437

MAY 2005

437

Influence of Fiber Surface Purity on Wicking Properties of Needle-Punched Nonwoven after Oxygen Plasma Treatment JO VERSCHUREN,1 PETER VAN HERZELE, KAREN DE CLERCK,

AND

PAUL KIEKENS

Department of Textiles, Ghent University, Belgium ABSTRACT Polyester and meta-aramid nonwoven samples were treated at reduced pressure in a radiofrequency oxygen plasma for between 5 and 30 seconds. The hydrophilic effect of the plasma treatment was assessed by three consecutive vertical wicking tests on each sample. The influence of cleaning of the fiber surface before as well as after the plasma treatment was determined. The presence of fiber surface additives had a positive influence on first-time wicking. On the other hand, when the layer of surface additives is plasma treated rather than the fiber polymer surface, the treatment effect is easily washed away. Log(time) wicking curves and summed wicking height plots are introduced as a means to help the interpretation of wicking data.

In the course of the last five decades numerous publications and patents have reported on the successful use of an electrical discharge– under vacuum or at atmospheric pressure–for the improvement of a wide range of textile properties, such as hydrophilicity [1, 14], dye exhaustion [9, 16, 18], adhesion [10, 11] and shrinkproofing [4, 5]. Generally, the fiber surface is cleaned from fiber surface additives (fiber processing chemicals) prior to plasma treatment. These surface additives improve frictional and electrostatic properties as required during fiber processing, such as web formation and needle-punching. Their removal from the fiber surface occurs via various methods, ranging from industrially feasible processes such as gentle washing with a detergent [21], as well as scouring and bleaching [2, 19], to extensive laboratory (Soxhlet) extraction in one or more solvents (acetone [15], benzene and water or methanol [8, 12], perchloroethylene, CCl4 and ethanol [6], CCl4, acetone and dimethyl sulfoxide (DMSO) [13], etc.) Although the latter cleaning methods are often required for fundamental studies, they are not feasible in the textileindustry context due to time, cost and environmental constraints. As a consequence, when one has the ambition to introduce a plasma technology into the textile industrial setting, the actual state of the fiber surface should be taken into account, and pre-plasma cleaning of

1 To whom correspondence should be addressed: Jo Verschuren (Mr.) Ghent University Department of Textiles Technologiepark 907 9052 Gent Belgium Telephone: ⫹32 9 264 57 40 (direct), ⫹32 9 264 57 35 (secretary) Fax: ⫹32 9 264 58 46 e-mail: jo. [email protected]

Textile Res. J. 75(5), 437– 441 (2005) DOI: 10.1177/0040517505054170

the fiber surface should be done in a feasible washing step. The textile structures treated in this study were made of 15 micron thick fibers. In this case a 0.4% weight-offiber (%w.o.f.) addition of surface additives covered the fibers with a layer of average thickness 15 nm. This thickness is comparable to the widely acclaimed instant working depth of a plasma (5 to 20 nm). Accordingly, for most as-received industrial fabrics it is the surface additives that are plasma treated, rather than the fiber polymer surface. Only when this layer has been removed by plasma physical and chemical etching is the full fiber polymer surface exposed and treated. The purpose of this study was to assess the influence of fiber surface additives on the result of a standard plasma treatment aimed at improving the hydrophilic properties of a hydrophobic fiber polymer surface. Both wettability and rewettability were tested through the wicking properties of the (un)treated samples. Wicking properties were shown to be of use in the assessment of plasma-induced changes in fiber surface hydrophilic properties [3, 7, 20].

Experimental SAMPLING

AND

PLASMA TREATMENT

The samples were 490 g/m2 poly(metaphenylene isophthalamide), (meta-aramid/MA) or 550 g/m2 poly(ethylene terephthalate), (polyester/PET) nonwoven fabrics. Both needle- punched nonwovens were reinforced by a wide maze fabric of the same fiber type, ensuring dimensional stability. Surface additives were present at a concentration of 0.40 ⫾ 0.05 %w.o.f., as determined © 2005 Sage Publications

www.sagepublications.com

438 from a 1 hour Soxhlet extraction with acetone (DIN54278-1-1995). A 20 cm ⫻ 20 cm sample area was plasma treated centro-symmetrically between two perforated aluminum plate electrodes that were 4 cm apart [17]. Pumpdown to 13 Pa (100 mTorr) took 3 minutes, after which a flow of oxygen increased the pressure to 53 Pa (400 mTorr). The estimated gas flow velocity at 53 Pa was 18 to 20 cm/second perpendicular to the nonwoven surface. Two minutes after pressure equilibrium a 13.56 MHz, 71 mW/cm3 (375 mW/cm2 nonwoven) plasma was ignited for 5, 10, 15 or 30 seconds. The short treatment durations reflect industrial productivity. For a treatment length of 10 or 15 m in commercial vacuum plasma reactors a 5 to 30 second treatment corresponds to a fabric speed between 120 to 20 m/minute and 180 to 30 m/minute, respectively. After the plasma treatment the oxygen gas flow was maintained for another 2 minutes before vacuum break. WICKING TESTS The treated samples were left in a controlled climate (20 ⫾ 2°C, 65 ⫾ 2% R.H.) for 24 ⫾ 2 hours before proceeding with DIN 53924 vertical wicking tests. The result of the test is a wicking curve, expressing the wicking height (a.k.a. capillary rise, in mm) in a vertically positioned sample as function of wicking time (up to 600 seconds). In this study the abscissa is expressed as log (time), by which the wicking curves are more or less linearized. The use of a logarithmic time scale has the advantage of (a) stressing the industrially important early moments of wicking, and (b) avoiding the pile-up of data points at short wicking times. From the plasma-treated sample, four 3 cm ⫻ 20 cm repeat subsamples were cut. The influence of the plasma treatment duration on wicking can further be presented by integration of the wicking height over the full wicking time, resulting in “summed wicking height” (SWH) plots. An overview of the experimental setup for wicking is given in Table I. Wicking tests were done for both the MA and PET nonwovens on as-received samples and on samples of which the fiber surface was cleaned prior to the plasma treatment. Pre-cleaning of samples was done in a 1% solution of isopropanol in water at 40°C to remove all water-soluble matter. After rinsing, and gentle centrifuging to remove capillary water, the samples were oven dried for 2 hours at 105 ⫾ 2°C. The influence of plasma treatment duration was assessed by comparison with a non-plasma-treated sample with similar samples plasma treated for 5, 10, 15 or 30 seconds. All of the samples were subjected to three consecutive wicking tests (runs). First-time wicking of as-received and

TEXTILE RESEARCH JOURNAL TABLE I. Overview of wicking test procedure. as-received

pre-cleaned

first run

first run 3 drying 4

second run

second run 3 washing 4

third run

third run

plasma-treated samples produced a brown-yellowish band moving with the rising water front. This band was presumed to consist of water-soluble products created or modified during the plasma treatment. The possible inhomogeneities in hydrophilicity thus created were fixed on the fiber surface by overnight drying of the wetted samples, and assessed during a second wicking run. The same samples were then gently hand-washed in a 1% isopropanol solution as described before. Then a third wicking run was done on these washed samples. The use of isopropanol was required to prevent traditional surface active agents from interfering with the spectroscopic analysis performed on the samples in parallel with the wicking tests (results to be published later). Special care was taken not to change the nonwoven structure during the different sample handling steps. Whereas the consecutive wicking runs for as-received samples reflect the stability of the plasma-treated layer of surface additives towards water, the runs for pre-cleaned samples reflect the stability of the plasma-modified fiber polymer surface. From the four repeat determinations, the following ranges of 95% confidence limit were determined for the wicking height in as-received MA samples: 0.5–1.3 mm (first run), 0.5–2.9 mm (second run) and 0.5–1.7 mm (third run). For the pre-cleaned samples the confidence limits are: 0.5–1.5 mm (first and second runs), 0.5–1.9 mm (third run). For each of the wicking times separately the corresponding wicking heights are significantly influenced by the treatment duration, as determined from a one-way ANOVA. Results for the PET samples are very similar.

Results and Discussion WICKING

OF

AS-RECEIVED SAMPLES

The averaged wicking curves for as-received MA samples are presented in Figure 1. Fiber surface additives create an environment for considerable first-time wicking, even without plasma treatment. The influence of treatment duration is modest. The wicking height increases in a reasonably linear manner with log (time) up to 3 to 5 minutes of wicking, after which it becomes constant. In the second run the influence of plasma

MAY 2005

439 in a linear fashion, indicating a regained homogeneity of the fiber surface throughout the sample. The separation of the curves with regard to treatment duration was good, indicating how longer treatment durations were required before the actual fiber polymer surface can be modified by the reactive plasma species. It can be expected that optimized plasma treatments of possibly several minutes will result in good wicking properties of the as-received samples, even after washing. The SWH plots for as-received samples (Figure 2) show a systematically lower wicking performance for PET samples in comparison with MA samples, in the first as well as the second wicking run. This could be due to a difference in porous structure and the use of surface additives with a lower hydrophilicity. In the second run the influence of plasma treatment duration has increased, but more so for the MA samples. The wicking properties of washed samples were far inferior even to the wicking properties of as-received samples that were not plasma treated, and this was true for both fiber types. The third wicking run SWH plots show that for the combination of textile structures and plasma treatment conditions applied in this study it required a plasma treatment of longer than 30 seconds before the fiber polymer surface began to be modified through the layer of surface additives.

FIGURE 1. Averaged wicking curves for as-received MA samples, as function of plasma treatment duration. The top and center graphs are for the first and second wicking runs, respectively. The bottom graph presents the four repeat wicking curves for the third run (washed samples). Legend: plasma treatment of 0 seconds (X), 5 seconds (E), 10 seconds ({), 15 seconds (‚), 30 seconds (䊐). The arrows indicate increasing plasma treatment duration. Note the difference in ordinate scale.

treatment duration was increased but overall wicking performance was decreased. This can be expected because of the elution– during the first run– of hydrophilic substances along the rising water front. This elution results in a hydrophilic concentration gradient increasing with sample height, as reflected in the non-linearity of the log (time) plots. For the as-received samples that were not plasma treated the maximum wicking height in the second run was more than halved as compared to the first run (from 82 to 38 mm). This indicates the high water solubility of the applied fiber surface additives. After washing of the samples the wicking was even lower, with an inhibition period of at least 15 seconds. For all samples treated for up to 10 seconds the maximum wicking height was close to zero. For the longer treatment durations the log (time) wicking curves behave

FIGURE 2. Summed wicking height (SWH) plots for as-received MA (white) and PET (black) samples, as function of treatment duration. Legend: first (E,F), second (‚,Œ) and third (䊐,■) wicking runs.

WICKING

OF

PRE-CLEANED SAMPLES

The behavior towards consecutive wishing tests of pre-cleaned MA and PET samples was similar to that of the as-received samples, although the wicking properties themselves had changed considerably (Figure 3). Water sorption of the non-plasma-treated sampleswas always minimal, as expected from the fact that a moderately

440

FIGURE 3. Averaged wicking curves for pre-cleaned MA samples as function of plasma treatment duration. The top and bottom graphs are for the first and third wicking runs, respectively. Legend: plasma treatment of 0 seconds (X), 5 seconds (E), 10 seconds ({), 15 seconds (‚), 30 seconds (䊐).

hydrophilic layer of surface additives was removed during the washing treatment, exposing a more hydrophobic fiber polymer surface. After the plasma treatments of varying duration the separation of the first-time wicking curves is clear for both fiber types and from the shortest wicking times onwards. The curves are linear with log(time), indicating a homogeneous fiber surface. Results for the second run of the MA samples were nearly identical to those for the first run, which is due to a better stability of the plasma treatment effect to water in comparison with an as-received surface (Figure 4). Such surface stability was not found in the pre-cleaned PET samples, for which second run wicking results were reduced considerably. The behavior of pre-cleaned PET samples was similar to results for first and second wicking runs for as-received samples; namely, reduced maximum wicking heights, an inhibition of about 30 seconds and a loss of curve parallelism (not shown). It is suggested that the plasma treatment causes the PET polymer surface to become partly water soluble, resulting in elution-induced inhomogeneities and loss of hydrophilic properties. The SWH plots show that for pre-cleaned samples even the shortest plasma treatment is effective in giving the hydrophobic fiber surface a considerable hydrophilic property. They also clearly indicate how the successive wicking and washing procedures have a limited influence on MA sample properties and a larger influence on PET sample properties. In view of the

TEXTILE RESEARCH JOURNAL

FIGURE 4. Summed wicking height (SWH) plots for pre-cleaned MA (white) and PET (black) samples, as function of treatment duration. Legend: first (E,F), second (‚,Œ) and third (䊐,■) wicking runs.

partial aliphatic and complete aromatic structures of PET and MA, respectively the difference in wicking behavior of pre-cleaned samples is possibly the consequence of the formation during the plasma treatment of low molecular weight products with a different solubility in water. Note that using SWH values for summarizing wicking data has the advantage over the maximum wicking height (MWH, for example, after 10 minutes wicking) of better representing the behavior of the total wetted textile fiber surface. As an example, samples showing or not showing an inhibition period could have the same MWH, although the SWH value will be lower for the sample showing inhibition.

Conclusions The purpose of this study was to assess–with the use of consecutive wicking tests–the influence of fiber surface additives on the effect of an oxygen plasma treatment, and on the stability of the effect towards a treatment in water. The presence of a layer of surface additives was shown to have a profound influence on the efficiency of the plasma treatment. This observation is expected to be as significant for atmospheric plasma processing as it is for plasma processing in vacuo, and valid for all textile fiber surfaces covered with watersoluble additives. However, it is obvious that not all plasma applications “suffer” from the presence of sur-

MAY 2005

441

face additives. Many treatments aim only at the improvement of the first-time wettability; in this case there is no problem if the treated surface additives are washed away during the first contact with water. For the textile structures treated in this study, to obtain a given wicking effect a clean fiber surface requires longer treatment durations than a surface covered with surface additives. This is a result of the difference in resistance towards plasma oxidation between the surface additives and the pure fiber polymer surface. The longer treatment duration required for a pure fiber surface is compensated by an improved resistance of the effect towards a treatment with water. In the SWH plots the wicking behavior of samples treated in a range of plasma conditions and post-plasma wet treatments can be evaluated, and this in a single graph. Still, the individual log(time) wicking curves remain the basis for information on the industrially important early wicking behavior, and on inhomogeneities at the fiber surface. Although wicking tests do not produce the surface morphological and (physico)chemical information obtained from Atomic Force Microscopy (AFM) and x-ray Photoelection Spectroscopy (XPS) and contact angle measurements, its value lies in the possibility to characterize a textile structure in its entirety. As such, the determination of wicking and related properties is a complementary tool for the characterization of textiles whose processing or application benefits from the fiber surface modification by a plasma.

7. 8.

9.

10.

11.

12.

13.

14.

15.

ACKNOWLEDGEMENTS This study was supported by BOF project TW11V/ 983473, Ghent University

16.

17.

Literature Cited 1. Belin, R.E., The Effect of Corona Treatment on the Cohesion between Fibres and on their Wettability, J. Text. Inst. 67, 249 –252 (1976). 2. Benerito, R. R., Ward, T. L., Soignet, D. M., and Hinojosa, O., Modifications of Cotton Cellulose Surfaces by Use of Radiofrequency Cold Plasmas and Characterisation of Surface Changes by ESCA, Textile Res. J. 51, 224 –232 (1981). 3. Ferrero, F., Wettability Measurements on Plasma-treated Synthetic Fabrics by Capillary Rise Method, Polymer Testing 22, 571–578 (2003). 4. Hesse, A., Thomas, H., and Ho¨cker, H., Zero-AOX Shrinkproofing Treatment of Wool Top and Fabric. Part I: Glow Discharge Treatment, Textile Res. J. 65, 355–361 (1995). 5. Hesse, A., Thomas, H., and Ho¨cker, H., Zero-AOX Shrinkproofing Treatment of Wool Top and Fabric. Part I: Collagen Resin Application, Textile Res. J. 65, 371–378 (1995). 6. Ivanova, V. P., Andreevskaja, G. D., Friedrich, J., and

18.

19.

20.

21.

Ga¨hde, J., Untersuchungen zum Einfluss einer Oberfla¨chenmodifizierung von Fasermaterialien im nichtthermischen Plasma einer Hochfrequenz-Entladung auf die Adha¨sionsfestigkeit zu Epoxidharz. Part 1: Messung der Adhesionsfestigkeit, Acta Polymerica 31, 752–756 (1980). Kreiselmeier, G., Faserstoffmodifizierung durch ionisierte Gase, Textiltechnik 34, 73–76 (1984). Lee, K. S., and Pavlath, P. E., Low Temperature Plasma Treatment of Wool. II. Effects of Variables on Shrinkage and Yarn Strength, J. Polymer Sci.-Polymer Chem. Ed. 12, 2087–2090 (1974). Lee, M., Ryu, J., Wakida, T., and Sato, Y., Effect of Low-temperature Plasma Treatment on Surface Characteristic and Dyeing Properties of Wool, Chemistry Express 7, 241–244 (1992). Masse, P., Cavrot, J. P., Franc¸ois, P., Lefebvre, J. M., and Escaig, B., Adhesion Improvement of High Modulus Polyethylene Fibers by Surface Plasma Treatment: evaluation of Pull-out Test, Polymer Composites 14, 247–251 (1994). Pitt, W. G., Lakenan, J. E., and Brent Strong, A., The Influence of Plasma Gas Species on the Adhesion of Thermoplastic to Organic Fibers, J. Appl. Polym. Sci. 48, 845– 856 (1993). Sarmadi, A. M., and Kwon, Y. A., Improved Water Repellency and Surface Dyeing of Polyester Fabrics by Plasma Treatment, Textile Chemist and Colorist 25, 33– 40 (1993). Sarmadi, A. M., Ying, T. H., and Denes, F., Surface Modification of Polypropylene Fabrics by Acrylonitrile Cold Plasma, Textile Res. J. 63, 697–705 (1993). Spence, P., Wadsworth, L., and Chihani, T., Plasma Surface Treatment for Improved Wettability and Rewettability, in “Proc. Inda-Tec97,” Cambridge, MA, 8 –10 September 1997. Stefecka, M., Rahel, J., and Cernak, M., Atmosphericpressure Plasma Treatment of UHMW PE Fibers, J. Mater. Sc. Letters 18, 2007–2008 (1999). Sung-Spitzl, H., Plasma Pre-treatment of Textiles for Improvement of Dyeing Processes, Int. Dyer May 20 –23 (2003). Verschuren, J., Cold Plasma Technology for Textile Products, taking into Account their Specific Properties, in “Proc. 1st RUG-FTW PhD Symposium,” Ghent University, 5 December 2000. Wakida, T., Lee, M., Sato, Y., Ogasawara, S., Ge, Y., and Niu, S., Dyeing Properties of Oxygen Low-temperature Plasma-treated Wool and Nylon 6 Fibres with Acid and Basic Dyes, J. Soc. Dyers & Colorists 112, 233–236 (1996). Wong, K. K., Tao, X. M., Yuen, C. W. M., and Yeung, K. W., Topographical Study of Low Temperature Plasmatreated Flax Fibres, Textile Res. J. 70, 886 – 893 (2000). Wong, K. K., Tao, X. M., Yuen, C. W. M., and Yeung, K. W., Wicking Properties of Linen Treated with Low Temperature Plasma, Textile Res. J. 71, 49 –56 (2001). Wong, W., Chan, K., Yeung, K. W., and Lau, K. S., Morphological Modification of Poly(ethylene terephthalate) Surface after Excimer Laser Irradiation, J. Text. Eng. 46, 32– 41 (2000). Manuscript received 2005; accepted 2005.