271

Textile Research Journal

Article

Study of Dyeing Behavior of Polyester Fibers with Disperse Dyes Abstract

Sorption isotherms of disperse dyes from water with small and big molecules on various polyester fibers were obtained at 130, 115, and 105°C. Isotherms were plotted using a large range of initial dye concentration. A remarkable deviation from Nernst behavior was observed for all plotted isotherms. Langmuir fitting was tested and appeared to be in good agreement with experimental curves. Distribution of disperse dyes between polyester substrate and water and the saturation solubility of these dyes in polyester fibers was compared by considering the crystallinity and surface area of the fibers.

Soufien Dhouib1 and Abdelaziz Lallam Laboratoire de Physique et Mécaniques Textiles de Mulhouse, ENSITM, 11 rue Alfred Werner, 68093, Mulhouse, Cedex, France

Faouzi Sakli Unité de Recherches Textiles de Ksar Hellal, Institut Supérieur des Etudes Technologiques de Ksar Hellal, Avenue Hadj Ali Soua, 5070, Ksar Hellal, Tunisia

Key words polyester, microfiber, dyeing, disperse dyes, sorption isotherms, Langmuir fitting, Nernst fitting, aggregation, crystallinity, surface area

The dyeing of polyester fibers has been widely studied. However, the modeling of the sorption isotherms of disperse dyes on polyester fibers is still the subject of discussion. In the present study the influence of fiber characteristics and dye size on sorption isotherms of disperse dyes on polyester fibers at high temperature was investigated. A survey of the literature showed that until the 1990s a linear distribution of Nernst type has always been used as a reference for sorption isotherms of disperse dyes on polyester fibers [1–4]. In 1995, Nakumara et al. [5] proposed a dual-mode model in which they combined a Langmuir component with the Nernst classical model for the sorption isotherms of two purified disperse dyes on polyester fibers. During the last few years, some other research has led to sorption isotherms of commercial disperse dyes on polyester fibers obeying the Langmuir law, but without a convincing explanation [6, 7]. The present study was performed to try to find a theoretical foundation to explain such a result.

Experimental The polyester filament yarns, 72dtex/88fil (partially oriented yarn, fineness of monofilament 1.95 dtex), 50 dtex/

Textile Research Journal Vol 76(4): 271–280 DOI: 10.1177/0040517506061243

22 fil (fully oriented yarn, 2.27 dtex), 40 dtex/44 fil (fully oriented yarn, 0.9 dtex) were manufactured by SETILA S.A. All yarns were transformed into knitted fabrics then treated in alkaline solution of 1 g/L non-ionic detergent and 2 g/L Na2CO3 at 80°C for 30 minutes with a 30 : 1 liquor ratio. Later, the fabrics were heat-set at 180°C for 30 seconds without any tension. The crystallinity, the diameter and surface area were determined for all fibers before any treatment (1) and after heat-setting (2). The crystallinity was obtained by measuring the densities with a density gradient column of carbon tetrachloride and petroleum ether. Equation (1) [8] shows the relation between the crystallinity rate (χ%) and the density (ρ) of polyester fibers. The surface area was estimated by the use of the BET equation of krypton gas adsorption on polyester samples [9] (Table 1).1 1.455 ( ρ – 1.335 ) χ ( % ) = 100 × ----------------------------------------0.120 × ρ

(1)

1 Corresponding author: Tel.: (00216) 73475900; fax: (00216) 73475163; e-mail: dhouib.soufien@laposte. net/soufien.dhouib@ isetkh.rnu.tn Also at: Unité de Recherches Textiles de Ksar Hellal, Institut Supérieur des Etudes Technologiques de Ksar Hellal, Avenue Hadj Ali Soua, 5070, Ksar Hellal, Tunisia

www.trj.sagepub.com © 2006 SAGE Publications

TRJ

272

Textile Research Journal 76(4)

Table 1 Fiber characteristics. Fineness (dtex) Treatment Fiber FOY 0.9 POY 1.95 FOY 2.27

Diameter (µm)

Shrinkage after heat-setting (%)

Crystallinity (%)

Surface area (m2/g)

(1)

(1)

(2)

(2)

(1)

(2)

(2)

0.9 dtex 1.95 dtex 2.27 dtex

9.3 13.8 14.7

9.3 21.3 14.8

1.9 58.7 3.21

38 10 35

51 39 51

0.33 0.14 0.18

It was not possible to measure the surface area at the dyeing temperature with the laboratory equipment available. The estimated values obtained allowed us to compare the surface area of the different fibers. All dyeings were carried out with a well-stirred dyebath using a Mathis laboratory-scale dyeing machine at a liquor ratio of 20 : 1 and pH 5. Two dyes were selected (Figure 1) according to their molecular weights, which were obtained from Ciba, and used in the dyeing process without further purification: dye 1, CI Disperse Yellow 42 (molecular weight = 369 g) and dye 2, CI Disperse orange 3 (molecular weight = 242 g). No dispersing agent was added to the dyebath during dyeing. After dyeing, the samples were soaked twice in boiling water for 5 minutes then transferred to a fresh pot of cool water. Finally, surface dye was removed by rinsing the

specimens in a cold acetone bath for 30 s and then blotted and dried. Dye on the polyester substrate was measured by colorimetric analysis after its extraction with dimethylformamide (DMF) using a Kumagawa extractor. The disperse dyes were purified by Soxhlet extraction with benzene for 10 h as described in the previous paper [10]. Calibration curves were prepared using solutions of the pure dyes. Experimental isotherm curves were plotted using a large range of initial dye concentrations. The quantity of commercial dye added to the dyebath was varied between 0.1 and 20% in order to be able to determine the isotherm curves with great precision. The quantity of pure dye on the commercial powders was about 49.5% for dye 1 and 31.8% for dye 2.

Results and Discussion

Figure 1 Chemical formula of dye 1 and dye 2.

It was clear that all the sorption isotherms of disperse dyes on all polyester fibers at dyeing temperatures 130, 115, and 105°C obeyed the Langmuir law with a correlation coefficient R > 0.99 as presented in Table 2. As shown in Table 2 and Figures 2–7, the greatest saturation concentration (Cs) was always obtained with the fiber POY 1.95 for both dyes and at all temperatures in spite of it having the lowest surface area after heat-setting. This result can be explained by its low crystallinity, which gave it more amorphous zones that were able to receive more dye molecules in comparison with the other fibers. Indeed, Table 1 shows that despite its very high shrinkage, the fiber POY 1.95 was still the most amorphous after heat-setting. However, its diameter became greater than that of the fiber FOY 2.27 and its specific area became the lowest of all. Moreover, it was noted that at all dyeing temperatures when sorption isotherms of dye 1 on FOY 2.27 and FOY 0.9 fibers reached the saturation concentration Cs, the dye concentration on the fiber POY 1.95 continued to increase with increasing dye concentration in the bath at equilibrium (Figures 2, 4 and 6). This fact indicates that the measured sorption isotherms cannot be adjusted to the Nernst model in which the fibers saturate with dyebath saturation [1, 2, 4, 11], as otherwise, we would need to accept that dye solubility in the

Study of Dyeing Behavior of Polyester Fibers With Disperse Dyes S. Dhouib et al.

273

Table 2 Values of Langmuir fitting parameters of dye sorption isotherms. Disperse Yellow 42

Disperse Orange 3

Parameters of Langmuir fitting

FOY 0.9

POY 1.95

FOY 2.27

FOY 0.9

POY 1.95

FOY 2.27

130°C

KL (L/g) Cs (g/kg) Cs (mol/kg) KLCs (L/kg) R

7.2 49.7 13.5 357 1.00

3.3 92.7 25.1 304 0.99

2.6 64.0 17.3 168 1.00

3.7 50.6 20.6 188 1.00

3.8 62.8 25.5 238 1.00

2.5 56.9 23.1 139 0.99

115°C

KL (L/g) Cs (g/kg) Cs (mol/kg) KLCs (L/kg) R

12.7 36.8 9.9 463 0.99

3.0 92.7 25.1 277 1.00

15.7 28.2 7.6 443 0.99

9.2 33.0 13.4 303 0.99

8.2 31.6 12.9 259 1.00

105°C

KL (L/g) Cs (g/kg) Cs (mol/kg) KLCs (L/kg) R

28.1 24.3 6.6 682 0.99

8.24 60.1 16.3 495 0.99

27.1 11.8 3.2 320 0.99

12.7 26.1 10.6 333 0.99

14.7 24.2 9.8 355 0.99

T (°C)

Langmuir equation: Cf = KLCsCb/(1 + KLCb), where Cf = dye concentration on the fibre; Cb = dye concentration in the bath; Cs = saturation value for Langmuir sorption; KL = Langmuir constant; R = correlation coefficient between experimental results and Langmuir sorption curve.

dyebath can take different concentration values, which is impossible. These results are very important because they confirm that the saturation of polyester fibers was not obtained after saturation of the dyebath but after saturation of fiber sites that are able to receive dye molecules. Such sites can be assimilated to porous regions or dislocations within amorphous zones in which the dye can be bound to the fiber with hydrogen bonding and Van Der Waals

Figure 2 Sorption isotherms of dye 1 on polyester fibers at 130°C: ◦ = FOY 0.9,
= POY 1.95,  = FOY 2.27.

forces. This hypothesis agrees very closely with the theory of Langmuir isotherms as found in our experiments. Thus, we can say that the real deviations from the ideal Nernst behavior that occur in the distribution of disperse dyes between polyester substrate and water seems to be due to two phenomena bound to the characteristics of the fiber and the dye behavior within the fiber. The first one is the reduction of the volume able to fix dye molecules when

TRJ

TRJ

274

Textile Research Journal 76(4)

Figure 3 Sorption isotherms of dye 2 on polyester fibers at 130°C: ◦ = FOY 0.9,
= POY 1.95,  = FOY 2.27.

Figure 4 Sorption isotherms of dye 1 on polyester fibers at 115°C: ◦ = FOY 0.9,
= POY 1.95,  = FOY 2.27.

the dye concentration on the fiber increases. The second one may be the aggregation of the dye inside the fiber that can obstruct some pores which were able to accept dye in the molecular state. This tendency to aggregation should be higher when dye size increases. Figure 8 represents a

dyeing model that we have established to explain such a phenomenon. In addition, we observed that sorption isotherms of dye 1 at 115°C on FOY 0.9 and FOY 2.27 fibers (Figure 4) present a point that approximates to the general form of

Study of Dyeing Behavior of Polyester Fibers With Disperse Dyes S. Dhouib et al.

275

Figure 5 Sorption isotherms of dye 2 on polyester fibers at 115°C: ◦ = FOY 0.9,
= POY 1.95,  = FOY 2.27.

Figure 6 Sorption isotherms of dye 1 on polyester fibers at 105°C: ◦ = FOY 0.9,
= POY 1.95,  = FOY 2.27.

the Langmuir adjustment curve when it approaches the dye saturation concentration; something that we did not detect with dye 2. In fact, we found that at 115 and 105°C the experimental curves fitted better to curves composed of two parts. The first one ranged from zero to this point and obeyed (as shown in Figures 9 and 10) the Langmuir

law or a dual-mode sorption model composed of both Nernst and Langmuir models, whereas the second one corresponded to a straight line of equation Cf = Cs. Such a phenomenon is related to the fact that at 115 and 105°C, the fiber is not sufficiently swollen and its pores are still tight. Thus, when the dye concentration Cf on the high crys-

TRJ

TRJ

276

Textile Research Journal 76(4)

Figure 7 Sorption isotherms of dye 2 on polyester fibers at 105°C: ◦ = FOY 0.9,  = FOY 2.27.

Figure 8 Dyeing model of polyester fiber dyed with high concentration of big (on the left) and small (on the right) dye size.

talline fibers (such as FOY 0.9 and FOY 2.27) approaches the saturation value, the large amount of dye adsorbed on the fiber accelerates the aggregation of dye molecules inside the fiber. So a multilayer adsorption is produced and the pores of the fiber are quickly plugged, the thing that prevents the dye from reaching the accessible interior regions (Figure 8) and so adsorption of the dye is abruptly stopped. The behavior of the fiber POY 1.95 did not appear to be affected by the reduction of the dyeing temperature to 115 or 105°C because of its development of amorphous zones.

With reference to earlier studies that have led to sorption isotherms for disperse dyes with Nernst behavior, we suggest that such a result may be due to the fact that most plotted isotherms have been measured at low dyeing temperatures, generally less than 100°C and very rarely up to 110°C. At such low dyeing temperatures, most previous studies have been made with small dye sizes (as for dye 2 herein) that can diffuse through the polyester substrate much more easily under such conditions. In this case, the decrease of the dyeing temperature displaces the dyeing equilibrium towards the fiber. Furthermore,

Study of Dyeing Behavior of Polyester Fibers With Disperse Dyes S. Dhouib et al.

277

Figure 9 Sorption isotherms of dye 1 on FOY0.9 (top) and FOY2.27 (bottom) fibers fitted to curves composed of two parts: the first part obeys the Langmuir law and the second one corresponds to a straight line of equation Cf = Cs:
= 105°C, ◦ = 115°C,  = 130°C.

dye saturation in the bath and the fiber become lower and, in addition the use of low dyeing temperatures leads to an earlier saturation of the fiber, thereby limiting the first part of the sorption isotherms to the weak concentrations that are close to the origin of the curve. At this point it should be noted that for low dye bath concentrations, the Langmuir curve can be easily assimilated to the ideal Henry behavior with a slope equal to KLCs. As the dye concentration of

sorption isotherms with disperse dyes at low temperatures are limited to low or middle concentrations, it was easy to fit such experimental curves to Nernst-type isotherms. On the other hand, if the dye molecule size was large, the use of low dyeing temperature would cause an early aggregation of the dye in the fibers that were slightly swollen and, as a result, the isotherms obtained seem to resemble those shown in Figures 9 and 10. Thus, the first part of the isotherm can

TRJ

TRJ

278

Textile Research Journal 76(4)

Figure 10 Sorption isotherms of dye 1 on FOY0.9 and FOY2.27 fibers fitted to curves composed of two parts: the first part obeys a dualmode sorption model composed of both Nernst and Langmuir models and the second one corresponds to a straight line of equation Cf = Cs:
= 105°C, ◦ = 115°C,  = 130°C.

also be fitted to the Nernst model with low error. The deviation from linearity for the highest concentrations that occurs with pure and commercial dyes is explained by various events such as the increasing amounts of dispersing agent included in the increasing amounts of commercial dye in the bath, errors in measured dye concentration, and the presence of certain associated disperse dyes on the aqueous solution [3].

Furthermore, it is easily seen in Figures 2 and 3 that at 130°C the microfiber FOY 0.9 was able to fix less dye than the classical fiber FOY 2.27, in spite of its greater surface area. Hence, the surface area is not well correlated to the dye adsorption ability of the fiber at this temperature; however, other parameters have a real influence on fiber accessibility such as the crystallinity and probably the geometry and the pore size of the fiber. As FOY 0.9 and FOY2.27 fib-

Study of Dyeing Behavior of Polyester Fibers With Disperse Dyes S. Dhouib et al. ers have close crystallinity after heat setting, this result can be explained by supposing that even though the total volume of pores of the classical fiber is lower than that of the microfiber sample, the pores of the classical fiber are larger. At high dye concentration, when the dye began to aggregate within the fibers, the classical fiber FOY 2.27 would fix a greater quantity of dye. Indeed, the dye aggregation could block the passage of other dye molecules to deeper sites if the pore diameters were small (Figure 8). This hypothesis is further justified when observing the values of Langmuir constants KL and KLCs that correlate with the slopes of the first parts of the isotherm curves. These constant values show that for low dye concentrations (when the aggregation phenomenon had not still occurred) the dye quantity adsorbed by the microfiber FOY 0.9 was greater than the classical fiber FOY 2.27 due to the greater surface area of the microfiber. At low temperatures (115 and 105°C), swelling of the microfibres (as FOY 0.9) is easier than that of the classical fibers (as FOY 2.27) due to the larger surface areas of the microfibers, and this gives them a greater adsorption area to the dye [12]. This explains why, at these temperatures, the ability of fiber FOY 0.9 to fix dye molecules became better in comparison with fiber FOY 2.27 and consequently the Cs values of fiber FOY 2.27 became the slowest when dyeing at 115 or 105°C whatever the dye used (Table 2). The fiber POY1.95 always preserved the highest saturation values due to its very large amorphous domain and this facilitated its swelling and increased the number of sites which were able to fix dye molecules. On the other hand, in comparison with those achieved with dye 1, the sorption isotherms of dye 2 on polyester fibers show some differences not in the form of curves but rather in the proportionality of values of Langmuir fitting constants. Indeed, with dye 2, the values of KL and KLCs became smaller. This indicates that the affinity of dye 1 to polyester fibers was better than that of dye 2. Furthermore, as shown in Table 2, at 105°C the mass and molar concentrations of Cs were higher when using dye 2. This means that, in this case, the fibers were slightly swollen and their pores were not dilated enough. Then, the small dye molecules (as dye 2) were fixed inside the fibers more easily and in a much more significant number. At 130°C, in a saturation state, the molar concentration [Cs (mol/kg)] of dye 2 remained the highest; but, the mass concentration [Cs (g/kg)] became slightly lower than those of dye 1. In fact, the swelling of the fiber increased with the increasing temperature; consequently, the fiber pores became larger by transformation of some pores of type A or B (Figure 11) into larger ones like type C, and so the fibers became able to receive many more dye molecules for both dye sizes. The number of molecules fixed was large as the dye size was small; however, the molar concentrations (Cs) of dye 2 were reduced to values lower than 1.53 times the molar concentration of dye 1; moreover, as the molar

279

Figure 11 Different types of fiber’s pores able or not to receive dye molecules: = dye 1, = dye 2.

mass of dye 1 was 1.53 times that of dye2, obviously a lower mass concentration would be obtained with dye 2. Finally, as shown in Table 2, for all the dyes and fibers, when the dyeing temperature decreased, the KL values increase distinctly whereas the Cs values decrease. This was due to two phenomena related to the reduction of dyeing temperature: first, there is a decrease of the solubility of disperse dyes in the bath that displaces the dyeing equilibrium to the fiber; second a reduction of the fiber accessibility has occurred due to the decrease of fiber swelling. Therefore, the fiber is saturated with lower dye concentrations.

Conclusion The sorption isotherms of disperse dyes on polyester fibers from water are well correlated to Langmuir law. The most important parameters for dye saturation on the fiber are temperature, dye size and volume of accessible domains able to receive dye molecules. This volume is very dependent on fiber crystallinity. Both fiber surface area, and dyeing temperature have a second-order effect. The temperature will have more influence when fiber accessibilty is poor.

Literature Cited 1. 2.

3.

Burkinshaw, S. M., “Chemical Principles of Synthetic Fibre Dyeing,” Blackie Chapman & Hall, London, 1st edn, 1995. Cegarra, J., Puento, P., and Valdeperas, J., “The Dyeing of Textile Materials – the Scientific Bases and the Techniques of Application,” G.B. Paravia & C. Torino, Italy, 1992. McDowell W. and Weingarten R., New Experimental Evidence about the Dyeing of Polyester Material with Disperse Dyes, JSDC, 85(12), 589–597 (1969).

TRJ

TRJ

280 4.

5.

6. 7.

Textile Research Journal 76(4) Peters, R. H., “Textile Chemistry, Volume III, The Physical Chemistry of Dyeing,” Elsevier Scientific Publishing Company, Amsterdam, 1975. Nakamura, T., Ohwaki, S., and Shibusawa, T., Dyeing properties of polyester microfibers, Textile. Res. J. 65(2), 113–118 (1995). Dieval, F., Thèse de doctorat en sciences pour l’ingénieur, U.H.A, Mulhouse, 1998. Park, R. H., Casetta, M., and Koncar, V., Diffusion of disperse dyes into microfibres and conventional polyester fibres, Coloration Technol. 118, 319–324 (2002).

8.

Aspland, J. R., A series on dyeing – chapter 7 – The application of non ionic dyes to fibers: nonionics and their sorption by man made fibers, Text. Chem. Col, 24(10), 36–40 (1992). 9. Viallier, P., Thèse, Université de Lyon, 1970. 10. Shukla, S. R., and Dhuri, S. S., A Practical application of the Kubelka–Munk Theory in Polyester Dyeing, Am. Dyestuff Report. 32–52, April, (1992). 11. Viallier, P., Comportement tinctorial des microfibres, l’origine des problèmes, L’industrie Textile 1293, 71–74, December (1997). 12. Dauber, Bunn, and Braun, Proc. Roy. Soc., A226, (1954).