September 11, 2008

Particle Release From Woven Cellulosic Substrates

By Manian, Avinash P Lenninger, Margit; Bechtold, Thomas; Steinlechner, Erik

Abstract The liming propensity of a woven cellulosic substrate is investigated as a function of substrate characteristics and of different parameters in the test environment. The results from the linting propensity tests parallel those observed in liquid-solid extraction processes, where solutes are extracted from a matrix of other insoluble solids by selective dissolution in a liquid. The equation quantifying solute extraction in such systems also proves to be a good fit for the linting propensities observed in this work. Hence, it maybe possible to regard linting phenomena as extraction processes and to use the equation quantifying solute extraction to predict substrate linting as a function of both substrate characteristics and the test environment. Key words particle release, cellulosic fabrics, COD tests, inter-fiber friction, linting

(ProQuest-CSA LLC: ... denotes formulae omitted.)

In controlled environments such as clean rooms and hospital operating rooms, a major source of contamination is particles released from fabrics used in clothing or as wipes. There are different methods available to characterize particle release from fabrics, which are largely fibers and also referred to as "lint" [1- 5]. However, fabric linting propensities are strongly influenced by the conditions employed in tests and there is little correlation between results obtained from different test methods [1], which makes it difficult to formulate a unified model to characterize fabric linting propensities.

Mattina and Paley [6] attempted to characterize fabric linting by quantifying particle release as a function of the mechanical stress applied to fabrics. They placed wet fabric samples flat in a shallow tray and dragged known weights across their surface through defined distances at defined velocities. After each application of stress, the tray was filled with water to extract the free particles generated, and particle amounts in water were quantified with a particle counter. The authors empirically related particle release to applied stress with

P = P^sub 0^ + kS^sup n^ (1)

where P is the cumulative number of particles released up to each measurement point, P^sub 0^ is the number of particles released under zero stress, S is the cumulative stress applied to fabric (J/ m^sup 2^) and k and n are empirical constants, believed to be related to fabric construction parameters and material type.

The equation, however, does not account for environmental influences on particle release from substrates. The environment in which substrates are tested or used exerts a strong influence on their linting propensities as evinced by the wide variability in results obtained from different tests, many of which employ conditions that simulate actual use. In our work, we measured linting from a woven cellulosic substrate and attempted to characterize its linting propensity as a function of both its structural characteristics and the test environment.

Experimental details


The substrate used in this study was a composite structure comprised of four layers of woven cotton fabrics sewn together at the edges such that the fabric ends were tucked inwards into the structure. The cotton fabrics were made of 16.3 tex and 24.5 tex warp and fill yarns at a thread density of 16 ends and 12 picks per centimeter. The substrate dimensions were 40 x 40 cm^sup 2^ and weighed on average 38.42 +- 0.28 g.

Linting Tests

Linting propensities were evaluated by rinsing substrates by hand in specific volumes of deionized (DI) water in n successive sets of r rinses each, at a rate of around 35-40 dips per minute with each rinse set carried out in a fresh volume of rinse water. The lint released was collected by filtering the rinse water through glass- fiber filters of 1 im porosity (PALL, Type A/E 47 mm), after which the filters with lint were allowed to dry overnight at ambient conditions.

Lint Quantification

Lint quantification was achieved by subjecting the collected lint to chemical oxygen demand (COD) tests. The COD test is commonly used in wastewater analyses as an indirect measure of the amount of organic material in water. The organic material is digested in a strong chemical oxidant; the amount of oxidant consumed in the reaction is measured and expressed as "milligram of O2 per liter of water", and used as a quantitative measure of the amount of organic material in water. The general chemical reaction is [7]

... (2)

where n = 4x + y - 2z.

The filters with lint, after drying, were introduced along with 10 ml DI water into digestion tubes, and the COD tests were performed according to the photometric method described in ISO 15705:2002 [8] with test solutions prepared in the laboratory with analytical grade reagents. In exception to the prescribed procedure, mercury (II) sulfate was not included in reagent solutions since DI water was used in all experiments and significant levels of chlorine were not expected in the system; also, 15 ml reagent solutions were used in analyses instead of the 5 ml prescribed in the standard.

The glass-fiber filters were not affected by the chemical digestion process and remained as a solid residue in the digested solutions. Photometric measurements of the digested solutions were performed on a Shimadzu UV-VIS spectrophotometer after centrifuging the cooled solutions at 1000 rpm for 5 min in a laboratory centrifuge. A blank sample containing only the filter was introduced in each set of tests and all COD values were corrected for blank.

The COD values from successive rinse sets were added together as shown in the following equation and the cumulative values were used to chart substrate linting propensities with increasing rinse levels:

... (3)

where COD^sub N^ is the cumulative COD value after N rinse sets and COD^sub n,r^ is the COD value obtained in the nth rinse set with r number of rinses.

In all COD measurements, cellulosic solids were dispersed in a total solution volume of 25 ml made up of 10 ml DI water and 15 ml reagent solution. Hence, the volume term in the COD unit "mg O2/l" refers to the total volume of test solutions and the amount of oxidant consumed in the chemical digestion, expressed in equivalent amounts of oxygen, is proportional to the amount of solids. In this manner, a volumetric concentration unit was used to quantify gravimetric amounts of released fibers.

Binder Treatments

The degree of linting from a fabric will be influenced by the degree of cohesion/adhesion among its fibers. To alter the level of binding between fibers in substrates and to thereby study the influence of cohesion/adhesion among fibers on linting, substrates were treated with commercial binder formulations commonly used in textile processing. Two binder types were used: binder A, an aqueous dispersion of Polyurethane commonly used in three-dimensional foam printing; and binder B, an aqueous dispersion of polyacrylate commonly used in pigment printing. The binders were applied on substrates from formulations containing 20-100 g/1 of the commercial product with a pad-dry-cure process, by padding substrates at 2 bar and drying in a laboratory stenter at 100[degrees]C for 6 min. Substrates treated with binder A were cured at 150[degrees]C for 6 min and those treated with binder B at 170[degrees]C for 1 min, based on recommendations of the manufacturer. There were no significant differences in wet pickup between the two binder treatments and the values ranged from 124% to 115% as the binder concentration increased from 20 to 100 g/l.

Inter-fiber Friction Measurements

The levels of cohesion/adhesion among fibers were evaluated by measuring the inter-fiber friction in yarns from both binder- treated and untreated substrates. A method to measure inter-fiber friction in yarns was described in a previous communication from this laboratory [9], which consists of mounting yarn specimens on a yarn twist tester at increasing levels of pre-tension and counting the number of reverse twists required for the yarn to be pulled apart in two owing to fibers slipping past each other. This method did not yield satisfactory results for yarn specimens obtained from binder-treated substrates. The yarns were not pulled apart under the levels of pre-tension commonly used in the test, and were observed to break as a result of fiber rupture under higher levels of pre- tension. Therefore, a modified method was used to measure yarn inter- fiber friction in this study.

First, the yarn twist in untreated substrates was determined on a Zweigle Yarn Tester D312 (Zweigle-Reutlingen, Germany) at a pre- tension of 5 cN across a gauge length of 25 cm, and was found to be 236 +- 7 twists/25 cm and 178 +- 5 twists/25 cm in the warp and fill yarns, respectively.

Fresh specimens of warp and fill yarns were then carefully isolated from test substrates so as to avoid significant twisting/ de-twisting in specimens, and mounted on the yarn tester with a 5 cN pre-tension across a 25 cm gauge length. The yarn specimens were then given reverse twists (236 and 178 reverse twists for warp and fill yarns, respectively), removed from the yarn tester and mounted on an Instron Tensile Strength Tester at the same pre-tension and gauge length. The maximum force required to pull the untwisted yarn apart in two, measured at a 1 cm/min rate of extension, was used as a direct measure of the inter-fiber friction in yarns. The broken yarn ends were examined under an optical microscope to confirm that the break in untwisted yarns was a result of fiber slippage rather than fiber rupture. The specimens were conditioned in a standard atmosphere of 20 +- 2[degrees]C and 65 +- 4% relative humidity for over 24 h before tests, which were performed under the same conditions. The individual fabric layers in substrates were numbered in sequence from one to four, starting from the layer that was uppermost during pad-dry-cure treatments. In untreated substrates, the fabric layers were numbered from one of the outermost layers arbitrarily designated as the first layer. The inter-fiber friction was measured in warp and fill yarns from the first and third fabric layers.

Statistical analyses of the data were performed at a 0.05 level of significance with the software SPSS(R). The error bars in graphs represent data variability in terms of +-1 standard deviation.

Results and Discussion

Preliminary Tests of the COD Method

A series of preliminary tests were conducted to determine the suitability of the COD method to quantify lint amounts, by subjecting 0.002-0.010 g of different cellulosic materials (including the experimental substrate) to COD tests. The blank sample in these tests contained only the test solution. The COD values (corrected for blank) of the different materials are shown as a function of their weight in Figure 1.

Within the weight range examined, there was a linear increase in COD with increase in material weight with no significant differences between the different material types. A linear regression of the data yielded the following relationship between material weight and its COD value (R^sup 2^ = 0.9765; standard error, epsilon = 47.48):

COD (mg O2/l) = 107008.82 x Material weight (g) (4)

The simple and direct correlation between material weight and its COD value shown in (4) made it possible to quantify lint amounts in terms of weight. However, lint quantification in terms of COD values provided higher sensitivity and greater resolution in comparisons of substrate linting propensities.

Linting Propensity Tests

A set of experiments was conducted to study the influence of test parameters on substrate linting propensities. The effect of rinse water volume on substrate linting propensities was examined by subjecting untreated substrates to 80 rinses in eight successive sets of 10 rinses each in: (a) 1.51, (b) 3.0 1 and (c) 4.5 1 DI water. The tests with 3.0 and 4.5 1 DI water were accomplished in 5 1 plastic jugs. The tests with 1.5 1 DI water were accomplished in 21 plastic jugs since the substrates were not completely submersed in this volume of water when rinsed in 5 1 jugs. The resulting linting propensities are shown in Figure 2.

The range of COD values in the linting propensities shown in Figure 2 correspond to 0.004-0.030 g of released lint, which amount to 0.01-0.08% of total substrate weight. There was little difference between linting propensities obtained in rinse volumes of 1.5 and 3.0 1 water, but there was a significant increase in substrate linting propensities with increase in rinse volume from 3.0 to 4.5 1 and the differences increased with increasing rinse levels.

To examine the effect of rinsing scheme on substrate linting propensities, untreated substrates were subjected to 80 rinses in 3.0 1 DI water in 5 1 plastic jugs using two schemes: (d) four successive rinse sets with 10 rinses each in the first three sets and 50 rinses in the last set, and (e) a single rinse set of 80 rinses. The values from series D and E were compared with those from series B above, as shown in Figure 3.

The range of COD values in the linting propensities shown in Figure 3 correspond to 0.003-0.022 g of released lint, amounting to 0.01-0.06% of total substrate weight. The linting amount obtained in the single rinse set from series E was similar to those obtained in the first rinse sets from both series B and D, despite the large differences in the number of rinses per set between the three series. The linting propensity values from series D were lower than those from the first four rinse sets in series B by small but statistically significant amounts, which appeared to remain the same across the four rinse sets compared. These differences, remarkable especially with regard to the first three rinse sets as the experimental conditions were identical in both series, result from operator variability and highlight a shortcoming of the hand- rinsing technique in linting measurements.

It was also observed that within the individual test series, the differences in lint amounts between successive rinse sets were similar regardless of the number of rinses per set. This suggests that there was lint transfer from substrates to rinse water and vice versa, and that equilibrium was established between the two processes within the minimum number of rinses per set, namely 10 rinses.

These phenomena in linting propensities are analogous to those observed in multi-stage liquid-solid extraction or leaching, where solvents are used to selectively extract solutes of interest from a matrix of other insoluble solids. The cumulative amounts of solutes extracted in such processes, when fresh solvent is used in each extraction stage, may be determined with the following equation [10, 11]:

where G^sub A^ is the cumulative amount of solute extracted, G^sub T^ is the total amount of extractable solute present in solid, V^sub A^ is the total volume of extracting solvent, V^sub B^ is the volume of solvent adhering to solids, K is the distribution ratio of solute between solvent and solid and N is the number of extraction stages.

We fit Equation (5) to the data obtained in linting propensity tests by considering released lint to be the extracted solute. The values of V^sub A^ (volume of rinse water), V^sub B^ (volume of water adhering to substrates) and GA (cumulative linting values obtained at the end of each rinse set) corresponding to N (number of rinse sets) from series A-D were introduced in (5) and the values of G^sub T^ (total amount of releasable lint) and K (distribution ratio of lint between water and substrate) which best fit all experimental series were iteratively determined by the least squares method.

To determine the values of V^sub B^, fresh substrates were rinsed in 1.5, 3.0 and 4.51 DI water in 2 and 5 1 plastic jugs as in the actual tests. At the end of each rinse, the excess water in substrates was allowed to drain freely and the wet towels were weighed. The difference between substrate weights from before rinsing (dry weight) and after each rinse (wet weight) were used to estimate the volume of water adhered to substrates and the values are listed in Table 1 along with those of V^sub A^.

Since the primary differences in experimental conditions among the diff erent series were in V^sub A^ and V^sub B^, it should have been possible ideally to obtain common values of G^sub T^ and K for series A-D. However, it was found that values of G^sub T^ = 4540 mg O^sub 2^/l and K = 0.0039 which best fit series B and C were not suitable for series A and D, as seen from the values of R^sup 2^ and standard error (e) listed in Table 1, In separate curve fitting exercises it was found that values of G^sub T^ = 3425 mg O2/l and K = 0.0039 yielded the best fit for series D, with R^sup 2^ and e values of 0.9603 and 87.80, respectively. For series A, the best fit with R^sup 2^ and e values of 0.9836 and 77.41, respectively, was obtained by maintaining G^sub T^ = 4540 mg 02/l and K = 0.0039 but changing V^sub A^ to 2.41. The values of 3425 and 4540 mg O2/l for G^sub T^ correspond to 0.032 and 0.042 g of releasable lint, amounting to 0.08 and 0.11% of total substrate weight, respectively.

The high values of R^sup 2^ and low values of e shown above indicate a good fit between (5) and data from the linting tests. The lower value of G^sup T^ for series D is attributed to a lower amount of stress imparted to substrates during this series of tests as compared to the other test series, leading to a decrease in the generation of releasable lint. In series A, the substrates were rinsed with 1.51 water in 21 jugs to ensure complete submersion of substrates in water, which was not found possible using 5 1 jugs. The resulting difference in the value of V^sub A^ for obtaining the best fit reflects the influence of the degree of contact between substrate and water, in addition to that of total rinse water volume, on substrate linting propensities.

In the discussion above, we examined the influence of the test environment on substrate linting propensities. In the discussion that follows, we examine the influence of substrate structure (in terms of the degree of binding among its constituent fibers) on its linting propensities.

Effect of Binder Treatments

The effect of binder treatments on the inter-fiber friction in yarns from the treated substrates is shown in Figure 4, which shows the maximum force required to pull apart untwisted yarns by fiber slippage for yarns isolated from the first and third fabric layers.

There were no significant differences in maximum force between the warp and fill yarns from a given fabric layer, so the values for the two yarns were averaged and one set of data points was plotted for each layer. The yarns from the untreated substrate (0 g/l binder concentration) broke apart under the force of pre-tension alone and were assigned nominal force values of zero.

In general, binder A treated substrates exhibited higher values of maximum force in comparison to binder B treated substrates. The force values increased with increase in binder concentration of up to 80 g/l for binder A and 60 g/l for binder B but did not change significantly thereafter, suggestive of an upper limit to the gains achieved in inter-fiber friction with increasing binder concentration. There were significant differences in the maximum force required to pull apart yarns from the two fabric layers tested, which was used as a measure of the degree of binder penetration in substrates. Hence, it was concluded mat there was a greater depth of binder A penetration in substrates, as lower differences in inter-fiber friction between fabric layers were observed in these substrates. The linting propensities of binder- treated substrates were evaluated in 3.0 1 DI water using the methodology described for untreated substrates in series D. The cumulative COD values after each rinse set are listed in Table 2 and those after four rinse sets are plotted as a function of binder concentration in Figure 5.

In apparent discord with increased cohesion/adhesion among fibers, the linting propensities in substrates treated with binder A were higher than that in the untreated substrate. When comparing the cumulative values obtained after four rinse sets, a dramatic increase by close to 160% was observed in progressing from untreated substrates to those treated with 20 g/l binder A. Further increases in binder concentration resulted in a comparatively marginal, but statistically significant, increase of 12-14% in COD values levels up to binder concentrations of 60 g/l beyond which the values did not change significantly. In linting tests on binder A treated substrates, we observed the presence of binder residues on filters along with the collected lint. Hence, the higher linting propensities are attributed to the presence of unfixed binder residues in substrates that were released during tests.

There was no evidence of binder residues being released in tests with binder B treated substrates and, in concordance with their increased inter-fiber friction, in a comparison of cumulative COD values after four rinse sets these substrates exhibited 21-60% lower COD values as compared with the untreated substrate.

It is difficult to evaluate the weights of released particles from the COD obtained in tests on binder-treated substrates, as the measured values are comprised of contributions from both fibers and binder residues released from substrates. However, it is possible to obtain rough estimates on the basis of a few simplifying approximations. If the smallest oxidizable unit in cellulose is regarded to be -CHOH which has a molecular weight of 30 units and requires two oxygen atoms for complete oxidation and the smallest oxidizable unit in polyurethane to be -CH^sub 2^ which has a molecular weight of 14 units and requires three oxygen atoms for complete oxidation, then the average molecular weight units oxidized per oxygen atom for the two materials, 15 for cellulose and 4.67 for polyurethane, are in a ratio of approximately 3:1. Hence, the weight of binder A residues oxidized per unit COD maybe approximated as being three times that of cellulose as calculated with Equation (4). Similarly, the smallest oxidizable unit in polyacrylate maybe regarded to be -CH^sub 2^-CH-COOH, which has a molecular weight of 72 units and requires six oxygen atoms for complete oxidation yielding a value of 12 molecular weight units oxidized per atom of oxygen for binder B which is 0.80 times that required for cellulose. Hence, the weight of binder B residues consumed per unit COD maybe approximated as being 1.25 times that of cellulose as calculated with Equation (4).

Such approximations maybe used to estimate particle amounts from the COD values listed in Table 2. For example, particle amounts corresponding to the cumulative COD value of 3729.71 mg O2/l obtained after the fourth rinse set for substrates treated with 100 g/1 binder A may be estimated as ranging from 0.035 to 0.105 g depending on the proportion of fibers and binder residues in the released particles. That of 798.13 mg O2/l obtained after the fourth rinse set for substrates treated with 100 g/1 binder B may be estimated as corresponding to 0.007-0.009 g of released particles.

We fit Equation (5) to the data from the linting tests on binder- treated substrates. No significant differences were found in V^sub B^ between untreated and binder-treated substrates so V^sub B^ was maintained at 0.108 1, while V^sub A^ was 3.0 1. The value of K persisted at 0.0039 in the best fit curves for all bindertreated substrates, while the values of G^sub T^ changed with binder type and concentration, as listed in Table 3 which includes the R^sup 2^ and e of the fits and also the corresponding weight approximations calculated as earlier.

The values of G^sub T^ were higher for binder A treated substrates as compared with untreated substrates and those treated with binder B and, in general, increased with increasing binder A concentration, but decreased with increasing binder B concentration. The higher values of G^sub T^ in binder A treated substrates substantiates the previous attribution of their higher linting propensities to the increased presence of releasable particles. On the other hand, treatments with binder B resulted in a lowering of releasable particle amounts in substrates and thereby reduced substrate linting propensities.

The linting propensities observed in this work conform to what is generally understood about the effect of substrate on linting or particle release. Particle release increases with increase in stress applied on substrates and decreases with increase in degree of adhesion/cohesion among its fibers. In apparent exception to the latter concept, substrates treated with binder A exhibited higher particle release despite increased inter-fiber friction among their constituent fibers, which was traced to the presence of binder residues among the released particles. Treatments with binder A appeared to result in an increase in the amounts of releasable particles owing to the presence of unfixed binder residues and there is a causal relationship between releasable particle amounts and particle release.

Apart from substrate structural characteristics, the test environment also influenced substrate linting propensities. An increase in rinse volume resulted in increased substrate linting. There was evidence of a two-way transfer of lint between substrate and the environment, and there appeared to be an establishment of equilibrium between the two processes. The phenomena observed in substrate linting propensities in this work were similar to those observed in multi-stage liquid-solid extraction processes where solvents are used to selectively extract solutes of interest from a matrix of other insoluble solids. The equation quantifying solute amounts recovered in liquid-solid extractions (Equation (5)), which is applicable for extractions by solute dissolution and may not be strictly valid for particle release by mechanical means, was used to characterize substrate linting propensities. From the empirical evidence obtained in this work, the equation proved to be a good descriptor of substrate linting propensities.

According to (5), fabric linting propensities increase with increase in the amount of releasable lint (G^sub T^), which increases with increase in applied stress and/or decrease in cohesion/adhesion between its fibers. Under a given set of conditions for testing or usage, the values of G^sub T^ are also expected to increase progressively with time as repetitive applications of stress will likely diminish the forces of cohesion/ adhesion among the constituent fibers in fabrics. According to the equation, fabric linting increases with increase in the distribution ratio (K) of lint between water and substrate, which is likely to be influenced by material type and may differ for different polymer types. Linting also increases with increase in the ratio V^sub A^/ V^sub B^. Hence, fabrics with a higher capacity to imbibe water, and thereby higher V^sub B^, will exhibit comparatively lower linting.

The combined effect of different factors on substrate linting propensities as described by (5) appears to tally with our intuitive understanding of linting phenomena. It also appears to match what is routinely observed in systems where there is relative motion between fabrics and water or other liquids, e.g. when fabrics are used as wipes.


The linting propensities of a woven cellulosic substrate have been determined by hand rinsing substrates in DI water followed by quantification of released lint with COD tests. The hand rinsing technique of linting propensity measurements suffers from the drawback of being highly susceptible to operator variability, but the potential for error was minimized by employing a single operator in all tests. It was found that COD tests could be used to quantify amounts of released lint, because the values were linearly related to fiber weight and were sensitive to small changes in fiber amounts. However, COD tests indiscriminately measure the load of organic material released from substrates, and do not yield any information on the type or nature of particles released.

The trends obtained in substrate linting propensities paralleled those observed in liquid-solid extraction processes, where liquids are used to selectively extract solutes by dissolution from a matrix of other insoluble solids. The empirical evidence obtained in this work showed that the equation describing solute recovery in liquid- solid extractions closely fits linting propensity trends. The combined effect of different factors on linting propensities as described by the equation also conforms to what is known about linting in other systems with relative motion between fabrics and water or other liquids, and hence may be used as the basis of a model to predict substrate linting propensities as a function of both substrate characteristics and the test environment.


Extracts from this paper were presented at Ambience 05, International Scientific Conference on Intelligent Ambience and Weil- Being, Tampere, Finland, 19-20 September 2005. We are grateful to the Versuchsanstalt-Textil of HTL-Dorn-birn for the use of their testing facilities. We are also grateful to Anton Heinzle and Klaus Geismayer of HTL-Dornbirn for their help with some of the mathematics in this work. This work was made possible by the financial and material support received from the Christian-Doppler Research Society (Vienna, Austria) and Lohmann & Rauscher GmbH. References

1. Neumann, T, Bohnen, J., and Kruessman, H., Determination of Particles in Laundered Textiles, in "Proceedings of the 40th International WFK Detergency Conference", Strasbourg, France, 2002, pp. 192-198.

2. Paley, S. J., Cleanroom Wipers: State-of-the-Art Evaluation Techniques, Cleanrooms, 10, 52-56 (1996).

3. Bhattacharjee, H. R. and Paley, S. J., Comprehensive Particle and Fiber Testing for Cleanroom Wipers, J. Inst. Environ. Sci. Technol., 41, 19-25 (1998).

4. Atterbury, O., Bhattacharjee, H. R., Cooper, D. W., and Paley, S. J., Comparing Cleanroom Wipers with a Dry Abrasion Resistance Test, MICRO, 15, 83-100 (1997).

5. Wagner, J. R., and Hagner, H. A., An Image-analysis Technique to Measure Linting of Hospital and Clean-room Gowns, Tappi, 80, 167- 172 (1997).

6. Mattina, F. M., and Paley, S. J., Assessing Wiping Materials for Their Potential to Contribute Particles to Clean Environments: Constructing the Stress-Strain Curves, J. Inst. Environ. Sci. Technol., 34, 21-28 (1991).

7. Graner, C. A. E, Zuccari, M. L., and Pinho, S. Z., Determinacao da Demanda Quimica de Oxigenio em Aguas por Espectrofotometrica Simultanea dos Ions Cromio (III) e Dicromato, Ecletica Quimica, 23, 31-44 (1998).

8. ISO 15705:2002: "Water Quality-Determination of the Chemical Oxygen Demand Index (ST-COD)-Small-scale-Sealed-tube Method".

9. Campos, R., and Bechtold, T., Fiber Friction in Yarn-a Fundamental Property of Fibers, Textile Res. J. , 73, 721-726 (2003).

10. Bungay, H. R. III, "Leaching - Stagewise Calculations", http:/ /, (accessed 3 March 2006).

11. Bittel, A, Flussig-Flussig-Extraktion, in "Ullmans Encyklopadie der Technischen Chemie", Vol. 2/1, 3rd edn, Foerst, W. (ed), Urban & Schwarzenberg, Munich, 1961, pp. 77-105.

Avinash P. Manian, Margit Lenninger and Thomas Bechtold1

Christian-Doppler Laboratory for Textile and Fiber Chemistry in Cellulosics, Institute of Textile Chemistry/Physics, University of Innsbruck, Hoechsterstrasse 73, 6850 Dornbirn, Austria

Erik Steinlechner

Lohmann & Rauscher GmbH & Co. KG, Research and Development, Kirchengasse 17, 2525 Schoenau, Austria

1 Corresponding author. Tel: +43-5572-28533; Fax: +43-557228629; e-mail: [email protected]

Copyright Textile Research Institute Aug 2008

(c) 2008 Textile Research Journal. Provided by ProQuest LLC. All rights Reserved.