Cellulose nanofilaments and method to produce same

Abstract

Cellulose nanofilaments from cellulose fibers, a method and a device to produce them are disclosed. The nanofilaments are fine filaments with widths in the sub-micron range and lengths up to a couple of millimeters. These nanofilaments are made from natural fibers from wood and other plants. The surface of the nanofilaments can be modified to carry anionic, cationic, polar, hydrophobic or other functional groups. Addition of these nanofilaments to papermaking furnishes substantially improves the wet-web strength and dry sheet strength much better than existing natural and synthetic polymers. The cellulose nanofilaments produced by the present invention are excellent additives for reinforcement of paper and paperboard products and composite materials, and can be used to produce superabsorbent materials.

Claims

1. Cellulosic nanofilaments, wherein the nanofilaments are natural wood or plant fibers, the nanofilaments comprising: a ratio of DP.sub.nanofilaments/DP.sub.initial is at least 0.75, wherein DP.sub.nanofilaments is a degree of polymerization of the cellulosic nanofilaments and DP.sub.initialis an initial degree of polymerization of a source natural wood or plant fibers, the nanofilaments further comprising a length of at least 100 m, and a width of about 30 to about 300 nm, wherein the nanofilaments are physically detached from each other, and are substantially free of fibrillated cellulose, wherein the nanofilaments have an apparent freeness value of over 700 ml according to Paptac Standard Testing Method C1, and, wherein a suspension comprising 1% w/w nanofilaments in water at 25 C. under a shear rate of 100s.sup.1 has a viscosity greater than 100 cps.

2. The nanofilaments according to claim 1, wherein an aqueous suspension of over 0.1% w/w fails to settle according to a settling test described in GB 2 296 726.

3. The nanofilaments according to claim 1, wherein an aqueous suspension of less than 0.05% w/w settles to 50% volume according to the settling test described in GB 2 296 726.

4. The nanofilaments according to claim 1, wherein the length is between 100 m and 500 m.

5. The nanofilaments according to claim 1, comprising a surface charge of at least 60meq/kg.

6. The nanofilaments according to claim 1, wherein the ratio of DP.sub.nanofilament/DP.sub.initial is at least 0.80.

7. A mineral paper comprising at least 50% by weight of mineral filler and at least 1%, and up to 50% cellulose nanofilaments according to claim 1.

8. The paper according to claim 7, having mineral content up to 90%.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1A is a micrograph of a softwood kraft fiber cellulose raw material according to one embodiment of the present invention, viewed through an optical microscope;

(2) FIG. 1B is a micrograph of the cellulose nanofilaments produced from the raw material of FIG. 1A according to one embodiment of the present invention viewed through an optical microscope;

(3) FIG. 2 is a micrograph of cellulose nanofilaments produced according to one embodiment of the present invention viewed through a scanning electronic microscope;

(4) FIG. 3 is a schematic representation of a cellulose nanofilamentation device according to one embodiment of the present invention;

(5) FIG. 4 is a block diagram for production of the cellulose nanofilaments according to one embodiment of the present invention;

(6) FIG. 5 is a bar chart of the tensile energy absorption of never-dried wet web at 50% (by dry weight) solids content including varying amounts of the cellulose nanofilaments according to one embodiment of the present invention in comparison with a prior art system;

(7) FIG. 6 is a graph of tensile energy absorption (TEA in mJ/g) of never-dried wet web versus dosage of cellulose nanofilaments (dry weight %) according to one embodiment of the present invention;

(8) FIG. 7 is a graph of tensile energy absorption (TEA in mJ/g) of a dry sheet including cellulose nanofilaments according to one embodiment of the invention in comparison with a prior art system;

(9) FIG. 8 is a graphic plot of tensile energy absorption (TEA in mJ/g) of wet-web containing 30% PCC as a function of web solids versus cationic CNF (dry weight %) according to another embodiment of the present invention in comparison with a prior art;

(10) FIG. 9 illustrates a cross-section view of a nanofilamenting device according to one embodiment of the present invention; and

(11) FIG. 10 illustrates a sectional taken along a cross-section lines 10-10 of FIG. 9, illustrating one embodiment of a peeling agitator including blades according to one embodiment of the present invention.

DESCRIPTION OF THE INVENTION

(12) It is an objective of the present invention to provide a cellulosic material made from natural fibers, that is superior to all the cellulosic materials disclosed in the above mentioned prior art in terms of aspect ratio and the ability to increase the strength of paper, tissue, paperboard and plastic composite products. It is a further objective of this invention to provide a strengthening agent made from natural fibers whose performance is superior to existing commercial strengthening polymeric agents including starches and synthetic polymers or resins. It is another objective to provide a strength agent made from natural fibers that not only improves dry strength, but also the strength of the moist web before sheet drying. An additional objective of the invention is to provide fibrous reinforcing materials for the composite manufacture. Yet another objective of the invention is to provide fibrous materials for superabsorbent products. Still another objective is to provide a method or a device and a process to produce the high-performance cellulosic material from natural fibers.

(13) Accordingly, we have discovered that cellulose nanofilaments produced from natural fibers using our method have performance superior to conventional strength polymers and are different from all the cellulosic materials disclosed in prior art. Our nanofilaments are neither cellulosic fibril bundles nor fibers branched with fibrils or separated short fibrils. The cellulose nanofilaments are individual fine threads unraveled or peeled from natural fibers and are much longer than nanofibres, micro fibrils, or nano-celluloses as disclosed in the prior art. These cellulose filaments have a length preferably from 100 to 500 micrometers; typically 300 micrometers; or greater than 500 micrometers, and up to a couple of millimeters, yet have a very narrow width, about 30-300 nanometers, thus possess an extremely high aspect ratio.

(14) Because of their high aspect ratio, the cellulose nanofilaments form a gel-like network in aqueous suspension at a very low consistency. The stability of the network can be determined by the settlement test described by Weibel and Paul (UK Patent Application GB 2296726). In the test, a well dispersed sample with a known consistency is left to settle by gravity in a graduated cylinder. A settled volume after a given time is determined by the level of the interface between settled cellulose network and supernatant liquid above. The settled volume is expressed as the percentage of the cellulose volume after settling to the total volume. The MFC disclosed by Weibel et al. has a settled volume greater than 50% (v/v) after 24 hours settlement at an initial consistency of 1% (w/w). By contrast, the CNF made according to this invention never settles at 1% consistency in aqueous suspension. CNF suspension practically never settles when its consistency is over 0.1% (w/w). The consistency resulting in a settled volume of 50% (v/v) after 24 hours is below 0.025% (w/w), one order of magnitude lower than that of MDC or MFC disclosed by Weibel et al. Therefore, the CNF of the present invention is significantly different from the MFC or MDC disclosed earlier.

(15) CNF also exhibits a very high shear viscosity. At a shear rate of 100 s.sup.1, the viscosity of CNF is over 100 centipoises when measured at a consistency of 1% (w/w), and 25 C. The CNF is established according to Paptac Standard Testing Method C1.

(16) Unlike the nanocelluloses made by chemical methods, the CNF of the present invention has a degree of polymerization of the nanofilaments (DP) very close to that of the source cellulose. For example, the DP.sub.nanofilaments of a CNF sample produced according to this invention was 1330, while the DP.sub.initial of the starting softwood kraft fibers was about 1710. The ratio of DP.sub.nanofilaments/DP.sub.initial approaches 1 and is at least 0.60; more preferably at least 0.75, and most preferably at least 0.80.

(17) Because of its narrow width of the CNF, and shorter length relative to the original fibers, the CNF in an aqueous suspension can pass through the screen without forming a mat to obstruct water flow during freeness test. This enables CNF to have a very high freeness value, close to the carrier liquid, i.e. water itself. For example, a CNF sample was determined to have a freeness of 790 ml CSF. Because a freeness tester is designed for normal-size papermaking fibers to determine their fibrillation, this high freeness value, or apparent freeness, does not reflect the drainage behavior of the CNF, but an indication of its small size. The fact the CNF has a high freeness value whereas the freeness of the nanofibers of Koslow is near zero is a clear indication that the two families of products are different.

(18) The surface of the nanofilaments could be rendered cationic or anionic, and may contain various function groups, or grafted macromolecules to have various degrees of hydrophilicity or hydrophobicity. These nanofilaments are extraordinarily efficient for improving both wet-web strength and dry paper strength, and functioning as reinforcement in composite materials. In addition, the nanofilaments improve significantly fines and filler retention during papermaking. FIGS. 1A and 1B show micrographs of starting raw material fibers and cellulose nanofilaments produced from these fibers according to the present invention, respectively. FIG. 2 is a micrograph of the nanofilaments at a higher magnification using a scanning electronic microscope. It should be understood that microfibrillated cellulose is defined as a cellulose having numerous strands of fine cellulose branching outward from one or a few points of a bundle in close proximity and the bundle has approximately the same width of the original fibers and typical fiber length in the range of 100 micrometers. Substantially free is defined herein an absence or very near absence of the microfibrillated cellulose.

(19) The expression the nanofilaments are physically detached from each other means that the nanofilaments are individual threads that are not associated or attached to a bundle, i.e. they are not fibrillated. The nanofilaments may however be in contact with each other as a result of their respective proximity. For a better understanding, the nanofilaments may be represented as a random dispersion of individual nanofilaments as shown in FIG. 2.

(20) We have also discovered that the nanofilaments according to the present invention may be used in the manufacture of mineral papers. The mineral paper according to an aspect of the invention comprises at least 50% by weight of mineral filler and at least 1% w/w, and up to 50% w/w cellulose nanofilaments as defined above. The term mineral paper means a paper that has as the main component, at least 50% by weight, a mineral filler, such as calcium carbonate, clay, and talc, or a mixture thereof. Preferably, the mineral paper has a mineral content up to 90% w/w with adequate physical strength. The mineral paper according to this invention is more environmentally friendly comparing to commercial mineral papers which contain about 20% by weight of petroleum-based synthetic binders. In the present application, a treated paper product comprises the cellulose nanofilaments produced herein while a non-treated paper product lacks these nanofilaments.

(21) In addition, we have discovered that the said cellulosic nanofilaments can be produced by exposing an aqueous cellulose fiber suspension or pulp to a rotating agitator, including blade or blades have a sharp knife edge or a plurality of sharp knives edges rotating at high speeds. The edge of the knife blade can be a straight, or a curved, or in a helical shape. The average linear speed of the blade should be at least 1000 m/min and less than 1500 m/min. The size and number of blades influence the production capacity of nanofilaments.

(22) The preferred agitator knife materials are metals and alloys, such as high carbon steel. The inventors have discovered by surprise that contraintuitively, a high-speed sharp knife used according to the present invention does not cut the fibers but instead generates long filaments with very narrow widths by apparently peeling the fibers one from the other along the length of the fiber. Accordingly, we have developed a device and a process for the manufacture of the nanofilaments. FIG. 3 is a schematic presentation of such a device which can be used to produce the cellulosic nanofilaments. The nanofilamenting device includes 1: sharp blades on a rotating shaft, 2: baffles (optional), 3: pulp inlet, 4: pulp outlet, 5: motor, and 6: container having a cylindrical, triangular, rectangular or prismatic shape in cross-section along the axis of the shaft.

(23) FIG. 4 is a process block diagram where in a preferred embodiment the process is conducted on a continuous basis at a commercial scale. The process may also be batch or semi-continuous. In one embodiment of the process, an aqueous suspension of cellulose fibers is first passed through a refiner (optional) and then enters into holding or a storage tank. If desired, the refined fibers in a holding tank can be treated or impregnated with chemicals, such as a base, an acid, an enzyme, an ionic liquid, or a substitute to enhance the production of the nanofilaments. The pulp is then pumped into a nanofilamentation device. In one embodiment of the present invention several of nanofilamentation devices can be connected in series. After nanofilamentation, the pulp is separated by a fractionation device. The fractionation device could be a set of screens or hydro cyclones, or a combination of both. The fractionation device will separate the acceptable nanofilaments from the remaining pulp consisting of large filaments and fibers. The large filaments may comprise unfilamented fibers or filament bundles. The term unfilamented fibers means intact fibers identical to the refined fibers. The term filament bundles means fibers that are not completely separated and are still bonded together by either chemical bonds or hydrogen bond and their width is much greater than nanofilaments. The large filaments and fibers are recycled back to the storage tank or directly to the inlet of nanofilamentation device for further processing. Depending on the specific usage, the produced nanofilaments can bypass the fractionation device and be used directly. The nanofilaments generated may be further processed to have modified surfaces to carry certain function groups or grafted molecules. The surface chemical modification is carried out either by surface adsorption of functional chemicals, or by chemical bonding of functional chemicals, or by surface hydrophobation. The chemical substitution could be introduced by the existing methods known to those skilled in the art, or by proprietary methods such as those disclosed by Antal et al. in U.S. Pat. Nos. 6,455,661 and 7,431,799.

(24) While it is not the intention to be bound by any particular theory regarding the present invention, it is believed that the superior performance of the nanofilaments is due to their relatively long length and their very fine width. The fine width enables a high flexibility and a greater bonding area per unit mass of the nanofilaments, while with their long length, allows one nanofilament to bridge and intertwine with many fibers and other components together. In the nanofilamentation device, there is much more space between agitator and a solid surface thus there can be greater fiber movement than in the homogenizers, disk refiners, or grinders used in the prior art. When a sharp blade strikes a fiber in the nanofilamentation device, it does not cut through the fiber because of the additional space, and lack of solid support to retain the fiber such as bars in a grinder or the small orifice in a homogenizer. The fiber is pushed away from the blade, but the high speed of the knife allows nanofilaments to be peeled off along the length of fiber and that without substantially reducing the original length. This in part explains the long length of the cellulose nanofilament obtained.

EXAMPLES

(25) The following examples are presented to describe the present invention and to carry out the method for producing the said nanofilaments. These examples should be taken as illustrative and are not meant to limit the scope of the invention.

Example 1

(26) Cellulose nanofilaments (CNF) were made from a mixture of bleached softwood kraft pulp and bleached hardwood kraft pulp according to the present invention. The proportion of softwood to hardwood in the blend was 25:75.

(27) The mixture was refined to a freeness of 230 ml CSF prior to the nanofilamentation procedure, liberate some fibrils on the surface of the feed cellulose. Eighty g/m.sup.2 handsheets were made from a typical fine paper furnish with and without calcium carbonate filler (PCC), and with varying amounts of the nanofilaments. FIG. 5 shows the tensile energy absorption (TEA) of these never-dried wet sheets at 50% solids content. When 30% (w/w) PCC was incorporated into the sheets, the TEA index was reduced from 96 mJ/g (no filler) to 33 mJ/g. An addition of 8% CNF increased the TEA to a level similar to that of unfilled sheets. With higher levels of CNF addition, the wet-web strength was further improved, by 100% over the non-PCC standard. At a dosage level of 28%, the wet-web tensile strength was 9 times higher than the control sample with a 30% w/w PCC. This superior performance has never been claimed before with any commercial additives, or with any other cellulosic materials.

Example 2

(28) Cellulose nanofilaments were prepared following the same method as in Example 1, except that unrefined bleached hardwood kraft pulp or unrefined bleached softwood kraft pulp were used instead of their mixture. A fine paper furnish was used to make handsheets with 30% w/w PCC. To demonstrate the effect of the two nanofilaments, they were added into the furnish at a dosage of 10% before sheet preparation. As shown in Table 1, 10% CNF from hardwood improved the wet-web TEA by 4 times. This is a very impressive performance. Nevertheless, the CNF from softwood had even a higher performance. The TEA of the web containing CNF from softwood was nearly seven times higher than that of the control sample. The lower performance of the CNF from hardwood compared to CNF from softwood is probably caused by it having shorter fibers. Hardwood usually has a significant amount of parenchyma cells and other short fibers or fines. CNF generated from short fibers may be shorter too, which reduced their performance. Thus, long fibers are a preferable starting material for CNF production, which is opposite to the MFC that prefers short fibers as disclosed by Suzuki et al (U.S. Pat. No. 7,381,294).

(29) TABLE-US-00001 TABLE 1 Wet-web strength of the sheets containing 30% PCC and nanofilaments Nanofilaments addition (w/w %) TEA index at 50% solids Control 0 33 CNF made from hardwood 10 139 kraft CNF made from softwood 10 217 kraft

Example 3

(30) Cellulose nanofilaments were produced from 100% bleached softwood kraft pulp. The nanofilaments were further processed to enable the surface adsorption of a cationic chitosan. The total adsorption of chitosan was close to 10% w/w based on CNF mass. The surface of CNF treated in this way carried cationic charges and primary amino groups and had surface charge of at least 60 meq/kg. The surface-modified CNF was then mixed into a fine paper furnish at varying amounts. Handsheets containing 50% PCC on a dry weight basis were prepared with the furnish mixture. FIG. 6 shows the TEA index of the wet-web at 50% w/w solids as a function of CNF dosage. Once again, the CNF exhibits extraordinary performance in wet-web strength enhancement. There is an increase in TEA of over 60% at a dosage as low as 1%. The TEA increased linearly with CNF dosage. At an addition level of 10%, the TEA was 13 times higher than the control.

Example 4

(31) Cationic CNF was produced by following the same method as in Example 3. The CNF was then mixed into a fine paper furnish at varying amounts. Handsheets containing 50% w/w PCC were prepared with the furnish mixture following PAPTAC standard method C4. For comparison, a commercial cationic starch was used instead of CNF. The dry tensile strength of these handsheets is shown in FIG. 7 as a function of additive dosage. Clearly, the CNF is much superior to the cationic starch. At a dosage level of 5% (w/w), the CNF improved dry tensile of the sheets by 6 times, more than double the performance yielded by the starch.

Example 5

(32) Cellulose nanofilaments were produced from a bleached softwood kraft pulp following the same procedure as in Example 2. Handsheets containing 0.8% nanofilaments and 30% PCC were prepared. For comparison, some strength agents including a wet-strength and a dry-strength resin, a cationic starch were used instead of the nanofilaments. Their wet-web strength at 50% w/w solids content is shown in Table 2. The nanofilaments improved TEA index by 70%. However, all other strength agents failed in strengthening the wet-web. Our further study showed that the cationic starch even reduced wet-web strength when PCC content in the web was below 20%.

(33) TABLE-US-00002 TABLE 2 Tensile strength of wet-webs containing nanofilaments and conventional strength agents Dosage Additive (%) TEA index (mJ/g) Control 0 33 CNF 0.8 57 Wet strength resin 0.8 31 Dry strength resin 0.8 32 Cationic Starch 2 33

Example 6

(34) Cellulose nanofilaments were produced from a bleached softwood kraft pulp following the same procedure as in Example 2, except that the softwood fibers were pre-cut to a length of less than 0.5 mm before nanofilamentation. The CNF was then added to a fine paper furnish to produce handsheets containing 10% w/w CNF and 30% w/w PCC. For comparison, nanofilaments were also produced from the uncut softwood kraft fibers. FIG. 8 shows their wet-web tensile strength as a function of web-solids. Clearly, the pre-cutting reduces significantly the performance of CNF made thereafter. On the contrary, pre-cutting is preferable for the production of MFC (U.S. Pat. No. 4,374,702). This illustrates that the nanofilaments produced according to the present invention are very different from the MFC disclosed previously.

(35) To further illustrate the difference between the cellulosic materials disclosed in prior art and the nanofilaments according to the present invention, handsheets were made with the same furnish as described above but with 10% of a commercial nanofibrillated cellulose (NFC). Their wet-web strength is also shown in FIG. 8. The performance of NFC is clearly much poorer than that of nanofilaments, even worse than the CNF from precut fibers according to the present invention.

Example 7

(36) Cellulose nanofilaments were produced from a bleached softwood kraft pulp following the same procedure as in Example 2. The nanofilaments have extraordinary bonding potential for mineral pigments. This high bonding capacity allows forming sheets with extremely high mineral filler content without addition of any bonding agents like polymer resins. Table 3 shows the tensile strength of handsheets containing 80 and 90% w/w precipitated calcium carbonate or clay bonded with CNF. The strength properties of a commercial copy paper are also listed for comparison. Clearly CNF strengthens well the high mineral content sheets. The CNF-reinforced sheets containing 80% w/w PCC had tensile energy absorption index over 300 mJ/g, only 30% less than that of the commercial paper. To the knowledge of the inventors, these sheets are first in the world containing up to 90% w/w mineral filler reinforced only with natural cellulosic materials.

(37) TABLE-US-00003 TABLE 3 Tensile strength of mineral sheets reinforced with nanofilaments Tensile Mineral Nano- Long Breaking energy Mineral content filaments fibre length absorption type (%) (%) (%) (km) (mJ/g) PCC 80 6 14 1.25 315 PCC 90 4 6 0.56 134 Clay 90 4 6 0.99 230 Commercial 17 0 83 3.65 436 copy paper

Example 8

(38) Cellulose nanocomposites with various matrices were produced by casting in the presence and absence of nanofilaments. As illustrated in Table 4, nanofilaments improved significantly tensile index and elastic modulus of the composite films made with styrene-butadiene copolymer latex and carboxymethyl cellulose.

(39) TABLE-US-00004 TABLE 4 Tensile strength of nanocomposite reinforced with nanofilaments CNF Tensile index Elastic modulus Matrix (%) (N .Math. m/g) (km) Styrene-butadiene 0 2.06 3.0 copolymer Styrene-butadiene 7.5 7.26 50 copolymer Carboxy methyl 0 49.7 521 cellulose Carboxy methyl 7.5 63.5 685 cellulose

Example 9

(40) Cellulose nanofilaments were produced from a bleached softwood kraft pulp following the same procedure as in Example 2. These nanofilaments were added into a PCC slurry, before mixed with a commercial fine paper furnish (80% bleached hardwood/20% bleached softwood kraft) w/w. A cationic starch was then added to the mixture. First-pass retention (FPR) and first-pass ash retention (FPAR) were determined with a dynamic drainage jar under the following conditions: 750 rpm, 0.5% consistency, 50 C. For comparison, retention test was also conducted with a commercial retention aid system: a microparticle system which consisted of 0.5 kg/t of cationic polyacrylamide, 0.3 kg/t of silica, and 0.3 kg/t of anionic micropolymer.

(41) As shown in Table 5, without retention aids and CNF, the FPAR was only 18%. The microparticle improved the FPAR to 53%. In comparison, using CNF increased the retention to 73% even in the absence of retention aids. Combination of CNF and the microparticle further improved retention to 89%. Clearly, CNF has very positive effect on filler and fins retention, which brings additional benefits for papermaking.

(42) TABLE-US-00005 TABLE 5 CNF improves first-pass retention and first-pass ash retention Retention aid FPR, FPAR, Furnish chemicals % % Pulp + 50% PCC + 14 kg No 54 18 starch Pulp + 50% PCC + 14 kg 0.5 kg CPAM + 0.3 kg 74 53 starch S/0.3 kg MP Pulp + (50% PCC + 5% CNF) + No 84 73 14 kg starch Pulp + (50% PCC + 5% CNF) + 0.5 kg CPAM + 0.3 kg 93 89 14 kg starch S/0.3 kg MP Note: 1. Dosages in kilogram are based on one metric ton of whole furnish; 2. CPAM: cationic polyacrylamide; S: silica; MP: micropolymer.

Example 10

(43) Cellulose nanofilaments were produced from a bleached softwood kraft pulp following the same procedure as in Example 2. The water retention value (WRV) of this CNF was determined to be 355 g of water per 100 g of CNF, while a conventional refined kraft pulp (75% hardwood/25% softwood) w/w had a WRV of only 125 g per 100 g of fibers. Thus CNF has very high water absorbency.

Example 11

(44) Cellulose nanofilaments were produced from various pulp sources following the same procedure as in Example 2. A settlement test was conducted according to Weibel and Paul's procedure described earlier. Table 6 shows the consistency of CNF aqueous suspension at which the settlement volume equals to 50% v/v after 24 hours. The value for a commercial MFC is also listed for comparison. It is observed that the CNFs made according to the present invention had much lower consistency than the MFC sample to reach the same settled volume. This low consistency reflects the high aspect ratio of the CNF.

(45) Table 6 also shows the shear viscosity of these samples determined at a consistency of 1% (units), 25 C. and a shear rate of 100 s.sup.1. The viscosity was measured with a stress-controlled rheometer (Haake RS100) having an open cup coaxial cylinder (Couette) geometry. Regardless of the source fibers, the CNFs of the present invention clearly had much higher viscosity than the MFC sample. This high viscosity s caused by the high aspect ratio of CNF.

(46) TABLE-US-00006 TABLE 6 Consistency resulting in 50% settled volume and viscosity of 1% w/w suspension of various CNF samples and a commercial MFC sample. Viscosity at a shear rate Consistency of 100 s.sup.1 of resulting in 50% 1% w/w settled volume suspension after 24 hrs with water Samples (%) (cP) CNF from NBSK.sup.1 market pulp 0.018 127 CNF from never-dried 0.016 144 unbleached softwood kraft pulp CNF from never-dried 0.016 135 bleached softwood kraft pulp CNF from bleached hardwood 0.022 129 kraft market pulp.sup.2 A commercial MFC 0.38 10.4 Note: .sup.1North Bleached Softwood Kraft; .sup.2The fines in the hardwood pulp had been removed before making CNF.

(47) FIG. 9 illustrates a nanofilamentation device or nanofilamenter 104 according to one embodiment of the present invention. The nanofilamenter 104 includes a vessel 106, with an inlet 102 and outlet (not illustrated but generally found a the top of the vessel 106). The vessel 106 defines a chamber 103 in which a shaft 150 is operatively connected to drive motor (not shown) typically through a coupling and a seal arrangement. The nanofilamenter 104 is designed to withstand the conditions for processing cellulosic pulp. In a preferred embodiment the vessel 106 is mounted on a horizontal base and oriented with the shaft 150 and axis of rotation of the shaft 150 in a vertical position. The inlet 102 for the raw material pulp is in a preferred embodiment found near the base of the vessel 106. The raw material cellulosic pulp is pumped upward towards the outlet (not illustrated). The residence time within the vessel 106 varies but is from 30 seconds to 15 minutes. The residence time depends on the pump flow rate into the nanofilamenter 104 and any recirculation rate required. In another preferred embodiment the vessel 106 can include an external cooling jacket (not illustrated) along the vessel full or partial length.

(48) The vessel 106 and the chamber 103 that it defines may be cylindrical however in a preferred embodiment the shape may have a square cross-section (see FIG. 10). Other cross-sectional shapes may also be used such as: a circular, a triangle, a hexagon and an octagon.

(49) The shaft 150 having a diameter 152 includes at least one peeling agitator 110 attached to the shaft 150. A plurality or multiple peeling agitators 110 are usually found along the shaft 150 where each agitator 110 is spaced apart from another, by a spacer typically having a constant length 160, that is in the order of half the diameter 128 of the agitator 110 or so. Clearly each blade 120, 130 has a radius 124 and 134 respectively. The shaft rotates at high speeds up to (about 20,000 rpm), with an average linear speed of at least 1000 m/min at the tip 128 of the lower blade 120.

(50) The peeling agitator 110 (as seen in FIG. 10) in a preferred embodiment includes at least four blades (120,130) extending from the center hub 115 that is mounted on or attached to the rotating shaft 150. In a preferred embodiment a set of two smaller blades 130 project upward along the axis of rotation, and another set of two blades 120 are oriented downward along the axis. The diameter of the top two blades 130 is in a preferred embodiment from 5 to 10 cm, and in a particularly preferred case is 7.62 cm (from the tip to the centre of the shaft). If viewed in cross-section (as illustrated in FIG. 10) the radius 132 of blades 130 varies from 2 to 4 cm in the horizontal plane. The lower blade set 120 may have a diameter varying from 6 to 12 cm, with 8.38 cm being preferred in a laboratory installation. The width of the blade 120 is generally not uniform, it will be wider at the centre and narrower at the tip 126, and roughly 0.75 to 1.5 cm at the central portion of the blade, with a preferred width at the center of the blade 120 of about 1 centimeter. Each set of two blades has a leading edge (122, 132) that has a sharp knife edge moving in the direction of the rotation of the shaft 105.

(51) Different orientations of the blades on the agitator are possible, where blades 120 are below the horizontal plate of the center hub and blades 130 are above the plate. Furthermore, blades 120 and 130 may have one blade above and the other below the plate.

(52) The nanofilamenter 104 includes a gap 140 spacing between the tip 126 of blade 120 and inner surface wall 107. This gap 140 is typically in the range of 0.9 and 1.3 cm to the nearest vessel wall where the gap is much greater than the final length of the nanofilament obtained. This dimension holds also for bottom and top agitator 110 respectively. The gap between blades 130 and the inner surface wall 107 is similar to or slightly larger than that between the blade 120 and the wall surface 107.