Preparation and Characterization of Xylan Derivatives and Their Blends

Although hemicellulose is found widely in nature, it is currently under-utilized as a raw material for commercial applications. It would be desirable to find new uses for hemicellulose in order to add value to this agro-based material. A common type of hemicellulose is xylan, which is found in a number of wood species and in cotton. In this work we prepared cationic and anionic xylan derivatives and characterized them by ¹³C NMR, FT-IR, size exclusion chromatography (SEC), thermal analysis, and rheology. In particular, the ¹³C NMR spectra of carboxymethyl xylan (CMX) and quaternary ammonium-adducted xylan (QAX) were fully assigned with the help of samples with different degrees of substitution. SEC indicated that the beechwood xylan showed a bimodal molecular weight distribution, but with derivatization the distribution tended to become unimodal. Thermal analysis and rheology studies did not uncover any surprises; the solution of xylan and its derivatives exhibited mostly Newtonian behavior. The blends of CMX and QAX produced a precipitate at almost all ratios, indicating the formation of a polyelectrolyte complex. When cationic and anionic xylan samples were added together to paper, the paper dry strength increased. Thus, the combination of cationic/anionic xylan may be of interest in selected applications.

Xylan; 13C NMR; Anionic xylan; Cationic xylan; Polyelectrolyte complex; Dry paper strength

Along with cellulose and starch, xylan is among the most abundant polysaccharides on earth. It comprises mostly of d-xylose units linked as β-1,4 linkages. Typically it constitutes 10-35% in hardwoods and 10-15% in softwoods. The xylan backbone is usually substituted with other sugars, but the types of substitution vary. The most common substitutions in hardwoods are acetyl and 4-O-methylglucuronosyl, whereas in softwood they include arabinosyl and 4-O-methylglucuronosyl [[1] , [2] ].

A popular approach to produce new products from xylan is to derivatize it in order to impart additional functionalities and to improve its properties [[2] -[4] ]. For example, the synthesis of carboxymethyl xylan (CMX) has been reported in several publication, e.g., Petzold et al. [[5] ], Alekhina et al. [[6] ], Saghir et al. [[7] ], Peng et al. [[8] ], Ren et al. [[9] ], Simkovic et al. [[10] ], Konduri and Fetehi [[11] ], Hettrich et al. [[12] ]. Quaternary ammonium adducts of xylan have been made and reported by many workers, including Schwikal et al. [[13] ], Ren et al. [[14] ], and Kong et al. [[15] ]. Amphoteric xylan containing both carboxymethyl and quaternary ammonium groups have also been reported [[16] ]. The effect of cationic xylan on pulp has been investigated [[17] , [18] ].

Despite keen interest in xylan and its derivatives in recent years, it is currently still under-utilized as a raw material for commercial applications. It would be desirable to explore new uses for xylan in order to add value to this abundant agro-based material. In this work, we have derivatized beechwood xylan to produce CMX and quaternary ammonium adduct of xylan (QAX). The 13C NMR spectra of these polymers have been re-examined and improved spectral assignments obtained. The properties of the resulting blends have been examined, including their effects on paper strength. As far as we know, studies of the blends of CMX and QAX have not been reported before.

In the literature, there have been many papers reporting mixtures of a cationic polysaccharide and an anionic derivative of a different polysaccharide, e.g., alginate-chitosan [[19] , [20] ], anionic starch-chitosan, and cationic starch-pectin [[21] ], cationic starch-carboxymethyl cellulose [[22] ], and quaternized hemicellulose and carboxymethyl cellulose [[23] ]. There have been fewer examples of the blends of cationic and anionic derivatives of the same polysaccharides; some examples include cationic and anionic starch [[24] , [25] ], cationic and anionic cellulose [[26] , [27] ], and cationic and anionic guar [[28] ]. The blends of cationic and anionic xylan fall in this latter category.

Beechwood xylan, hydrochloric acid, sodium monochloroacetate and 1,2-dimethoxyethane were all acquired from Sigma Aldrich, St. Louis, MO, USA. Sodium hydroxide, 2-propanol, ethanol, and dialysis tubing were all purchased from Fisher Scientific, Fair Lawn, NJ, USA. Glycidyltrimethylammonium chloride (QUAB® 151) was purchased from SKW QUAB Chemical, Inc., Saddle Brook, NJ. The starch samples (Cato® 255 and Opti-Pro® 650) were obtained from Ingredion Inc., Westchester, IL, USA.

The procedure was adapted from previously reported methods [[5] , [6] ]. Beechwood xylan (5 g) was dissolved in 25 mL of a 25% aqueous NaOH (6.25 g) solution. To the slurry media, 2-propanol (35 mL) was added. The reaction mixture was vigorously stirred for 30 min at 30 °C. Then sodium monochloroacetate (0.25-9 g) was added and the temperature of the reaction bath raised to 65 °C for 70 min. The reaction mixture was neutralized with dilute acetic acid and the polymer precipitated with ethanol. The polymer was washed once with 100 mL of 65% aqueous ethanol and four times with 100 mL of absolute ethanol. The product was dried at 60 °C in a vacuum oven and the weight of the product was recorded.

The synthesis followed the method reported by Schwikal et al. [[13] ]. Beechwood xylan (5 g) along with 17.5 mL of distilled water was heated under reflux for 15 min. The mixture was cooled to room temperature and a solution of 0.9 g NaOH in 5 mL distilled water was added to the solution. After 30 min at room temperature, 1,2-dimethoxyethane (35 mL) was added to the dissolved xylan, and 7.0-24.6 g of 70% glycidyltrimethylammonium chloride (GTMAC, also known as 2,3-epoxypropyltrimethyl ammonium chloride) were added dropwise to the solution. The mixture was stirred for 24 h at room temperature. The solution was neutralized with dilute HCl and the product precipitated with 200 mL of absolute ethanol overnight. The product was dried in a vacuum oven, reconstituted and then dialyzed using a membrane with a molecular weight cutoff of 3500 g/mol in water for approximately 2 days. The product was dried again and then weighed.

All NMR spectra were obtained on a Bruker 500 MHz NMR spectrometer at room temperature with D2O as a solvent. 13C NMR spectra (125.8 MHz) with broadband 1H decoupling were recorded using a pulse sequence with 30° pulse angle and 3 s recycle delay. Standard instrumental setups were used for all other NMR experiments. All 13C chemical shifts were referenced to tetramethylsilane at 0 ppm.

Fourier-transform infrared (FT-IR) spectroscopy was conducted on solid samples using a Bruker Vertex 70 spectrometer (Billerica, MA, USA) coupled with the Opus software. Each spectrum was scanned in the frequency range of 650-4000 cm−1. Typically 64 scans were taken at a spectral resolution of 8 cm−1.

The xylan and CMX were analyzed using a procedure similar to those reported before for neutral water-soluble polysaccharides and carboxymethyl cellulose with a low degree of substitution (DS) [[29] -[31] ]. The instrument used included a Waters HPLC system (Milford, MA, USA), with a Waters 515 HPLC pump, a Waters 717 plus autoinjector, a Waters column oven and a Waters 2414 refractive index detector. Two Tosoh Haas TSK-Gel GMPWXL columns were attached in series, operating at 30 °C. The system was controlled by Waters Empower 3 software. The mobile phase used comprised 50 mM lithium acetate at pH 8.4. Samples of neutral polysaccharides were prepared for analysis by dissolving 10 mg of polymer in 20 mL of mobile phase (0.05%) with overnight tumbling. The analyte solutions were filtered through a 0.45-µm Nylon membrane prior to analysis. For each analysis 100 µL of the analyte solution was injected into the unit. The flow rate was 0.5 mL/min and the run time 75 min per sample. Calibration was made using poly(ethylene oxide) narrow molecular weight standards.

The cationic derivatives were analyzed using a similar procedure as above, except that the neutral and anionic polysaccharides were analyzed under slightly alkaline conditions, whereas the cationic polysaccharides were analyzed under somewhat acidic conditions. Thus, the mobile phase comprised 300 mM lithium acetate at pH 4.8. Another difference was that the columns and the refractometer were operated at 40 °C.

The measurements were conducted on an Anton Parr rheometer (MCR 102, Ashland, VA, USA). The steady-state viscosity was recorded with a cone-and-plate geometry (cone angle: 20; diameter: 40 mm; truncation: 56 µm) and a Peltier unit controlled at 25 °C. Flow curves were obtained on 5% solutions of the samples. The shear rate was gradually increased from 1 to 1000 s−1 at 1-s increments.

Thermogravimetric analysis (TGA) was performed under nitrogen (Model Q500 analyzer, TA Instruments, New Castle, DE, USA). Sample masses between 3.0 and 8.0 mg were heated from 25 to 600 °C at a rate of 10 °C/min. Data were processed with TA Instruments Universal Analysis 2000 software and with Excel spreadsheet. Differential scanning calorimetry (DSC) was carried out on a Model Q20 instrument (TA Instruments), using stainless steel pans at a temperature from 0 to 300 °C and a heating rate of 10 °C/min under nitrogen.

For the testing of dry strength of paper, we followed a procedure adapted from an ASTM method (D 828-97) for paper testing [[32] ]. A similar method was reported previously by several workers on the effects of other additives on paper [[33] -[35] ]. In this test, Whatman filter paper #1 (24 cm diameter) was cut into seven 2.54 cm × 15.24 cm rectangular strips and coated with different polymer solutions. The coating procedure entailed making up the solutions of CMX and QAX (at 5% w/w) with distilled water, and each solution was applied to the seven strips of paper using a soft paint brush; the paper strips were then air-dried. For the mixture of CMX and QAX, two procedures were used. In the first procedure, the paper strips were coated consecutively with CMX solution and QAX solution, and then air dried. In the second procedure, the paper strips were coated with CMX solution and air-dried; it was then coated with the QAX solution and then dried in air. In all cases, the weight of the xylan additives were obtained by taking the difference in dry weight of paper alone and paper with additives. For comparison, the same test was also conducted with two starch samples (Cato® 255 and Opti-Pro® 650 from Ingredion, Inc.). The starch solution were prepared by heating with stirring 1 g of starch with 8 g of distilled water at 90 °C for at least 15 min. The starch solutions (at room temperature) were then used to coat the paper using the brush procedure. A control sample was simply paper coated with distilled water.

All the test paper strips were heat-pressed at 120 °C and 0.25 MPa for 10 min (benchtop press, Model 3856, Carver Inc., Wabash, IN, USA). Tensile tests of paper were conducted on a Zwick tensile tester (Model Z005, Ulm, Germany) at a speed of 1 mm/min with a solid clamp in tension mode. Data collected included the tensile modulus and maximum tensile strength at break (both in MPa). Seven paper strips were tested for each formulation. Analysis of variance and a Tukey means comparison test (α = 0.05) were used to assess the differences in the paper’s mechanical properties.

Beechwood xylan was converted to CMX through the etherification reaction adapted from the literature [[5] , [6] ]. Several CMX samples made (Table 1) had DS values that varied from 0.1 to 0.9. The effects of NaOH and monochloroacetate on the DS on CMX have been previously reported by several workers [[5] -[12] ]. In agreement with previous work, the DS increased with increasing amounts of sodium monochloroacetate and NaOH added. In particular, through the use of a specific combination of monochloroacetate and NaOH addition, we could dial in the DS value desired.

Samples of CMX, synthesized through the reaction of beechwood xylan with monochloroacetic acid (MCA)

aMCA was added as the sodium salt; molecular weight was 116.5

bXylan molecular weight (on the anhydroxylose basis) was assumed to be 132

For the synthesis of QAX, we followed a procedure reported by Schwikal et al. [[13] ]. Several QAX samples generated through these processes are summarized in Table 2. As expected, the DS value could be directly correlated with the amount of GTMAC added to the reaction mixture. Rather high values of could be achieved (up to DS 1.9).

Samples of QAX, synthesized through the reaction of beechwood xylan with GTMAC

aGTMAC was supplied as 70% solution in water; the above weight was corrected for 0.7

bMolecular weight of GTMAC = 151.64

cXylan molecular weight (on the anhydroxylose basis) was assumed to be 132

The 13C NMR spectra of xylan and CMX are shown in Fig. 1. In agreement with the literature [[36] , [37] ], the 13C peaks for unreacted xylan (Fig. 1a) are found at 103 ppm (C1), 78 ppm (C4), 74 ppm (C3), 73 ppm (C2), and 63 ppm (C5). Smaller peaks at 98.8 ppm (g1), 83.5 ppm (g4), 73.6 ppm (g5), 72.6 ppm (g2), 60.8 ppm (gM, CH3O), 78.1 ppm (g3, overlapped with other peaks), and 177.4 ppm (g6, carboxylic acid, outside the spectral region) are related to the 4-O-methyl glucuronic acid that branches out of the xylan backbone [[36] ] (Scheme 1).

13C NMR spectra in the 50-110 ppm region of (a) xylan, (b) CMX with DS ~ 0.4, and (c) CMX with DS ~ 0.9. The letter g refers to carbons from methyl glucuronic ester; the numbers refer to carbon numbers on xylan (Scheme 1) and CMX (Scheme 2), and u and s denote unreacted and substituted carbons, respectively. The peaks labelled as x1 and x2 are due to isopropyl carboxymethyl ether and glycolic acid, respectively

Numbering scheme for xylan and the methyl glucuronic ester side chain

Numbering scheme for CMX (left) and quaternary ammonium adducted xylan (right). For illustration, the substituents are shown to be substituted at position 2 only. In the actual products, substitution occurs singly at position 2 or 3, and doubly at both positions 2 and 3

For CMX, Peng et al. [[16] ] had earlier assigned the peaks at 68.6 and 71.0 ppm to the methylene carbon in carboxymethyl moiety; the same assignment was made by Peng et al. [[8] ] and Ren et al. [[9] ], after correcting for the 13C shift reference. We have reviewed the 13C NMR assignments of the CMX spectra, using CMX samples with different DS values and the help of empirical 13C shift additivity relationships [[38] , [39] ]. In this way, more detailed assignments were obtained. Thus, for the monosubstituted xylan, the C2 and C3 xylan carbons are found at 83 and 84 ppm. The C2 and C3 carbons of the disubstituted xylan are found at about 81 ppm. The substituted C4 is found near the unsubstituted C4, near 76 ppm. The C6 carbon (CH2 of carboxymethyl) is found at 72 ppm. (The assignment of this C6 carbon has been confirmed with DEPT-135 experiment, not shown in Fig. 1.) The substituted xylan C1 and C5 resonances occur at about the same location as the unsubstituted location. In addition, there are peaks found in some spectra for isopropyl carboxymethyl ether (marked as x1 in Fig. 1) and glycolic acid (marked as x2), due to the reaction of MCA with isopropanol and water, respectively. These assignments are directly noted in Fig. 1. The numbering scheme for CMX carbons is given in Scheme 2.

The DS for CMX can be obtained from the 13C NMR data using two approaches. One approach is to take the area of the carboxyl peaks (C7 in Scheme 2) at ca. 180 ppm and divide it by the average of C1 area (at 103 ppm) and C5 area (at 63 ppm). Another approach is to take the average of (1) the combined peak area for substituted C2 and C3 (in the 79-85 ppm range, minus the peak area for g4 in 4-O-methyl glucuronic acid) and (2) the peak area for C6 (at ca. 72 ppm, minus the peak area for g2 in 4-O-methyl glucuronic acid). This average value is then divided by the average of C1 area and C5 area to give the DS for the second approach. The DS values calculated from these two approaches are then averaged, and the average is reported.

The 13C NMR spectra for QAX at different values of DS are shown in Fig. 2. With increasing DS, new peaks grow and become bigger. These include the peaks due to C2 and C3 on monosubstituted xylan (83 and 82 ppm respectively), and disubstituted xylan (at 81-82 ppm), substituted C1 at about 101.5 ppm (slightly upfield of C1), substituted C4 at 75.8 ppm, and substituted C5 slightly upfield from unsubstituted C5 at 63.6 ppm. As for the peaks from the GTMAC residue, C6 occurs at 74.5 and 71.5 ppm, C8 at 68.4 ppm, C7 at 66.0 ppm, and C9 (N-methyl) at 54.9 ppm. (The assignment of C6 has been confirmed with DEPT-135 experiment, not shown in Fig. 2.) It may be noted that the assignments were previously reported in the literature [[13] , [16] , [40] ]. In this work, the assignments are compatible with the previous work, except for the more detailed assignments given here for substituted xylan C2 and C3 carbons.

13C NMR spectra in the 45-110 ppm region of (a) QAX, DS = 0.2, (b) QAX, DS = 0.7, (c) QAX, DS = 1.4, and (d) QAX, DS = 1.9. The letter g refers to carbons from methyl glucuronic ester; the numbers 1-8 refer to carbon numbers on QAX (Scheme 2), and u and s denote unreacted and substituted carbons, respectively

For QAX, the DS can be calculated by using the peaks areas for substituted C2 and C3 (in the 79-85 ppm range), C7 (66.0 ppm) and C8 (68.4 ppm). The average value for these three areas is then divided by the average of C1 and C5 peak areas to give the DS.

The FTIR spectra for xylan, CMX, QAX, and CMX/QAX mixture are shown in Fig. 3. The xylan spectrum appears similar to those published in the literature [[8] , [11] , [40] -[46] ]. Thus, the broad peak at around 3400 cm−1 is due to O-H stretching, and the peak at 2900 cm−1 due to symmetric C-H vibration. The somewhat broad peak at around 1610 cm−1 comes from glucuronic functionality and water [[42] , [45] ]. The peak at 1460 cm−1 corresponds to CH and OH bending in xylan [[42] , [44] ]. The large peak at 1043 cm−1 corresponds to C-O stretching in C-O-C linkages [[40] , [42] ]. The peak at 901 cm−1 corresponds to C1-H bending and ring frequency modes in β-glycosidic linkages between the xylose units [[41] , [44] , [45] ].

FTIR spectra of (a) xylan, (b) CMX, DS 0.4, (c) QAX, DS 1.9, (d) blend of CMX and QAX (2:1 weight ratio)

For CMX, the O-H stretching band at 3400 cm−1 noticeably decreases due to the reaction of the hydroxy unit, and the peak at 1600 cm−1 becomes sharper due to the carboxymethyl group present [[9] , [11] , [16] , [45] ]. Analogous assignments can also be obtained from carboxymethyl cellulose [[23] , [46] ]. The peak at ca. 1417 cm−1 has been variously attributed to CH2 scissoring [[9] , [11] , [46] ] and to symmetric -COO stretching [[16] , [23] , [45] ]. The peak at ca. 1320 cm−1 has been interpreted to correspond to C-H bending [[23] ] and -OH bending vibration [[9] , [46] ]. (It was pointed by a reviewer that according to the spectra in Fig. 3 the peak at 1460 cm−1 from the original xylan almost disappeared in the CMX spectrum, whereas the peak at 1320 cm−1 in CMX was practically absent in the original xylan. Thus, the CH and OH bending frequency probably moved from 1460 to 1320 cm−1 when xylan was converted to CMX.)

For QAX, the band at 1620 cm−1 is broader due to the presence of associated water. The new peak at 1475 cm−1 corresponds to the bending vibrations of CH3 and CH2 in the quaternary ammonium moiety [[16] , [23] ].

The molecular weight distributions of the polymers were determined using an aqueous SEC method similar to the ones published in the literature [[29] -[31] ]. The SEC curves are shown in Fig. 4, and the molecular weight data summarized in Table 3. The xylan sample seemed to have a somewhat bimodal molecular weight distribution. The higher molecular weight xylan component (Mn ~ 279,000 and Mw ~ 443,000 Da) accounted for 22% of the polymer, and the lower molecular weight component (Mn ~ 6500 and Mw ~ 15,300 Da) accounted for 78% of the polymer.

SEC: beechwood xylan (black), CM xylan DS 0.4 (green), and QAX, DS 1.9 (red). (Color figure online)

Xylans obtained from different sources and different processes may have different molecular weights. For xylan obtained from bleached birch Kraft pulp, a unimodal distribution with Mn 6500 Da and Mw 11,700 Da were obtained [[6] ]. Xylan extracted from Dendrocalamus membranaceus Munro gave a bimodal SEC curve, with overall Mn ~ 10,100 Da and Mw 47,200 Da [[16] ]. A commercial xylan from beechwood was reported to have Mn of 5800 Da and Mw of 9500 Da [[45] ]. Xylan obtained from steam defibrillated beechwood was shown to have Mw of about 10,000 Da [[47] ]. Hromadkova et al. [[31] ] extracted xylan from beechwood and found a bimodal distribution for xylan, with Mw at ca. 480,000 and 23,000 Da for the two molecular weight components. Our results are roughly compatible with the findings of Hromadkova et al. [[31] ].

When the xylan was derivatized to become CMX, the higher molecular weight component in xylan decreased, indicating the presence of some hydrolysis. The higher molecular weight component (Mn ~ 250,000, Mw ~ 328,000 Da) accounted for about 6% of the polymer, and the lower molecular weight (Mn ~ 13,300 Da, Mw ~ 23,000 Da) accounted for 94% of the polymer. When xylan was derivatized to form the QAX, the higher molecular weight component completely disappeared, probably because of hydrolysis and the higher DS achieved. A single molecular weight band was observed at Mn 20,000 Da and Mw 32,000 Da.

Earlier Peng et al. [[8] ] studied the rheology of CMX solutions and reported that the CMX exhibited a slightly shear-thinning behavior; they also reported that the CMX had a lower viscosity than the starting xylan. In our work, the viscosity values as a function of shear rate of 5% solutions of xylan, CMX and QAX are given in Fig. 5. Xylan by itself has a viscosity of about 0.01 Pa s; the curve shows some shear-thinning behavior. Both CMX and QAX have lower viscosities (around 0.003 Pa s) probably due to some degradation occurring during synthesis. The viscosity curves for CMX and QAX are flat, indicating almost Newtonian flow behavior.

Viscosity as a function of shear rate for 5% xylan (squares, purple dashed curve), 5% CMX DS 0.4 (diamonds, brown solid curve), 5% QAX DS 1.9 (circles, green dotted curve). (Color figure online)

The TGA data are shown in Fig. 6. All the samples showed loss of water at about 70-120 °C and degradation of xylan backbone and cleavage of substituents at > 200 °C; the overall shape of the curves were similar. Xylan by itself started to degrade at around 240-250 °C. QAX showed the degradation at about the same temperature, except that the weight loss happened at a narrower temperature range. The QAX sample also had less ash at higher temperatures because it had been dialyzed to remove inorganic matter. CMX started to exhibit weight loss at a slightly lower temperature, ca. 200 °C. These results are generally consistent with what was reported earlier [[6] , [16] , [45] ].

TGA curves for xylan (black solid line), CMX DS 0.4 (blue dashed line), QAX DS 1.9 (green dot-dashed line), and CMX/QAX blend (2:1 weight ratio, brown dotted line). (Color figure online)

The DSC data for xylan, CMX, QAX, and CMX/QAX blend are shown in Fig. 7. The samples were all dried in vacuum at ca. 100 °C prior to analysis and the water loss peak at ca. 100 °C was therefore not apparent except for the QAX sample. The exothermic peak at ca. 200-270 °C was due to the chain degradation of the xylan backbone, consistent with data in the literature for this type of xylan [[43] ]. For xylan the peak maximum was found near 240 °C. This maximum peak occurred at a slightly lower temperature (216 °C) for CMX, but at a higher temperature for QAX (270 °C).

DSC in nitrogen for (a) xylan, (b) CMX DS 0.4, (c) QAX DS 1.9, and (d) blend of CMX/QAX (weight ratio 2:1)

Solutions of CMX and QAX were made up with 1-5% concentrations in water. When the two solutions with the same concentration were mixed, a cloudy mixture was obtained from 0.1 to 0.9 weight fractions of these two components, indicating the formation of a polyelectrolyte complex. A blend of CMX/QAX (2:1 weight ratio) was prepared for FTIR, TGA and DSC studies by mixing the appropriate solutions, where the DS values of the CMX and the QAX were 0.4 and 1.9, respectively. The molecular weights (on the anhydroxylose basis) for CMX and QAX were about 159 and 420, respectively. Thus, the molar ratio of anionic charges in CMX and the cationic charges in QAX was roughly 1:1.

The FTIR spectrum of the CMX/QAX blend is given in Fig. 3d. As expected, it showed the presence of the peaks for both CMX (1600, 1417 and 1320 cm_SP_−1_sp_) and QAX (1620 and 1475 cm−1). The TGA data for the CMX/QAX weight blend are shown in Fig. 6. The weight loss started at about 230 °C, about half-way between the starting weight loss points for CMX and QAX. The DSC curve for the CMX/QAX blend (Fig. 7) showed a maximum DSC peak at ca. 245 °C, between the peak maxima for CMX and QAX. The fact that only one DSC peak was observed in the blend, instead of two peaks, suggests that the two polymers interact with each other.

Table 4 gives the results of paper testing. The paper control gave a tensile strength of 0.065 MPa. When either CMX or QAX was added at 5% concentration, the tensile strength increased to about 0.083 MPa. When both CMX and QAX were added, the tensile strength notably increased to about 0.115 MPa. The addition of CMX and QAX was carried out in two procedures. First, CMX solution and QAX solution were consecutively coated on paper and then air dried. Secondly, CMX solution was applied onto the paper strips and dried, and then the QAX solution was coated and dried. In both cases, similar tensile strength was obtained.

For comparison, the paper was also coated with two types of starch. It is known that starch can increase the dry strength of paper [[48] , [49] ]. Indeed, the use of Cato® 255 (an amphoteric starch) and Opti-Plus® (a cationic starch) both increased the dry strength to about 0.08 MPa, roughly comparable to the use of CMX and QAX (Table 4). However, as noted above, the concurrent use of both CMX and QAX produced a more pronounced strength improvement.

In this work, anionic and cationic derivatives of xylan were prepared and characterized by NMR, IR, SEC, TGA, DSC, and rheology. 13C NMR spectral assignments that were more extensive than those previously published were achieved. The SEC data indicated that the xylan sample used was bimodal in molecular weight distribution but the higher molecular weight component decreased in amount during the derivatization process. When cationic and anionic xylans were combined, they formed a polyelectrolyte complex. This blend of cationic xylan and anionic xylan may have useful properties in product applications. For example, the blend of these two xylan derivatives was found to enhance the dry strength of paper and may be of interest in that connection.

Thanks are due to Qi Zhao at Tulane University for NMR spectra, Jade Smith at USDA for help with thermal analysis, Chanel Fortier at USDA for help with the IR data, Mark Washall at Ashland Inc. for assistance with SEC data, and Wayne Britt of Ingredion Inc. for the starch samples. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. USDA is an equal opportunity provider and employer.

1 Sixta H, Handbook of pulp. 1, 2006, Weinheim, Wiley-VCH Verlag GmbH & Co. KGaA, 28, 30

• 2 Hansen NML, Plackett D, Biomacromolecules, 2008, 9, 1493, 10.1021/bm800053z
• 3 Naidu DS, Hlangothi SP, John MJ, Carbohydr Polym, 2018, 179, 28, 10.1016/j.carbpol.2017.09.064
• 4 Cheng HN, Gu QM, Polymers, 2012, 4, 1311, 10.3390/polym4021311
• 5 Petzold K, Schwikal K, Heinze T, Carbohydr Polym, 2006, 64, 292, 10.1016/j.carbpol.2005.11.037
• 6 Alekhina M, Mikkonen KS, Alén R, Tenkanen M, Sixta H, Carbohydr Polym, 2014, 100, 89, 10.1016/j.carbpol.2013.03.048
• 7 Saghir S, Iqbal MS, Hussain MA, Koschella A, Heinze T, Carbohydr Polym, 2008, 74, 309, 10.1016/j.carbpol.2008.02.019
• 8 Peng X, Ren J, Zhong L, Cao X, Sun RC, J Agric Food Chem, 2011, 59, 570, 10.1021/jf1036239
• 9 Ren JL, Sun RC, Peng F, Polym Degrad Stabil, 2008, 93, 786, 10.1016/j.polymdegradstab.2008.01.011
• 10 Šimkovic I, Kelnar I, Uhliariková I, Mendichi R, Mandalika A, Elder T, Carbohydr Polym, 2014, 110, 464, 10.1016/j.carbpol.2014.04.055
• 11 Konduri MKR, Fatehi P, Carbohydr Polym, 2016, 146, 26, 10.1016/j.carbpol.2016.03.036
• 12 Hettrich K, Dreshsler U, Loth F, Volkert B, Polymers, 2017, 9, 129, 10.3390/polym9040129
• 13 Schwikal K, Heinze T, Ebringerova A, Petzold K, Macromol Symp, 2006, 232, 49, 10.1002/masy.200551406
• 14 Ren JL, Peng X, Peng F, Sun RC, Adv Mater Res, 2011, 239-242, 463, 10.4028/www.scientific.net/AMR.239-242.463
• 15 Kong W, Ren JL, Wang S, Li M, Sun RC, Fibers Polym, 2014, 15, 943, 10.1007/s12221-014-0943-z
• 16 Peng X, Ren JL, Zhong L, Sun RC, J Agric Food Chem, 2012, 60, 1695, 10.1021/jf204522k
• 17 Schwikal K, Heinze T, Saake B, Puls J, Kaya A, Esker AR, Cellulose, 2011, 18, 727, 10.1007/s10570-011-9526-y
• 18 Vega B, Petzold-Weicke K, Fardim P, Heinze T, Carbohydr Polym, 2012, 89, 768, 10.1016/j.carbpol.2012.04.006
• 19 Xu Y, Zhan C, Fan L, Wang L, Zheng H, Int J Pharm, 2007, 336, 329, 10.1016/j.ijpharm.2006.12.019
• 20 Reddy AS, Kalyani S, Kumar NS, Boddu VM, Krishnaiah A, Polym Bull, 2008, 61, 779, 10.1007/s00289-008-1003-x
• 21 Rutiaga MO, Galan LJ, Morales LH, Gordon SH, Imam SH, Orts WJ, Glenn GM, Ninol KA, J Polym Environ, 2005, 13, 185, 10.1007/s10924-005-2949-5
• 22 Pettersson G, Wågberg L, Höglund H, Nordic Pulp Paper Res J, 2006, 21, 115, 10.3183/NPPRJ-2006-21-01-p115-121
• 23 Qi X, Liu S, Chu F, Pang S, Liang Y, Guan Y, Peng F, Sun RC, Materials, 2016, 9, 4, 10.3390/ma9010004
• 24 Lambert BP, Lowes J (1975) US Patent 3,873,336A, 25 Mar 1975
• 25 Samu R, Moulee A, Kumar VG, J Colloid Interface Sci, 1999, 220, 260, 10.1006/jcis.1999.6506
• 26 Lyness WI, Gloss RA, Bates NA (1976) US Patent 3,998,690A, 21 Dec 1976
• 27 Li MC, Mei C, Xu X, Lee S, Wu Q, Polymer, 2016, 107, 200, 10.1016/j.polymer.2016.11.022
• 28 Dasgupta SP (1991) US Patent Application 5,338,407A, 23 Dec 1991
• 29 Eremeeva TE, Bykova TO, Carbohydr Polym, 1998, 36, 319, 10.1016/S0144-8617(97)00259-2
• 30 Rinaudo M, Danhelka J, Milas M, Carbohydr Polym, 1993, 21, 1, 10.1016/0144-8617(93)90109-H
• 31 Hromadkova Z, Ebringerova A, Malovikova A, Macromol Symp, 2006, 232, 19, 10.1002/masy.200551403
• 32 ASTM (2002) Method D 828-97. Standard Test Method for Tensile Properties of Paper and Paperboard Using Constant-Rate-of-Elongation Apparatus
• 33 Park HJ, Kim SH, Lim ST, Shin DH, Choi SY, Hwang KT, J Am Oil Chem Soc, 2000, 77, 269, 10.1007/s11746-000-0044-2
• 34 Flory AR, Requesens DV, Devaiah SP, Teoh KT, Mansfield SD, Hood EE (2013) BMC Biotechnol 13:28., http://www.biomedcentral.com/1472-6750/13/28
• 35 Cheng HN, Villalpando A, Easson MW, Dowd MK, Int J Polym Anal Charac, 2017, 22, 699, 10.1080/1023666X.2017.1367986
• 36 Vignon MR, Gey C, Carbohyd Res, 1998, 307, 107, 10.1016/S0008-6215(98)00002-0
• 37 Fundador NGV, Enomoto-Rogers Y, Iwata T, ACS Symp Ser, 2013, 1144, 393, 10.1021/bk-2013-1144.ch026
• 38 Cheng HN, Bennett MA, Anal Chim Acta, 1991, 242, 43, 10.1016/0003-2670(91)87046-A
• 39 Cheng HN, TrAC Trends Anal Chem, 1994, 13, 95, 10.1016/0165-9936(94)87073-X
• 40 Ren JL, Sun SC, Liu CF, Chao ZY, Luo W, Polym Degrad Stab, 2006, 91, 2679, 10.1016/j.polymdegradstab.2006.05.008
• 41 Kakurakova M, Wellner N, Ebringerova A, Hromadkova Z, Wilson RH, Belton PS, Food Hydrocoll, 1999, 13, 35, 10.1016/S0268-005X(98)00067-8
• 42 Zhang X, Liu C, Zhang A, Sun RC, Polymers, 2017, 9, 345, 10.3390/polym9080345
• 43 Cheng HN, Furtado RF, Alves CR, Bastos MSR, Kim S, Biswas A, Int J Polym Anal Charact, 2017, 22, 35, 10.1080/1023666X.2016.1222491
• 44 Sun RC, Tomkinson J, Ma PL, Liang SF, Carbohydr Polym, 2000, 42, 111, 10.1016/S0144-8617(99)00136-8
• 45 Sousa S, Ramos A, Evtuguin DV, Gamelas JAF, Ind Crops Prod, 2016, 94, 682, 10.1016/j.indcrop.2016.09.031
• 46 Pushpamalar V, Langford SJ, Ahmad M, Lim YY, Carbohydr Polym, 2006, 64, 312, 10.1016/j.carbpol.2005.12.003
• 47 Ebringerova A, Heinze T, Macromol Rapid Commun, 2000, 21, 542, 10.1002/1521-3927(20000601)21:9<542::AID-MARC542>3.0.CO;2-7
• 48 Hubbe MA, BioResources, 2006, 1, 281
• 49 Lim WJ, Liang YT, Seib PA, Cereal Chem, 1992, 69, 237

PHOTO (COLOR)

PHOTO (COLOR)

PHOTO (COLOR)

PHOTO (COLOR)

PHOTO (COLOR)

PHOTO (COLOR)

PHOTO (COLOR)

PHOTO (COLOR)

PHOTO (COLOR)