samples to solubilize cellulose.Cross peaks are not assigned in the 2D rINEPT for 1H because they are continuing to be assigned for the plant cell wall and the correlation of 1H to 13C peaks only accounts for directly bonded nuclei, not double quantum coherences as seen in the CP-rINADEQUATE Figure 9C.The 1H-13C 2D rINEPT fingerprintof highly dynamic hemicellulose generally changes over the milling period.Carbon 1 of arabinose hemicellulose substitutions demonstrate a change in the number of proton environments over the milling process and a reduction in signal intensity.The carbon 1 of arabinose substitutions of hemicellulose decreases to 39.35 ± 3.92% and pectic hemicellulose to 29.62 ± 1.62% after 2 minutes of milling.Arabinose carbon 1 of arabinose substitutions on structural hemicellulose 14.81 ± 2.23% and pectic hemicellulose 5.69 ± 0.92% after 15 minutes of milling.This interpretation of lost signals in Figure 10 suggests that some highly dynamic lignin and hemicellulosecould become less dynamic.Further, the reduction of arabinose hemicellulose substitution signals in the CP-rINADEQUATE and in the rINEPT of polymers together suggest that arabinose, and arabinose substituted hemicellulose takes on a more intermediate dynamic in the plant cell wall.Tracking arabinose substitutions may show a potential marker for rigid and dynamic hemicellulose changes related to recalcitrance.Additionally, the recalcitrance hypothesis based on structural hemicellulose maintaining their rigidity is supported as signals are stronger for arabinose and 3f,AXylan remain in the CP-rINADEQUATE experiment.Reduction in lignin signals would indicate a loss of mobility associated with lignin condensation and emerging lignin cross-linkages are noted in the rINEPT experiment.