Linen Most Useful: Perspectives on Structure, Chemistry, and Enzymes for Retting Flax

The components of flax (Linum usitatissimum) stems are described and illustrated, with reference to the anatomy and chemical makeup and to applications in processing and products. Bast fiber, which is a major economic product of flax along with linseed and linseed oil, is described with particular reference to its application in textiles, composites, and specialty papers. A short history of retting methods, which is the separation of bast fiber from nonfiber components, is presented with emphasis on water retting, field retting (dew retting), and experimental methods. Past research on enzyme retting, particularly by the use of pectinases as a potential replacement for the current commercial practice of field retting, is reviewed. The importance and mechanism of Ca2+ chelators with pectinases in retting are described. Protocols are provided for retting of both fiber-type and linseed-type flax stems with different types of pectinases. Current and future applications are listed for use of a wide array of enzymes to improve processed fibers and blended yarns. Finally, potential lipid and aromatic coproducts derived from the dust and shive waste streams of fiber processing are indicated.


Introduction of Flax and Linen Fiber
e history of �ax �Linum usitatissimum L.) is long and important. e translation of its scienti�c name, �linen most useful" [1,2], aptly describes its versatility and importance to world economy. Linen, the long, strong �bers from �ax stems, is considered one of the earliest successes in textiles [3]. While evidence does not exist on how early people learned to separate �bers from the stems, �ax as a ma�or textile in ancient Egypt is well documented in depictions of its cultivation and processing [4]. Linen samples have been reported in the remains of Swiss lake dwellings dating back some 10,000 years [3]. Production and use expanded beyond the Mediterranean countries to central and northern Europe, making its way to Great Britain about 2,000 years ago from the Middle East by Phoenician traders [3]. Linen, as one of the primary �bers for Europe throughout the Middle Ages and the Renaissance period, was used extensively for clothing. Linen was important to Russia and its economy through various stages of its political history [5]. Flax became the greatest export item and the basis of economic life in Russia in the late 1800s and into the twentieth century. At one time, Russia produced about 80% of the world's �ber �ax crop and before 1936 was the greatest exporter of �ax.
Fiber �ax came to North America by European colonists. Production of �ax in Connecticut was reported as early as 1640 [6]. While �ax was grown in several regions of the United States, particular states developed well-organized commercial efforts, particularly Michigan and the Willamette Valley in Oregon [7]. e production of �ax �ber in Oregon in the early 1900s, led by efforts of the US Department of Agriculture and the Oregon Agricultural Experiment Station, was remarkably well documented and included production yields, processing and mills, and other advancements [5,8]. As in Europe, specially designed equipment to pull, turn, deseed, and scutch �ax was developed to increase agricultural e�ciency. Oregon's commercial enterprise for �ax �ber, however, ended in the 1950s due to introduction of synthetic �bers and loss of government subsidies.
e linen industry also declined in Europe due to the coming of synthetic �bers, such as nylon and polyester for apparel [3,4]. Cotton before then, however, had overtaken the high position of linen and industrial �ax �bers, which had existed for millennia, due to large amounts of inexpensive cotton and its improved mechanical processing. Generally, cotton has led as the natural �ber of choice since this time, with only short times of reversals such as blockades during the American Civil War (1861-1865) and disruptions caused by World War II (1939)(1940)(1941)(1942)(1943)(1944)(1945). Aer the war, the lower production levels returned.
In traditional linen-producing areas, promotional programs by linen industries in Northern Ireland and western European countries led to a strategic organization to promote linen in the 1960s [4]. In the 1980s, the FAO (Food and Agriculture Organization of the United Nations) sponsored workshops on �ax, and in 1993, the "FAO Flax Group" became the "European Cooperative Research Network on Flax, " with coordination through the Institute of Natural Fibres, Poznan, Poland [9]. is program continues to compile data on crop production, facilitates interaction of several working groups, and sponsors numerous workshops thus promoting global interests in �ax �ber [10]. Two publications per year still appear under its auspices.
Recently, however, production of textile �ax �ber has decreased in most of the reporting European countries and including the Baltic states and the Russian Federation [10]. In the European Union (EU) countries, production of �ax declined from 122,379 ha in 2005/2006 to 73,029 ha in 2010/2011. Declining EU subsidies since the 1990s have reduced production levels and regions. Production in China, which has varied over the last two decades, has occupied a prominent position for the last several years [11]. However, the quality of Chinese-produced linen is not high enough for textiles, and China imports considerable amounts of long �ber from France and Belgium. e Belgium market for the prices of long �ber, which varies based on quality, recently was reported to range from 130 to 210€/100 kg, with prices for unscutched short �ber of 30 to 50€/100 kg [10].
Aer �ax production ended in Oregon and effectively ended �ax �ber production in the US, efforts to renew a �ax �ber industry in the United States have occurred over the years. Most have not been successful. In 1998, Naturally Advanced Technologies (NAT) and the Alberta Research Council in Canada developed Crailar �ax. NAT and the Hanes clothing company promoted the development of this product for complementary use with cotton in textiles. While the method is proprietary, popular news stories [12] and promotional websites [13] indicated that Crailar �ax increased performance characteristics in textiles.
Despite the reduction in production and usage from previous times, linen imparts characteristics of comfort, drape, and a distinctive appearance that have maintained a share of the market, particularly the luxury market for textiles [4,10,11]. Blending of cottonized �ax, that is, short, re�ned �ax �bers, with cotton and other �bers offers a potential for nontraditional, �ax �bers to impart distinctive properties in textiles. For example, blending cotton with increasing levels of cottonized �ax �ber for rotor spinning improved air permeability and wicking rates for moisture and modi�ed the fabric structure [14]. e �ne linen used for painting canvasses also requires strong and clean �bers from linentype production systems, such as that from Belgium. Despite the reduced production of traditional linen and move to other sources, likely markets will exist for the long, strong, and clean linen �ber in speci�c, high-value applications along with cottonized �ax �bers for blends [10]. e ma�or source for �ax �ber in North America, however, is straw from the linseed industry. Linseed provides an industrial oil widely used in paints, varnishes, cosmetics, and linoleum [15]. In the past and even more recently, �ax seeds are being recognized as a health food, with the intact seeds providing a laxative effect and linseed products providing lignans and omega-3 fatty acids [16,17]. It is likely, then, that production of linseed will continue and provide a consistent source of stems for �ax �ber production.
In contrast to agronomic systems to optimize �ber production in �ax, linseed production seeks to optimize seed yield [18]. Short branching plants are grown until seeds are fully matured. Stems are thicker in these plants than in �bertype �ax plants. In North America, as well as Europe and Russia, linseed production is in colder climates, where �ber extraction from the stems is very difficult due to reduced microbial activities and poor retting (see later). Most of the North American �ax �ber is extracted from linseed stalks by hammer-milling to obtain all the �ber, regardless of length, to be used in specialty papers. e quality of linseed �bers processed in this manner is considered too low for apparel or other high-value uses under the current commercial production and cleaning operations. Even with this hammer milled product for paper and pulp, however, there is currently interest in producing a cleaner �ber product, that is, less shive or non�ber fractions, in order to reduce the amount of chemicals needed to remove lignin from the �ax in pulping.
Several research programs are in place to promote the use of �ax and other natural �bers from various sources, including linseed stalks, in new products and in particular biocomposites. Efforts are currently under way to improve processing methods and �nd more uses for value-added �bers in industrial, for example, biocomposite, applications. In fact, industrial applications with composites and nonwoven materials may provide the greatest potential for expanded use of �ax �bers in the future [19][20][21]. Canada, which is the world's leading producer of linseed [22], currently promotes expanding the �ax industry, both for �ber and seed, through genetics research programs and organizations such as the Saskatchewan Flax Development Commission, the Composites Innovation Centre, and Biolin Research, Inc. in Saskatchewan, Canada [19]. To this end, Canadian research into �ax has recently increased, particularly towards improving �ber quality of high-value applications such as biocomposites. North Dakota, the main producer of linseed in the US, follows the trend of the linseed industry in Canada.
About 20% of the straw from the linseed industry satis�es the specialty paper (mostly cigarette) industry in North America. e remainder of this straw by-product (greater than 1 million tons from western Canada) is now burned or chopped to spread on �elds [22]. With the closing of the Ecusta operation, only one large processor for �ax �ber from linseed straw, that is, Schweitzer Mouduit International, Winkler, Canada, still produces �ber and mostly for paper and pulp and lower-end uses. A very great opportunity exists with linseed straw to improve farm economy, replace synthetic �bers with natural �bers, and provide a valueadded product for myriad applications [19]. Consistency in supply and in �ber characteristics must be addressed when �ax �ber is sought for large-scale industrial usage. e degree of processing for �ber cleanliness will depend upon the end product desired, and for many products, the levels of cleanliness and processing costs are considerably less than for linen in textiles. With the North American textile industry based on short staple �bers (e.g., cotton), the possible use of new processing methods to provide textile-grade �bers from linseed straw should not be overlooked. e goal will be to �nd improved ways of processing the �bers for consistent quality and with properties for applications in high-end uses, even for apparel [18]. Globally, an urgent and increasing need exists for new sources and products for agricultural, manufacturing, and health industries. While linen from �ber-type cultivars is still a valuable commodity for textiles, the desire for natural �bers in biocomposites points to linseed stalks as a largely untapped source of industrial �bers. �roduction of a higher quality �ber for value-added applications from the linseed stalks, however, requires some changes in �ber production and processing. Biotechnological approaches can supply innovative, e�cient, and more directed methods for �ber production.

Structure and Chemistry of Flax Stems and
Relationship to Applications and Products e chemical composition and the anatomical location of constituents within the �ax stem de�ne processing, properties, and applications of �ax. e anatomy of the �ax stem and chemical composition in the cell types is well documented in many publications (e.g., [24,37,38]). Figure 1 shows the arrangement from outermost to innermost layers as follows: cuticle connected with a single layer epidermis, bast �bers in cortical region, and woody core tissues.

Cuticle and Epidermis.
e cuticle resides at the outermost part of the stem ( Figure 1). Lipids, including waxes and cutin, and aromatics comprise this layer and provide a protective barrier to water loss and to invading microbial pathogens into the internal stem tissues [24,25,39]. e cuticle can be readily observed with the histochemical stain oil red [40,41], which stains the wax in the cuticle a bright red color and thereby provides a distinctive, visible marker speci�c for the cuticle (Figure 2). Flax �bers do not stain with oil red, indicating no or a small amount of wax on �bers per se. By staining processed �bers with oil red, a quick, visible assessment of cuticle contamination is possible in "cleaned" �ax �ber [25,40]. e cuticle and the adjacent single layer of thin-walled, epidermal cells are closely connected, coming off as a unit ( Figure 1). e "outer layer" of cuticle/epidermis comprises 13%-24% (by weight) of the bast fraction of selected �ber and linseed cutivars (Table 1) [23]. is layer, and mostly the cuticle, provides a rich assortment of lipids, including waxes, cutin, and sterols, while aromatics are present in small amounts (Table 1). Near-infrared Fourier transform Raman microscopy indicated, by absorption at speci�c wavelengths in stem thin sections, the presence of wax and soluble aromatic pigments in the cuticle [42]. Ultraviolet absorption microspectroscopy further indicated an absorption near 280 nm indicative of aromatics in the cuticle but not in the epidermis [24]. e parenchyma cells, including those below the cuticle/epidermis layer, lack aromatics and are a pectin-rich layer generally amenable to degradation by microbes and enzymes (discussed later), allowing separation of the cuticle/epidermis fragment during proper retting.
e cuticle is generally impervious to microbial attack as observed in �eld-retting of �ax, although some disruption and penetration of the cuticle can occur by �eld-retting fungi [43,44]. With pectinolytic enzymes, the cuticle can be sloughed off and freed from the �ber [45]. When this cuticle/epidermis fragment is intact, however, as when �ax is underretted (Figure 2), it remains attached to several �ber bundles, resulting in large fragments and reducing the quality of �ber and yarn [25]. Total wax and cutin contents were signi�cantly higher ( ) in commercial grades of low versus high �uality �ber and yarn, likely because of the contamination by cuticle [25]. Cuticle, short �bers broken during processing, and woody core fragments comprise the dust particles that are emitted during �ber processing. e dust so generated is considerable, and dust bags or other suitable collection containers are used to maintain a clean environment. Potential lipid coproducts from the dust fraction, as well as from a purer cuticle fraction removed during experimental enzyme retting, have been evaluated [46,47]. A hot alcoholic extraction was effective in removing the wax components, which could then be separated by controlled cooling [47]. ese waxes and other lipids have potential as commercial products from this waste material of �ax �ber processing.

Inner
Core Cells. e central woody core tissues are the primary xylem and other structural cells, which provide support and water conduction for the plant (Figures 1 and 4). e core cells are about 65%-75% of stem material. e main sugars are glucose, representative of cellulose, and xylose, representative of hemicelluloses (Table 2) [24]. Other carbohydrate components are lower in amounts and represent hemicelluloses and pectins.
Almost all of the lignin in �ax stems is present in the core cells. e histochemical stain for lignin, acid phloroglucinol [48], stains the entire core cell walls a bright red color, indicating coniferyl-rich constituents throughout this tissue (Figures 4 and 5). Positive reactions with acid phloroglucinol for coniferyl lignin (monomethoxylated aromatic rings) and chlorine-sul�te for syringyl lignin (dimethoxylated aromatic rings) [48] indicated that both lignin types are present in core cells. Lignin values for core cells have been reported at 25%-30% [49]. Results, however, vary with the analytical method used, such as Klason lignin with 72% sulfuric acid or that derived from permanganate oxidation. Oen, Klason lignin gives higher values because components other than aromatics are in the residue. By alkaline extraction and sub-se�uent summation of similar structures, speci�c lignin types were determined. Using this method, aromatic levels totaled about 1.5%, with a higher level of guaiacyl to syringyl units (Table 3). e low S/G ratio of lignin components was also shown by others [49]. Of the low molecular weight aromatics, only small amounts of ferulic acid were found (Table 3) [24,49]. Nuclear magnetic resonance (NMR) spectrometry con�rmed that the aromatics were present as lignin [24]. Further, ultraviolet (uv) absorption microspectrophotometry of thin sections also showed a strong absorbance near 280 nm in the core cell walls and no absorbance beyond, but absorbance maxima shied slightly in cells more towards the stem center [24]. e presence of a single strong absorption near 280 nm, as well as lack of absorbance near 320 nm, suggests a complex structure of the lignin similar to that in woody plants [50].
During �eld retting by indigenous fungi, the core cell walls remain intact and are virtually impervious to fungal attack, showing a typically high resistance to degradation for this type of lignocellulose [24]. ese lignin-rich core cells make up a large portion of the "shive" waste fraction generated during �ber processing. �ith such a high proportion of stem material, shive waste is a huge by-product of commercial �ber production. Unless there are uses for it, shive becomes an economic de�cit in processing, and its disposal is necessary. Shive, however, has economic value in the overall �ax processing system, generally now �nding application in low-value uses typical of lignocelluloses, such as animal bedding, mulch, particle boards, and thermal energy from burning. ese additional sources of revenue provided by selling the shive, 9.00€/100 kg [10], are essential to guarantee a positive economic position of the �ax processing facility. More recently, the lignocellulosic nature of shive has shown to be exceptional as activated carbon for heavy metal absorption [51] and for removal of chlorinated hydrocarbon [52], outperforming selected commercial products. Extraction and separation of cellulose, hemicellulose, and lignin from �ax shives by various methods, notably pressurized low polarity water extraction and under different conditions of pH and temperature, are reported as effective means of obtaining aromatic feedstocks from this by-product of �ber processing [49,53]. e fact that the core cells, that is, shive, contain by far the most aromatics in the plant stem while �ber contains very little gave rise to a rapid method to assess �ber cleanliness. Near infrared re�ectance (NIR) spectroscopy coupled with chemometric models for ratios of shive : �ber from 0% to 100% was used to develop a method to predict shive content in cleaned �ber [24,[54][55][56][57]. is method has been further developed and employed for at-line assessment of quality in which commercial bales were assessed over several weeks [58]. In converse, the cleanliness of shive from residual �ber le� from processing can be similarly predicted in materials sought for clean woody lignocellulosic materials [59].

Bast
Fibers. e industrially important bast �bers are long, slender, and strong specialized cells that develop in bundles in the cortex region, located between the cuticle/epidermis layer and the innermost woody cells ( Figure  1). ese cellulose-rich cells are the source of linen and other commercial �bers. Based on quality properties such as length, strength, and cleanliness, the bast �bers are sought for highvalue apparel and other textiles, natural �ber composite reinforcement, and specialty papers such as cigarette, currency, and Bible sheets [15,18,19].
In nature, these �bers exist in bundles of individual (or ultimate) �bers encircling the ligni�ed core tissues (Figure 1). Ten to 40 spindle-shaped ultimate �bers, each 2 to 3 cm long and 15 to 20 m in diameter, form in 20 to 50 discrete bundles [3,37]. Nodes or ��bernodes� [60], which are dislocations perpendicular to the axis, occur in �bers and entire bundles, as clearly shown by microscopy ( Figure 6). What creates �bernodes is not clear, and kink bands are similar in structure and apparently give similar responses. Possibly, pressure exerted during mechanical handling or even through growth and expansion of tissues may create their presence. ese �bernodes are important, because dyes, enzymes, and other liquids preferentially react there �rst [61][62][63]. Cellulases also attack preferentially at the �bernodes, even across bundles ( Figure 7). Breaks during stress and compression tests occur preferentially at �bernodes and kink bands [64,65].
Cellulose is the main component in bast �bers, with values of 65%-80% of the dry weight reported [63]. In addition to cellulose, bast �bers contain pectins, hemicellulose, and aromatic compounds in small amounts (Tables  4 and 5). Field-retted �ber showed an expected increase in glucose (by weight) indicative of cellulose, while increases also occurred in mannose and galactose. ese noncellulosic sugars appear to be inherently part of the �ber [66]. e fact that hemicelluloses, such as galactoglucomannans and xylans are substantial components in �ax �bers has been shown by considerable research [63,[66][67][68]. Distinguishing characteristics of linen, such as high moisture regain, may be in�uenced by the presence of these noncellulosic carbohydrates within the cellulosic structure. Proteins and proteoglycans are also associated with secondary walls of �ax �ber and possibly provide structure [69].
Data on lipids associated with bast �bers were collected from several studies and presented in Table 5. Fibers that had been manually separated and cleaned of all other visible materials showed the presence of low levels of waxes, cutins, and sterols, with amounts of about 0.2% of �ber dry weight and 1/20th or less of levels in the cuticularized epidermis. Other work, however, on a series of dew-retted and waterretted �ax �bers from �urope [27] showed that lipids were present on �bers, with higher levels on the better quality (i.e., stronger, and �ner) water-retted ones ( Table 5). With these latter samples, it was not clear if the lipids represented residual cuticle or if these �bers had more inherent lipids on the surfaces. Chemical analyses of highly cleaned �bers from linseed �ax and mature plants (i.e., grown for seed) had higher lipid levels than did bast �bers of the �ber-type plants ( Table 5). ese compounds in linseed may re�ect a higher level of waxes on the �bers or may re�ect more difficulty in removing all the cuticle from �bers during processing. e use of oil red as a histochemical stain, however, did not indicate wax on the clean �ber surfaces, and so likely it is residual cuticle (Figure 2).
�hile the question of lignin in bast �bers frequently arises, most data indicate that the amount of aromatics present is small [23,24,28]. Localization of aromatics, using histochemical stains [24,70] and ultraviolet (uv) absorption microspectrophotometry [24] showed that aromatic compounds were limited to middle lamellae and cell corners in bundles, and by far, the greatest levels were in cell corners (Figures 3 and 5). e deposition of aromatics, however, as shown by both methods, was sporadic in the bundles. Furthermore, staining with acid phloroglucinol [48] suggested a coniferyl-type lignin. Other work, however, using solid phase 13 C NMR (nuclear magnetic resonance) spectrometry indicated that aromatic material in �ax �bers was predominately an anthocyanin, rather than lignin [71]. Spectroscopic analysis of water-soaked, manually-separated, and then enzyme-retted �bers, which were free of all non�ber materials, indicated only trace aromatics in �bers from �berand seed-type �ax stems [72]. So, while trace amounts of aromatics are found in �ber bundles, the amount is small and does not appear to impede �ber/core separation in either �ber-type or linseed straw [28]. Oen pockets of acidphloroglucinol staining are observed on �bers, without the presence of obvious shive material, and likely represent a residual from either core or cuticle or the aromatic material from cell corners. Possibly, heavily localized areas of aromatics that remain on retted �ber could in�uence �ber quality properties [73] or reduce processing efficiency.
�xtraction of �ax bast tissue, which included �bers and cuticle/epidermis, with a series of organic solvents (i.e., hexane, propanol, methanol, and water) and analysis by reverse phase high pressure liquid chromatography (HPLC) F 4: Digital photograph of the stem fraction separated from �bers during processing and stained with acid phloroglucinol. e stems are bright red and well differentiated from �bers. F 5: Light microscopy of thin section of �ax stem stained with acid phloroglucinol indicating lignin throughout all the core cell walls. Fibers do not stain with acid phloroglucinol, except for a few cell corners. and 13 C NMR indicated a variety of aromatic constituents including �avonoids and hydroxy-methoxy cinnamic acids [74]. e water extract from these �ax samples contained a complex mixture of compounds, including sugars and aromatics. e phenolic-containing extracts inhibited cellulase and pectinase activities, suggesting a possible in�uence on retting enzymes if such compounds were released. Based on the previous discussion, the most likely source of these aromatic compounds in this study was the cuticle of the bast layer, rather than the �ber.
In the cellulose of plant cells, generally, a structure of unbranched linear glucose units allows zones of higher order, that is, crystalline regions, as well as areas of lower order, that is, noncrystalline regions [75]. In �ax �bers, �-ray diffractometry shows region of crystalline and less-ordered structure, with the linear orientation of both regions higher in �ax than the other cellulosic �bers cotton and ramie [63]. e secondary cell walls of �ax �bers at maturity are reportedly locked in an almost axial direction [66], giving lower elongation and a more brittle nature for �ax �bers   4 1.7 ± 0.6 10.9 ± 0.8 6.9 ± 1.7 1 Cutin represented by analysis for 8,16-dihydroxyhexadecanoic acid. 2 Fibers manually separated without any evidence of non�brous material. 3 See Table 4 for description. 4 Enzyme-retted 2 X. 5 Not determined. Data adapted from [23][24][25]28]. compared to cotton �bers. While �ax �ber is primarily a cellulosic �ber, its chemistry and characteristics provide speci�c properties that di�er from cotton and many other natural �bers.

Other Cell Types and Structures.
Parenchyma, cambium, and the middle lamellae, which bind ultimate �bers in the bundles, are particularly rich in pectins, hemicelluloses, and other matrix polysaccharides as shown by response of these tissues to pectinolytic enzymes [26]. Lignin is lacking in these structures for the most part. �e separation of �bers from the woody core occurs at the outer surface of the cambium and is facilitated when stems have been stored in dry climates for an extended time. Proper retting also releases the cuticle/epidermis layer from the �ber bundles. It is the pectin-rich regions that are of prime importance in retting, and considerable work has been done on the pectinases and ways to degrade the pectin in �ax.

Pectin.
Pectin is a complex polysaccharide of many plant cell walls and plant tissues [76]. Pectin, while oen low in amounts, is strategically located and binds cell walls within plants [66]. While pectin is, therefore, particularly important in maintaining the structure of �ax stems, its degradation is of fundamental importance for retting and the resulting quality of �ax �bers [37,73].

ISRN Biotechnology
Chemically, pectin is a heteropolysaccharide consisting mainly of 1,4-linked -D-galacturonic acid, with various degrees of methylesteri�cation at the carboxyl position and with various attached side chains [77]. In some cases, pectin in primary plant cell walls may have a high proportion of oligosaccharide chains on the backbone and longer chains than the pectin in the middle lamellae [77]. NMR spectrometry indicated that a rhamnogalacturonan structure of type I pectin, which is a prominent form in plants, likely forms the backbone of the high molecular weight polysaccharides in �ax �ber [78]. In retting of �ax, pectin degradation was reported to be faster in �ax harvested during �owering than in mature �ax stems, and a residual pectin level of 7 to 10 g/kg remained aer retting [79].
Chelators to remove Ca 2+ or other divalent cations are known to improve retting. Sharma patented a chemical retting process using ethylenediaminetetraacetic acid (EDTA) [80]. Nonmethoxylated carboxyl groups on galacturonic acids are oen cross-linked by Ca 2+ to form stable bridges across pectin molecules [77]. Mid infrared microspectroscopy mapping of mature �ax �ber indicated that pectin types varied among plant types and regions [81], with the potential to in�uence retting e�ciency. Immunocytochemical staining methods, using gold-labeled antibodies against speci�c pectin structures, provide further indications that the sites of speci�c pectin types vary within areas and even layers of �ax �bers [82,83]. Both nonmethoxylated pectin and calcium levels are higher in the epidermal regions of the �ax stem and lower in the �bers [84]. Inductive coupling plasma (ICP) emission spectrometry showed that calcium levels in the cuticle/epidermis tissue was 5.5-fold greater than in the bast �bers of "Ariane" �ax [65]. High amounts of calcium in the rigid cuticle/epidermis fragment further stabilize an already formidable barrier to retting in �ax stalks. Endopolygalacturonase, a pectinolytic enzyme that is present in many enzyme mixtures, was reportedly inhibited by steric hindrance through calcium linkages in pectin [85].
e reported levels of pectin in �ax vary considerably and are in�uenced by various factors [86][87][88]. For decorticated, that is, processed, �ax �ber and cell walls of various cultivars, the pectin content ranged from 20.5% to 34%. Chemical treatment with dilute hydrochloric acid followed by ammonium citrate resulted in a pectin content for �ax �bers of 1.6% [88]. While only an approximation, the sum of uronic acids, rhamnose, and arabinose was 1.7% of the total carbohydrates for �eld-retted Ariane �ax �ber [24].

Retting of Stems for Fiber Extraction
To obtain bast �bers for commercial use, �ax stems undergo a process called retting to separate �bers from non�ber materials, namely, cuticle/epidermis and the woody core. e method �rst used for separating and cleaning linen �bers is not known and was "possibly an accidental observation of the fact that �ax stems…turned to �bers under certain conditions of exposure to weather or immersion in water…" [4]. e historical value for this information has importance in some contexts, and searches are underway to discover the method  early linen producers used to produce their �ne strong �bers for textiles.
Retting is primarily a microbial process. e main idea is to degrade the pectins and other cementing compounds that bind the bast �bers and �ber bundles to other tissues and thereby separate �bers from non�ber materials [3,18,37]. e separated �bers are then cleaned of non�ber materials by mechanical processing [18]. For long �bers used for linen, specialized equipment is employed� the �rst stage is called scutching [89], which uses a specialized system to beat and stroke long �bers to remove shive, and the second is hackling [90], relying on a specialized instrument to comb, straighten, and align �bers. Insu�cient retting, or underretting, results in poor separation of the cuticle/epidermis layer and the woody inner tissues (i.e., shive) from the �bers [25,31]. Subsequent cleaning is then problematic, because the non�ber materials become entangled in the �bers and reduce �ber yield, processing e�ciency, and ultimate �ber quality. Conversely, overretting can occur. In this situation, the cellulosic �bers are weakened by overly active cellulases (Figure 7), resulting in poor �ber quality. Retting is of ultimate importance in �ber yield and quality. Even though currently most of the linseed straw �ber is of fairly low quality and used for paper and pulp, a desire exists to have less shive in the hammer milled �ber material in order to reduce chemicals used for deligni�cation in making pulp. e fact that linseed straw is produced in regions unsuitable for dew retting is problematic in having high �ber yields and clean �ber.
Two main methods have been historically employed commercially to ret �ax for textile-grade �bers, namely, water retting and dew (or �eld) retting [3,37]. In traditional water retting, as practiced in western Europe when the linen industry was �ourishing, �ax stems were pulled and submerged in bodies of water, for example, lakes, rivers, and ponds, for �ve to seven days. Aerwards, retted stems were dried and sun bleached in the �eld. Anaerobic bacteria, primarily pectin degraders, were the main organisms in water retting. Because the process was understood to some extent, technological methods were adapted to improve the process. Retting pits or tanks were constructed where temperature could be controlled. Selected microorganisms were chosen to improve water retting. At times, tanks were aerated to modify the bacterial consortium and thereby the bacterial metabolism, reducing the problems of pollution and stench from anaerobic metabolism that caused widespread concern in areas of western Europe [37].
Water retting resulted in long, strong, and �ne �bers of excellent quality for apparel and other textiles. e high cost of this method and resulting pollution, with the stench residing in water-retted �bers, however, resulted in water retting being abandoned for the most part in the mid 1950s in western Europe [37]. Water retting has been mostly replaced by dew retting, although some water retted �ax was commercially available in the early 2000s [91]. Reports indicate that China, the largest producer of �ax, may still produce water retted �ber in aerated tanks, but the �ber quality is not reported to be of good quality for textiles [11].
Dew (or �eld) retting is the method used in western Europe for obtaining high quality �bers for textiles. Field retting, however, is reported to be the oldest method of retting �ax, practiced thousands of years ago by the Egyptians [37]. Field retting is carried out by pulling �ax stems and laying them in even layers of rows for the moisture to encourage indigenous fungi to colonize and grow on the stem. e farmers of western Europe reportedly produced the best dewretted �ber because of the climate and their knowledge of when to turn and harvest the �ax for uniform retting. In areas of proper climate and expertise, commercially dew retting works and has been the method of choice for linen and other �ax �ber production. Most of the world�s textile �ax �ber is produced by �eld retting [10].
e quality of �ax �ber has declined over the years since dew retting replaced water retting as the main method for getting textile grade �bers in western Europe [92]. In addition to lower quality, �eld retting results in an inconsistent quality �ber. Dew retting continues to be the main retting method over water retting because production costs are lower, however, and �ber yields are higher and there is no stench. Dew retting, though, has a number of disadvantages other than poor and inconsistent quality compared to waterretted �ber. Certain areas formerly known for their linen production are unable to ret because of the noncompliant climates, such as England, Scandinavia, and Ireland. Field retting only works with appropriate moisture and temperature for fungal activity. Another disadvantage of �eld retting is that large tracts of land are tied up for weeks until �ax is suitably retted. In intensive agricultural areas (e.g., for multiple cropping), farmers are disadvantaged by having land occupied by retting �ax. Dew-retted �ber is dirty due to the fungi and soil. e vagaries of weather constantly threaten the harvest. Too dry weather results in poor fungal growth and lack of proper retting� wet weather delays �eld harvest and also interferes with fungal growth, resulting in pockets of anaerobic degradation. Over-retting occurs with excessive growth of cellulolytic fungi in the retting consortium and results in weakened �ber. So, even in the best regions for �eld retting, crop losses of about one-third are expectedly for one reason or another.
From the early 1900s, technological efforts have been attempted to improve retting [37]. is subject has been brie�y and recently reviewed by this author [18,38]. Methods for improved retting included modifying water retting or dew retting to remove the inherent problems and to select speci�c retting microorganisms. Stand retting, where standing plants are dried with a herbicide, notably glyphosate (Nphosphonomethyl glycine), and allowed to ret by indigenous microorganisms in a modi�ed form of �eld retting, has been tried [93,94] and is still being developed [95]. Oen the �ber properties were shown to be improved over dew retting, although dry weather interfered with retting and was problematic with this herbicide [96]. is method is still promising, particularly with new forms of the herbicide, and has been used to produce test plots of �ax for cottonized �ber in England [97].
In addition to modi�cation of the traditional water and �eld retting methods, much research has focused on chemical retting approaches, sometimes with and sometimes without microorganisms or enzymes [80,[98][99][100][101][102][103][104][105][106]. As mentioned previously [80], the use of chelators, notably EDTA, has been pursued, and its value is shown by strong sequestration of Ca 2+ at various pHs [98] to disrupt the pectin linkages. Autoclaving �ax straw with the chelators EDTA and oxalate has been used with breeding programs to effectively extract unretted �ax �bers [101]. A patent exists for a mechanical process to produce �ber strips followed by a chemical/cooking process under pressure [102]. Flash hydrolysis or steam explosion treatment, with or without impregnation before steam treatment, has been used to remove pectins and hemicelluloses from decorticated �ax to produce small bundles and ultimate �bers [104][105][106]. Ultrasonic treatment, following decortication and opening of green �ax or hemp stalks, has been used to obtain �bers from diverse sources without the use of chemicals [103]. e use of low energy, uniform ultrasonic treatment, combined with enzymes, has shown increased activity of various enzymes for cotton fabrics [107] and could be a useful method to improve the e�ciency in retting �ax. Chemical separation has resulted in successful laboratory results, but at times, �ber properties are less satisfactory than those from other methods. Efforts are reported to be still underway to assess physical and chemical methods to separate �ber. e use of enzymes, focusing on pectinases, also has been researched for some time [37,[108][109][110]. None of these methods has replaced �eld retting as a commercial practice. Enzyme retting, however, has proven to offer promise as a biotechnological improvement and is still undergoing research and development for improving �ber quality.
3.1. Enzyme Retting. Sharma and his colleagues in Europe and the United Kingdom in the 1980s carried out a major effort on enzyme retting, primarily attempting to mimic water retting with a consortium of plant cell wall-degrading enzymes. ese efforts, as well as other topics related to �ax production for textiles, have been documented in numerous papers and in the book e Biology and Processing of Flax [111]. Since the plant cell wall is a complex lignocellulosic material, it was believed that a mixture of cellulases, hemicellulases, and most notably pectinases was required to ret �ax as occurred with the plant-associated natural microbial consortia in water or dew retting [110]. In hindsight, part of this approach could have been due to the lack or cost of speci�c enzymes available at the time, as shown by attempts to use highly pectinolytic microorganisms [110]. Fungal cultures contained this mixture of cell wall-degrading enzymes but with some different pro�les and activities. Microorganisms with high levels of pectinases were chosen, but most culture �ltrates also contained some cellulases. Sharma and colleagues enjoyed success in their work. Several commercial enzyme mixtures containing plant cell wall-degrading enzymes were tested. A product from Novo Nordisk (Copenhagen, Denmark) called SP 249 was used in a side-by-side test of enzyme versus water retting [108]. SP 249 was used at 3 g/L in an 11 : 1 liquid-to-solid ratio at 45 ∘ C for 24 hours. Eighty kg of �ax stalks were submerged for each of the two test methods. Aer retting, �ber yield and quality were equal for the two retting procedures. Oxidizing agents, however, were required to denature the enzymes and stop the continuing action of cellulases in the enzyme mixture. e study showed the successful application of enzyme retting in pilot plant scale, promising that retting of �ax could occur with enzymes and with a reduced time of retting [108]. e liquid method also ensured a more consistent product and could be carried out in any location with proper equipment. From the work of Sharma and colleagues, Flaxzyme, which was a patented liquid preparation of balanced cellulases, pectinases, and hemicellulases from Aspergillus species, was developed by Novo Nordisk (Copenhagen, Denmark) for retting �ax. is enzyme also resulted in �ber yield and properties equal to or better than �ber from water retting [37]. Later, Lyvelin (Lyven, Caen, France), a pectinase, but not pure, from Aspergillus niger was marketed speci�cally for retting of �ax. Despite these positive results and developments, an enzyme retting method did not replace dew retting. While the reasons are complex, likely enzyme costs and lack of industry support prevented further development. Flaxzyme is no longer sold under this name.

Further Research on Enzyme
Retting. In the 1990s, the United States was the largest per capita user of �axcontaining textiles, but no linen or �ax �ber for textiles was produced domestically. e only �ax grown was for linseed. is statistic prompted the Agricultural Research Service (ARS) of the US Department of Agriculture to begin research toward developing a �ax �ber program for textiles. e �rst goal in this initiative was to improve retting emphasizing enzymes. Results from the work in Europe, particularly with pectinases, were the basis of research. ere were other considerations, however, that became apparent for this work. Canada and the US northern plains states had a thriving linseed industry, with tons of waste straw available. Linseed straw removal aer harvest presented a problem for farmers. Producers wished to remove the straw from their �elds soon aer harvest, and the straw did not degrade readily. A small portion of this crop, estimated around 25%, was used for hammer milling and pulp for specialty papers, such as cigarette, currency, and Bible sheets. Most of the linseed straw was not used and was (and still is) burned to remove it from the �elds. Further, the US textile industry was tied to cotton �ber processing and not the long-line linen processing of Europe. In fact, no specialized wet-spinning equipment required for long-line linen yarn existed in the US. e textile spinning technology was based on short staple �bers like cotton and synthetic blends. A �ax tow product, which is short �ber as a by-product of long-line linen, was used in blends with cotton, but that product, too, was imported.
So, quickly, the ARS research effort incorporated linseed stems as a source of �bers, with attempts to improve retting and processing for higher, quality in the �bers. Further, since the �bers desired were short-staple like cotton, "total �ber production, " with collection of all bast �ber regardless of length, was employed with later "cottonizing" to shorten and re�ne the �bers.
e ARS �ax research focused on lowering enzyme amount, �nding purer and more active enzymes, and developing a protocol for enzyme retting. Earlier research indicated the value of calcium chelators for disrupting pectins in retting (see ealier). Tests with oxalic acid showed that chelators could greatly reduce the amount of enzyme needed for retting [45]. So, with following tests, the retting mixture was almost always an enzyme/chelator mix.
Excellent �eld retting of �ax had been noted in research experiments at Clemson University, South Carolina, for winter production of �ax. From this material, the major fungi colonizing the stems were cultured and isolated in a search for more active retting enzymes. One fungus stood out from the others as the most active retter of �ax [43]. is fungus, identi�ed eventually as Rhizopus oryzae sb NRRL 29086, produced a potent endopolygalacturonase (EPG) and few other enzymes in the �ltrate [30,112]. is enzyme was puri-�ed and tested for its ability to separate �bers in stems and compared in mixtures with potentially complementary cell wall-degrading enzymes. In these early tests, �ber separation was judged by light microscopy and the Fried's Test, which is an in vitro test to judge �ber separation from stems by comparing visual images [37].
Results indicated that this puri�ed EP� with oxalic acid alone was sufficient to separate �ax �bers (Table 6). e other potentially complementary enzymes tested, namely, pectin methyl esterase, xylanase, and cellulases, did not improve �ber separation [30]. Data, therefore, indicated that enzymes other than EP� were not required to separate �ber from non�ber fractions.
Based on these results, a search for a potential commercial enzyme, with high pectinase and low cellulase activity, indicated that Viscozyme L (Novozymes North America) had similarities to SP 249 and Flaxzyme. Oxalic acid in later tests proved not to be suitable as a chelator, as a precipitate formed and remained on the �bers. �hile tests continued on a series of chelators of different types and under different conditions [113,114], EDTA replaced oxalic acid because of efficiency at pH 5 to 10 and its commercial availability for the textile market. e retting test method of choice, then, was a combination of the commercial products Viscozyme L and EDTA (later using Mayoquest, a 36%-38% EDTA commercial product) for other comparisons and modi�cations. A series of Viscozyme/EDTA formulations, with increases in each of the enzyme and chelator, was used to ret a mature, �ber-type �ax. Rather than the Fried�s test, cotton �ber tests were used for strength by Stelometer [115] and �neness [116] using a modi�ed microaire system equilibrated with a series of �neness standards �bers [117]. An estimate of the percentage of �ne �ber, collected by passing retted and mechanically cleaned �ber through the Shirley Analyzer, was calculated from the starting material. e Shirley Analyzer, which is an instrument to separate and collect trash in cotton, provided a percent of �ne, cleaned �ber as an additional statistic to judge quality.
Experiments with a series of retting combinations [32] indicated that increasing Viscozyme levels increased �ne �ber yield but reduced strength, regardless of chelator levels ( Table  7). Increasing levels of chelator, within each enzyme series, increased �ne �ber yield and resulted in �ner �bers up to 18 mM. Chelator level alone did not affect strength. Fibers from this study were blended with cotton (50 : 50), spun as yarn in a miniature spinning system, and the yarn properties determined [118]. Results indicated that enzyme treatments affected yarn properties, with the highest enzyme level producing �ner �bers, easier yarn construction, and better quality yarns. Assessments are difficult to compare, however, because, for example, the highest level of enzyme produced a �ner but weaker �ber. Possibly, �ner but weaker �ber was also less stiff and brittle and therefore more amenable for blending with cotton. Both of these characteristics are important in yarn construction. e clearest result from this work was that the miniature spinning system has value in predicting the optimal retting formulations for yarn quality.
In order to develop a new pilot plant method for enzyme retting, a spray enzyme method, or a brief (2 min) soaking, to deliver the enzyme/chelator mixture was used to reduce liquid : solid ratio compared to the former method of retting �ax in submerged tests [119]. Since the cuticle/epidermis layer protected the internal stem tissues [45], methods were explored to facilitate the entry of enzymes into stems. Physically crushing stems with �uted rollers to breach this barrier improved enzyme retting over that by increased or reduced atmospheric pressures [120]. Although inhibitory, aromatic compounds had been shown to be released by chemical treatments [74], presoaking of �ber with water to remove these compounds showed no clear bene�t with enzyme retting and was not included as part of the enzyme retting protocol [121].
e enzyme retting method, termed SER, was tested on seed-and �ber-type plants, with various levels of enzyme and chelator. Ultimately, enzyme retting would have to be integrated to a commercial cleaning system for �ber production, but such cleaning systems did not exist in the US. Accordingly, arrangements were made to produce pilot scale amounts of enzyme-retted �ax for commercial processing in Europe. About 12 kg of retted �ax for each formulation was commercially cleaned by Ceskomoravsky len (CML) in Humpolec, Czech Republic, using a Uni�ed Line scutching system and La Roche cottonizing system [33]. ese �bers were tested for quality parameters and then spun into blended yarns with cotton at 50 : 50 and at 10 : 90 �ax : cotton amounts. Fibers were tested with modi�ed cotton testing equipment (Table 8) and the yarn properties by commercial testing equipment and methods (Table 9).
Results indicated that the enzyme mixtures retted both �ber-and seed-type plants (Table 8). Higher levels of enzyme reduced �ber strength but produced �ner and higher amounts of �ne �ber. Other than strength, �neness and �ne �ber yield were better than these characteristics in dew-retted �bers. Chelator levels did not seem to vary in their impact, with 25 mM amounts equal to the 50 mM levels. e seed �ax �bers were of less quality than �ber-�ax �bers with similar formulations. Retesting of the �bers 30 months later showed no further loss in strength, indicating the washing step aer enzyme retting was sufficient to stop further enzyme activity. Yarn properties were compared favorably between dew retted and enzyme retted at the higher level of enzyme (Table 9). Seed-�ax �bers processed into blends much like the �ber-�ax �bers, even the dew-retted sample. Speci�c areas for improvement were identi�ed. Of concern was the loss of strength with increasing level of enzyme.
Since retting, processing, and yarn construction are interrelated, a �ax pilot plant was constructed based on the Uni�ed Line at CML in order to have a commercial method for cleaning retted �ax stems [122]. is system was developed by engineers at CML but reduced in size and with each of four parts separately positioned for research. e parts were: 9-roller calender for breaking stems, 5-roller calender for further crushing shive, scutching wheel to produce a total �ber from the stems, and an upper pinned shaker to remove loose shive and straighten �bers.
Further, a series of test protocols was developed for objective test results of �bers. Fibers were tested using cotton equipment and protocols where applicable, such as strength and elongation by Stelometer [115]. Further test standards were developed for �ax, including color, �neness, and predicted shive through the Flax and Linen subcommittee of ASTM International [56,57]. e following tests methods adopted were (1) percent of �ne �ber yield produced by passing through a Shirley Analyzer, (2) tensile strength  [31]. Data from [32]. and elongation by Stelometer, (3) �neness based �rst on a modi�ed cotton air�ow method and later re�ned for a new ASTM test method [57], and (4) the percent shive in cleaned �ber using near infrared spectroscopy and chemometric models from �ber : shive combinations [54,55,59]. is latter method was accepted as a new ASTM test method in 2005 [57]. For certain assessments, the Fried's Test and light and scanning electron microscopy were used when appropriate. A color test method (D-6961-03) was approved as a new ASTM test method in 2003 [57]. Enzyme retting results in a lighter �ber color than that by �eld retting, and various enzyme retting formulations resulted in different color characteristics based on the CIELAB L * , a * , and B * values [123,124]. ese results suggest methods to tailor color properties for applications.
Enzyme mixtures that included cellulases, such as Viscozyme L, weakened �bers (Figure 7), as shown by previous work (Table 8). Purer enzymes were becoming more readily available at this time, and research had showed that puri�ed EPG alone could separate �bers from core without the other cell wall-degrading enzymes. Further, results indicated these methods worked on both �ber-type and seed-type �ax cultivars, indicating that these enzymes should be applicable to linseed straw. Commercial enzyme products, developed for various applications, were tested with the intent of �nding purer pectinases (i.e., low or no cellulases) that effectively retted �ax without loss of �ber strength. As these further tests were being carried out, Novozymes North America, Inc. (Franklinton, NC, USA) released a commercial pectate lyase (PL) product for removing the cuticle of cotton �bers as an environmentally friendlier way of scouring cotton, which traditionally used high levels of NaOH [125,126]. BioPrep 3000L is a liquid commercial PL produced by multiplying the native gene for alkaline PL in Bacillus lichniformis, placing  the genes back into the bacterium and allowing expression of these genes for high levels of enzyme production. Bioprep has a reported activity of 3,000 alkaline pectinase standard units (APSU)/g. We used a product marketed under the trade name Dextrol Bioscour 3000 (Dexter Chemical LLC, Bronx, NY) [35], which was shown to separate the �bers (Figure 8). At this time, other enzymes were developed and applied especially for enzyme retting [127,128]. Texazym BFE and Texazym DLG are propriety names for enzymes from Inotex Ltd., Dv�r �r�lovi, Czech Republic, speci�cally mentioned for use in �eld retting. �ther commercially produced enzymes for various applications of degradation of plant materials were selected. Fibers produced by various enzyme retting formulations were assessed through use of the �ber processing pilot plant and objective test methods [34]. All the enzymes were tested using suppliers' recommendations for optimal activity. e Fried's Test suggested initial levels and times for effective �ax retting of these various enzymes and formulations [34]. Texazym BFE effectively separated �ber from core at 2%, 5%, and 10% levels aer 24 h; only 10% BFE retted �ax at 7 h. e addition of EDTA (18 mM concentration) improved retting, showing effective �ber separation at 7 h for 5% BFE. EDTA at 18 mM concentrations improved retting of all enzymes except DLG, which alone was ineffective in �ber separation by this method. e effect of retting was further evaluated using BFE, DLG, Multifect Pectinase FE, and Bioprep 3000 L in several modi�cations of formulas and retting conditions [34]. While 1% BFE was effective at 24 h, the addition of EDTA facilitated enzyme retting with all levels of this enzyme. e 2% level appeared to be effective enough to warrant further study, and temperatures in the 50 to 60 ∘ C range were more effective than lower temperatures. Incubation of stems with DLG at 5%, even with EDTA,did not result in �ber separation by this test. Multifect Pectinex FE was effective at 0.2% with EDTA, but not without the chelator; lower levels were less effective than other enzymes even with EDTA. Addition of DLG as high as 0.5% did not improve �ber separation efficiency of 1% BFE plus EDTA by the Fried's Test. Similarly, addition of xylanase, to treat animal feed and reported low in cellulase activity, included up to 0.15% with Multifect FE plus EDTA did not improve �ber separation. Bioprep 0.05% at pHs 8 and 9 and with chelators effectively separated �bers from the core.
Many of the enzyme mixtures tested contain multiple types of enzymes active against plant cell walls, including cellulases. A companion study was carried out to test the activities of several commercial polygalacturonases from various sources [129]. Tests of enzyme activities and �ax �ber properties, including strength, indicated different cellulase activities within these products that affected �ber properties. Microscopic analysis and incubation of commercial �ax �bers with these enzymes over several days shows clear signs of �ber degradation by many. Texazym BFE and Bioprep, however, resulted in slight to no �ber destruction. Bioprep is listed as an alkaline pectate lyase, while the optimal conditions for activity of Texazym, which is not identi�ed as to type of pectinase, is similar to that for Bioprep. In contrast, Texazym DLG and Sigma cellulase were very destructive to �ax �bers. It should be noted that most of these commercial enzymes are not marketed for �ax retting. eir use, for example, Texazym DLG, however, could modify �ber and yarn properties as will be discussed later.
Stems retted with various enzyme formulations by the SER method and with �ber processed through the pilot plant are shown in Table 10. Fine �ber yield was highest for Texazym BFE and Bioprep formulations but not signi�cantly different from Viscozyme plus Mayoquest. All enzyme-retted and Shirley-cleaned samples were cleaner than unretted �ber, and differences were not large among the enzyme treatments. Differences, however, occurred in strength and �neness among retting formulations.
Further evaluation of the use of these enzymes for retting included the following: the amount of formulation uptake during brief (i.e., 2 min) soaking, �ne �ber yield, and cleanliness [35]. Uptake of the amount of various formulations of Bioprep was similar and about 300 mL (ranging from 272 to 408 mL) for 150 g initial �ber, giving a liquid-to-�ber ratio of 2-2.7 to 1.
Additional assessments tests were made on linseed varieties grown under commercial-type conditions in North Dakota [35]. Results showed that all retting enzymes were more efficient with chelating agents, particularly EDTA. EDTA has substantial Ca 2+ binding activity even at pH 5 [114], providing a positive effect of EDTA at low pHs, which is optimal for some enzymes such as EPG and Viscozyme. e binding capacity of EDTA for Ca 2+ is, however, considerably greater at alkaline pH [114], and the use of EDTA at a higher pH should be more efficient in separating �ber from core. It is well known, however, that PL requires Ca 2+ for activity [77]. e suggested method for cotton scouring with Bioprep is to �rst apply the enzyme and later apply the chelator (S. Salmon, Novozymes, personal communication).
e enzyme retting methods developed earlier indicated that Viscozyme could ret linseed varieties of �ax but reduced �ber strength. Tests were conducted on the efficiency of T 10: Properties of mature Ariane �ax �ber enzyme retted with various commercial products.
Retting formulation 1 Fine �ber yield (%) 2 Strength (g/tex) 3 Fineness (SSI) 4 Predicted shive (%) 5  Bioprep and Viscozyme to ret two linseed varieties, and �ber properties were determined. Hermes and Omega were grown to full seed maturity under production conditions in North Dakota. Stems were enzyme retted using formulations with Bioprep or Viscozyme in side-by-side tests (Table 11). e Omega sample had rain prior to baling, and substantial weathering had occurred as indicated by darkening of the straw. Hermes, in contrast, was light and showed no effects of weathering prior to enzyme retting. Bioprep effectively retted both cultivars and resulted in higher �ber yield and �ber strength than Viscozyme. Hermes was �ner a�er retting with Viscozyme plus chelator (Table 11). In this test, the chelator was used subsequent to the soaking with Bioprep but with the Viscozyme in a single solution.
Tests for incubation times with Bioprep, level of Bioprep (without chelator), and levels and incubation times of chelator were further assessed [35]. Based on �ne �ber yield and percentage shive content, incubation with Bioprep for 1 h followed by incubation with 18 mM EDTA for 24 h was equal or better than other conditions. Retting effectiveness, however, improved with increased amounts of Bioprep up to 0.5%, which was the highest level tested in this experiment and suggested that further increases in enzyme level may improve �ber separation. Furthermore, scanning electron microscopy of retted �bers indicated that Bioprep levels of 5% appeared to remove more contaminants than 0.1%.
Based on earlier results and general recommendations for bioscouring cotton with Bioprep (personal communication, S. Salmon, Novozymes North America, Inc.), a series of evaluations was carried out to optimize the use of Bioprep and EDTA for retting �ax (Table 12). Linseed variety Hermes was selected for these tests. e recommendation for bioscouring cotton was to treat with Bioprep about 15 min prior to adding chelators (S. Salmon, personal communication). e use of Mayoquest 200 to supply EDTA as chelator at 18 mM concentration, which had been determined from use with Viscozyme, appeared to work adequately with Bioprep. To further optimize the formulation and method for enzyme retting, Hermes was retted with a range of Bioprep levels from 0.1% to 5% and followed by chelator or combined with chelator in the formulation. e higher levels of "�ber" with the lower enzyme levels arise from �ber plus shive in varying amounts, as shown by predicted shive amounts. For Shirley-cleaned �ne �ber, Bioprep at 1.0% to 5.0% followed by chelator produced the highest �ber yields and the lowest shive contents, ranging from 1.5% to 2.3%. Bioprep at 5% did not produce higher yields or cleaner �bers than 1% or 2% levels. Shirley-cleaned �bers do not represent all the �bers that could be extracted in commercial, cottonizing systems. erefore, �ber yields from a single pass through the Shirley Analyzer were used only to rank enzyme formulations.
Fiber strength was maintained at all levels of Bioprep (Table 12), showing a signi�cantly greater strength than for �bers retted with Viscozyme (Table 11). erefore, a major objective of enzyme retting with increased �ber strength was reached with pectate lyase followed by EDTA. e commercial enzymes used in the present study represented a mixture of polysaccharidases, for example, cellulases and hemicellulases in some, as well as different types of pectinases. Viscozyme, or EPG, and pectate lyase were effective in attacking pectin and retting �ax, but the two enzymes have different optimal conditions for activity and different modes of action.
Advances for �ax �ber processing could occur with speci�c enzymes and systems. Polygalacturonase (PG) and pectate lyase (PL) are both depolymerizing enzymes for pectin but work in different ways and under different conditions. PG is reported to catalyze random hydrolysis of -1,4 polygalacturonic acid, and PL carries out a nonhydrolytic breakdown of pectates and pectinates by a transelimination split of the pectic polymer [77]. PL is activated by Ca 2+ and usually is active at higher pHs (e.g., [8][9][10] and temperatures (55-60 ∘ C) than PG. Research has been carried out to correlate �ber separation with the degradation of different "pectins, " that is, various functional groups and linkages, using a series of commercial PGs and PL with low xylanase and cellulase activities [130]. Retting efficiency was highly correlated (correlation coefficient of 0.99) with sparsely esteri�ed pectin, but correlations between retting and activities against other pectins were low. Since Ca 2+ binds acidic groups of pectin molecules and various types of pectin are in different regions of the bast [66,81], these data further reveal a coordinated mechanism for degradation of nonesteri�ed �ax pectins with chelators and pectinases. Further, these and other spectroscopic data [131] suggest that the pectins in the middle lamella and those binding the cuticle/epidermis to �ber bundles in �ax stems are targeted by this mechanism. Towards a more cost-effective enzyme retting system along these lines, other work [132] indicated that weak acid with enzymes was effective in separating bast �bers with reduced enzyme levels, likely by removing the Ca 2+ in pectin.
Work with commercial PL has indeed shown efficiency in separating bast �ber from stems ( Figure 8, Tables 10-12). Bioprep levels around 2% with 18 mM EDTA were optimal with the �ax samples used and conditions tested (Table  12). Fiber yield, �neness, and cleanliness were not improved with higher Bioprep levels. Sequential treatment of Bioprep followed by EDTA was the most effective for retting, but combining both enzyme and EDTA also retted �ax. �e procedure most effective for producing �ne, clean �ber was as follows: (a) saturate crimped �ax stems with Bioprep at 2%, (b) incubate for 1 h at 55 ∘ C, (c) without washing, resoak with 18 mM EDTA at pH 12, (d) continue incubation at 55 ∘ C for about 24 h total time, and (e) wash and dry �ber in preparation for mechanical cleaning.
Other work has shown the potential of alkaline pectinases such as PL to ret �ax and the bast plant ramie [127,133]. Bioprep-treated �ber was not tested in the miniature spinning system or for biocomposites to this author's knowledge. An engineered pectate lyase from Xanthomonas campestris, however, was developed and used at 37 ∘ C and pH 8.5 to "bioscour" commercially grown and decorticated linseed �ax [134][135][136]. Matched PL-treated and untreated �bers were then used in manufactured biocomposites. PLtreated �bers were cleaner and �ner than untreated �bers. In some biocomposites, PL-treated �bers performed better than similar but untreated �bers [134]. Other linseed straw samples, which had been le� in the �eld for a few weeks and then decorticated and treated with X. campestris PL for various times up to 46 hr followed by chelator, did not result in improved biocomposites [135]. Still further tests of commercially decorticated linseed straw showed that PLretted �bers, although �ner and cleaner than untreated ones, performed better in tensile strength tests but not in interfacial shear strength tests [136]. Further assessment is required to optimize use of Bioprep and other pectate lyases as a retting enzyme for �ax. It is clear that many factors in�uence the successful production and processing of �bers for textiles and biocomposites. Further, while weak cellulases, such as those in �iscozyme, reduced �ber strength, the resulting �ner �bers o�en spun better than �bers produced by lower �iscozyme levels in blends with cotton. Further assessment is required on �ber characteristics for speci�c applications (e.g., blended textiles or biocomposites), as well as the economics of enzyme retting. It is clear, however, that pectate lyases can separate �ber of non�ber components and retain �ber strength. e designed purpose of Bioprep as a cotton scouring agent, which acts by removing the cotton �ber cuticle, has shown to be effective in several large tests [125,126].
e ARS research on enzyme retting of �ax had ended by 2012, with the retirement of key individuals and closing of the USDA pilot plant. Research and development continues for retting and other �ber applications with enzymes [127,134]. Inotex (Dvur Kralove n.L., Czech Republic) has developed enzymes to assist with �eld retting [128], particularly towards producing consistent �bers in varying weather conditions and including use of oilseed straw (J. Marek, Inotex, Czech Republic, personal communication).
Genetics for plant modi�cation of �ax to improve �ber properties for linen and biocomposites are active areas of research. Related to the idea of improved retting, some research is focused on genetically modifying �ax for improved �ber extraction from linseed stems (Michael Deyholos, University of Alberta, Canada, personal communication). One goal of another program, FIBRAGEN, is to identify genetic markers, including those determining anatomy and physical properties, for expanding �ax markets in textiles and biocomposites (Jörg Müssig, Hochschule Bremen, University of Applied Sciences, Department for Biometrics, personal communication). Advances in plant modi�cation coupled with knowledge of speci�c action of enzyme systems for extracting �ber bode well for new systems to economically extract �bers of high and consistent quality and for directed purposes.

Enzymes for Postharvest Treatment of Flax Fiber and
Yarns. Perhaps one of the most effective uses of enzymes may be in postharvest treatments of �ax �bers to impart speci�c properties. Fiber-modifying enzymes are marketed for a variety of purposes, including enrichment of dew retting, repair of poor quality �ax (such as underretted material), cottonization of bast �ber tows for textiles or biocomposites, tailored �ber length, and processing of rovings to reduce noncellulosic content [128]. e use of enzymes pertinent to �eld retting includes spray applications a�er stalk pulling to minimize the effect of inclimate weather and to better utilize linseed stalks in Europe (J. Marek, Inotex, Czech Republic, personal communication). Weak cellulases may have applications where precision in limited attack on cellulose may be bene�cial, such as for cottonization or shortening of �ber. To this purpose, laboratory tests of �ax pulp treated with commercial pectinases and cellulases showed improved characteristics of hand sheets compared to those prepared by traditional (nonenzymatic) methods [137]. For pulping, the breakdown of the �ber bundles by pectinases and the shortening of the cellulosic �bers by attack of cellulases at the �bernodes improved some paper properties in laboratory studies. e nature of �ax �bers, that is, the lack of limiting lignin in bundles and presence of susceptible �bernodes, provided opportunities for use of these enzymes not possible in highly ligni�ed, woody sources of pulp. e authors further suggested that a more precise attack by speci�c enzymes may provide additional attributes in the pulp.
Research with an atomized enzyme delivery system showed that endoglucanases could be effectively delivered in small amounts onto �eld-retted �bers, likely resulting in attack at the �bernodes to reduce �ber length, strength, and elongation [138]. Application of the atomized method with endoglucanase and extended to other enzymes [36] modi�ed the properties in �ax �ber and in �ax�cotton blended (50�50) yarns (Table 13). ese enzymes were used as supplied, and such mixtures usually have multiple enzyme pro�les against �bers [34,109,129]. While further work is needed to assess speci�c activities, data suggest that all enzyme types were active in atomization, and various properties could be modi�ed. For example, lipase and arabinase improved certain yarn properties, such as increased strength and elongation and reduced neps and thick and thin places. Results further suggest that precise enzyme activities could tailor �ber and yarn properties.

Enzyme
Retting of Other Bast Plants. is paper has focused on �ax structure and composition, with potential for enzyme retting. Emphasis has been placed on the nature of the �ax bast �ber and bundle, speci�cally the lack of high levels of lignin, the binding of Ca 2+ in pectin molecules in the cuticle, and the presence of the more susceptible �bernodes and kink bands within the �bers and bundles. Pectinases, either polygalacturonases or pectate lyases, alone are able to separate �bers from cuticle and core. Flax is �ust one of many bast plants that are economically important for myriad uses throughout the world. Would the same enzyme work for other bast plants as for �ax � Research suggested that bast �bers with other characteristics may require other types of enzymes. Ramie, which is a nonligni�ed, celluloserich bast �ber-like �ax (unpublished data), has been retted with pectin lyase [133]. Enzyme retting of hemp, which is more heavily ligni�ed than �ax, showed some success, but different enzymes or protocols from those with �ax were needed [139]. �enaf is highly ligni�ed in the secondary walls and middle lamella of the bast �bers and bundles and has been retted by chemical means [140,141]. Use of a commercial enzyme, having cellulase and xylanase activities, with chelators and a crimping pretreatment separated the bast tissue to �ber bundles [139]. is process only produced coarse �ber bundles, and a delignifying process seems to be required for effective retting of kenaf. To this end, enzymes from noncellulolytic, lignin-degrading white rot fungi to remove aromatics and leave cellulose [142,143] may �nd applications. Successful, cost-effective, and commercial technologies will have requirements such as the following: selected �ax material, enzyme formulations and conditions to tailor �bers with speci�c properties, integrated cleaning procedures, objective assessment methods to assure high and consistent quality, and directed applications.

Summary and Conclusions
Flax has had a long and illustrious impact on human development for millennia. e long, �ne, and strong bast �bers provide apparel and other textiles, and other varieties of �ax provide linseed and its oil. e textile industry that once �ourished in western Europe has declined, but the desire for �ax and linen is still strong. �uality �bers are still marketed in Europe, and China and other regions desire more quality �ax �bers for products. Even the paper and pulp industries desire cleaner �bers to reduce the amount of chemicals required for deligni�cation. Biocomposites and nonwoven materials are predominant areas of interest, with the automotive industry continuing to focus on natural �ber composites.
In particular, biocomposites are sought for weight and cost savings, improved structural properties, processing bene�ts, and design �exibility and ease. Compared to glass, �ax �bers are lower in cost, lower in density, biodegradable, and similar in elongation at break� tensile strength is lower for �ax.
Woven �ax �bers as insets with resins particularly provide good strength and rigidity in composites. Substantial savings in energy costs are possible with natural �ber mats, which reportedly require about 80% less energy than those made with glass. Flax �ber provides a low cost alternative for glass �ber in reinforced composites. e replacement of glass �bers with �ax for this application, even with its important advantages, is nonetheless a considerable challenge. Consistency in supply and in �ber characteristics is required for �ax �ber to expand further into markets, especially higher value-added products. Reportedly, the best �ax �ber is still produced in western Europe, where climate and grower experience provide quality �bers for textiles. e drive for �bers in biocomposites and other industries has focused on getting �ax �ber from nontraditional linen plants, namely, linseed straw. Field retting, which is the primary method of �ax production, is problematic in that the �bers are o�en poor and inconsistent in quality. Climate is a major factor in quality, and outside western Europe, the major areas of �ax production are o�en in harsh climates for �eld retting. It is in regard to all these factors that research has focused on other methods, including enzyme retting, to improve retting and thereby �ber processing and quality.
Replacement methods for �eld retting have been sought for a long time, but currently, there are no such methods used commercially. Enzyme retting has been researched for several years and is still undergoing development. ere have been positive results, and there is considerable interest in postretting and treatment of roving and yarn to improve their properties. New developments in enzyme production by commercial companies have provided purer and more active pectinases that have promise in enzyme retting. Our work examined several commercial enzymes for retting and focused on the endopolygalacturonase-rich, mixed product Viscozyme plus EDTA and the purer alkaline pectate lyase product Bioprep followed by a commercial EDTA chelator. Protocols were developed on enzyme concentrations and conditions to separate �bers, which were then cleaned in a �ber processing pilot plant and characterized by objective test methods. While Bioprep-retted �bers had good properties of �neness, strength, and cleanliness, tests in textiles or composites have not been carried out. e vast amount of research on enzyme retting indicates that pectinases without the need of complementary enzymes are effective in separating �ber from �ax straw, even linseed straw. e inclusion of a chelator, such as EDTA, greatly reduces the amount of enzyme required and is particularly effective in separating �bers from the cuticle-epidermis layer in linseed straw. Fiber properties can be tailored with the use of speci�c enzymes.
Improving the quality and consistency of �ber from the huge biomass resource of linseed straw has great potential in addressing needs of myriad industries, even that of cottonized �ax �ber for textiles. e degree of processing for �ber cleanliness will depend upon the end product desired, and for some products, the requirements of cleanliness and processing costs are considerably less than for linen fabrics. e desire for quality apparel, however, continues to be important in �ax and linen products, and improved processing methods should not overlook this important and historical industry. To this end, the Crailar process reportedly uses enzymes in a proprietary process to produce so�, �ne �ax �bers for blending with cotton in an agreement with the Hanes clothing industry. One of the most important uses of enzyme might be in tailoring speci�c properties in postharvested materials to improve low-quality �bers.
Fibers and seeds are two historical products of �ax with traditional and continuing economic importance around the world. It is likely, however, that the usefulness of �ax will not be limited to �ust �ber and seed, as physical, chemical, and biotechnical methods uncover more products. Production of �ax �bers by enzymatic or other means of retting, followed by mechanical processing, generates bast �bers for many industrial needs and massive amounts of by-product wastes. is waste material, consisting of cuticle, shive, and �ber fragments, is already paid for and is localized at the processing plant. e potential for coproducts from processing �ax �ber is huge. Mention has been made of lipids �sterols, policosanol-type lipids, and waxes) from the cuticle in dust and of activated carbons and extracted aromatics and sugars from the shive. Currently, there is a burgeoning interest in microcrystalline cellulose from plants and their potential for value-added products ranging from biocomposites for medical devices to solidi�ed liquid crystals. Considerable work is still needed to overcome substantial problems and directed applications towards reaching the huge potential for cellulose nanocomposites. e highly crystalline and oriented nature of cellulose in �ax �bers warrants consideration for its properties in nanotechnology. e source, chemistry, structure, and crystalline nature of the native �ax bast �bers, particularly in regard to the response to speci�c enzymes, may offer a contribution to this growing area of research and technology.
Indeed, linen most useful, Linum usitatissimum, is poised to continue to expand as a supplier of useful products to mankind throughout the world.