Dyeing Property and Adsorption Kinetics of Reactive Dyes for Cotton Textiles in Salt-Free Non-Aqueous Dyeing Systems

In recent years, new concepts in textile dyeing technology have been investigated which aim to decrease the use of chemicals and the emission of water. In this work, dyeing of cotton textiles with reactive dyes has been investigated in a silicone non-aqueous dyeing system. Compared with conventional aqueous dyeing, almost 100% of reactive dyes can be adsorbed on cotton textiles without using any salts in non-aqueous dyeing systems, and the fixation of dye is also higher (80%~90% for non-aqueous dyeing vs. 40%~50% for traditional dyeing). The pseudo-second-order kinetic model can best describe the adsorption and equilibrium of reactive dyes in the non-aqueous dyeing systems as well as in the traditional water dyeing system. In the non-aqueous dyeing systems, the adsorption equilibrium of reactive dyes can be reached quickly. Particularly in the siloxane non-aqueous dyeing system, the adsorption equilibrium time of reactive dye is only 5–10 min at 25 °C, whereas more time is needed at 60 °C in the water dyeing system. The surface tension of non-aqueous media influences the adsorption rate of dye. The lower the surface tension, the faster the adsorption rate of reactive dye, and the higher the final uptake of dye. As a result, non-aqueous dyeing technology provides an innovative approach to increase dye uptake under a low dyeing temperature, in addition to making large water savings.


Introduction
In recent years, tremendous efforts have been focused on saving water and energy in the textile industry due to environmental pressure [1][2][3][4]. Some new dyeing methods [5][6][7][8] have been developed to reduce the use of salts and to save water during the dyeing process; for example, supercritical fluid dyeing, solvent dyeing, reverse micelle dyeing, non-aqueous dyeing, ionic liquid dyeing etc. Unfortunately, natural fibers cannot be effectively dyed with conventional water-soluble dyes in normal supercritical fluid [4,5]. Solvent dyeing and reverse micelle dyeing are often carried out using hydrocarbon solvents which are not environmentally friendly, such as hexane, cyclohexane and of non-aqueous dyeing technology is that little water is needed because dissolving the dye and other 48 chemical agents, and fiber swelling require only a small amount of aqueous solution [14][15][16].

49
Furthermore, all the aqueous solution can be completely absorbed by the cotton textile without using 50 any accelerating salts during the dyeing process. As for the re-usability of non-aqueous media, after 51 a short period of static separation non-aqueous media can be directly used for the next dyeing. Non-52 aqueous media which is adsorbed on the surface of cotton textile can be washed down using 53 surfactants and reused by flotation.

54
In our previous investigations [15,[17][18][19], a non-aqueous dyeing system was favorably prepared 55 using siloxane non-aqueous medium as the dyeing medium. Siloxane non-aqueous medium is a clear, 56 odorless, colorless, and non-oily cyclic siloxane fluid, which is widely used in consumer and 57 industrial applications. Research has demonstrated that siloxane non-aqueous medium is safe to both 58 human health and the environment [20][21][22]. In recent years, siloxane non-aqueous medium has been 59 applied widely in dry cleaning as a new medium [23]. Now, we have used this non-aqueous medium 60 to prepare a reactive dye/siloxane emulsion dyeing system for dyeing cotton textiles. As shown in 61 Figure 1, with the aid of a surfactant and co-surfactant, the dye solution can be evenly emulsified in 62 siloxane non-aqueous medium. Since reactive dye is totally incompatible with siloxane, there is a 63 strong affinity between the fiber and the dye, with the result that nearly 100% of reactive dye can 64 diffuse to the surface of the cotton textile under mechanical force. Therefore, the final uptake of    To study the influence of non-aqueous media surface tension on the adsorption of reactive dye, vinyl sulphone reactive dye (blue 19) and a combination of cyanuric chloride and vinyl sulphone reactive dye (red 195) were selected. The influences of dyeing media and surface tension of non-aqueous media on the adsorption kinetics of reactive dye were investigated with a model on the adsorption kinetics of reactive dye. In addition, the influence of surface tension on reactive dyeing level property was systematically discussed.

91
The molecular structures of these dyes are shown in Figure 2.

Determination of Surface Tension
A Krüss GmbH DAS20 using the captive drop method was used to measure the surface tension of different non-aqueous media. In the testing conditions, the needle diameter was 1.820 mm and flow rate was 60.2 mm/min.

Cotton Textile Dyeing
2.3.1. Non-Aqueous System 5.0 g cotton fabric was dyed with reactive dye (2%, o.w.f.) in 100 mL (mililitro) non-aqueous medium (liquor ratio = 20:1) using a dyeing machine (Sanjing Technology Company, Hangzhou, China). First, 2% (o.w.f.) of reactive dyewas added to 130% (o.w.f) of water. After stirring for 5 min, 7.5% (o.w.f) of sodium carbonate was added to reactive dye solution and stirred for another 5 min. Finally, the dye solution obtained in the previous step was added into 100 mL of non-aqueous medium. After adding the cotton fabric, dyeing began at a lower temperature (25 • C) for 15 min, and then the temperature was increased to 70 • C for 30 min, with a heating rate of 2 • C/min. After dyeing, the dyed fabric was washed in a soap solution containing 2.0 g/L sodium carbonate and 2.0 g/L standard detergent for 10 min, rinsed again with warm (50 • C) water and cold water, and finally dried at ambient temperature.

Aqueous Dyeing
For comparison with the conventional water-base dyeing, the same cotton fabric (5 g) was dyed with reactive dye (2%, o.w.f) in water base at a liquor ratio of 20:1. The amounts of NaCl and Na 2 CO 3 were 40 and 15 g/L, respectively. First, reactive dye and the cotton fabric were added into 100 mL of water at room temperature, and then the temperature was increased to 60 • C. After 10 min, half of the NaCl was added into the dye bath, and the rest of the NaCl was added in the next 10 min. After adsorption (30 min), the temperature was increased to 70 • C at 2 • C/min, then the Na 2 CO 3 was added into the dye bath and left for 30 min. After fixation, the dyed fabric samples were soaped and rinsed.

Determination of Adsorption Rate
The dyeing rate was monitored using the dye concentration. Small samples (1 mL) were taken at different times during the dyeing process. The concentration of dye was determined with a UV-vis spectrometer (Lambda 35, Perkin Elmer, Waltham, MA, USA).
Prior to dyeing, standard solutions of the two dyes were made and the absorbance was measured from 380~680nm wavelengths; therefore, a plot of absorbance versus wavelength could be obtained. Maximum adsorption wavelength of the dye was determined with a UV-vis spectrometer (C.I. Reactive Red 195 in 550 nm; C.I. Reactive Blue 19 in 596 nm). Calibration curves of absorbance vs. dye concentration (Beer-Lambert law) were performed separately for the two dyes. The calibration plot revealed a linear relationship in the range of dye concentration from 0 to 0.19 g/L. Therefore, it was necessary to dilute the sample with a higher concentration.

Color Depth of Dyed Fabric
The color strength of dyed fabric samples were measured with a Datacolor SF600X spectrophotometer (Datacolor, Lawrenceville, NJ, USA), then the K/S values were evaluated at λ max using the Kubelka-Munk equation: where R is the reflectance of the dyed fabric sample at the λ max absorption and K and S are the absorption and scattering coefficients, respectively.

Dye Uptake of Reactive Dye
The final uptake of reactive dye was calculated by the following equation. The amount of dye in the dyeing residues solution was determined using a UV-vis spectrometer.
where sorption refers to the dye uptake, C 0 and C 1 refer to the concentration (g/L) of initial dye solution and dyeing residues, respectively; and V 0 and V 1 refer to the volume (mL) of initial dye solution and dyeing residues, respectively.

Fixation Rate of Reactive Dye
Consideration of the percentage of fiber-reactive dyes is necessary to evaluate the dye fixation during the non-aqueous dyeing systems and in the traditional water dyeing system. The fixation ratio of reactive dye was calculated based on the following equation: where T refers to the fixation rate of reactive dye, and C 0 , C 1 , C 2 refer to the concentration (g/L) of the initial dye solution, dyeing residues and soaping solution, respectively. V 0 , V 1 , and V 2 refer to the volume (mL) of the initial dye solution, dyeing residues and soaping solution, respectively.

Level Dyeing Property
Twelve points on the fabric surface were randomly selected, and the K/S values of these points were tested with a Datacolor SF600X spectrophotometer (Datacolor, Lawrenceville, NJ, USA). The level dyeing property was calculated using the following equation, where the lower the value of S γ (λ), the better the level dyeing property:

Comparisons between the Dyeing Property of Reactive Dye in Different Non-Aqueous Dyeing Systems and a Traditional Water Dyeing System
Four different non-aqueous dyeing media were selected to study their influence on reactive dyeing for cotton fabric, and the color depths of the dyed fabric are plotted in Figure 3. Dyeing in a water base was also included for comparison. In the traditional water dyeing system, the color depth of cotton fabrics (K/S value) dyed with reactive blue 19 and red 195 were 12 and 13, respectively; while in non-aqueous media, the color depth of dyed cotton fabric samples dyed with reactive blue 19 and reactive red 19 were above 15 and 21, respectively.
where sorption refers to the dye uptake, C0 and C1 refer to the concentration (g/L) of initial dye 147 solution and dyeing residues, respectively; and V0 and V1 refer to the volume (mL) of initial dye 148 solution and dyeing residues, respectively.  Consideration of the percentage of fiber-reactive dyes is necessary to evaluate the dye fixation 151 during the non-aqueous dyeing systems and in the traditional water dyeing system. The fixation ratio 152 of reactive dye was calculated based on the following equation: where T refers to the fixation rate of reactive dye, and C0, C1, C2 refer to the concentration (g/L) of the 154 initial dye solution, dyeing residues and soaping solution, respectively. V0, V1, and V2 refer to the 155 volume (mL) of the initial dye solution, dyeing residues and soaping solution, respectively. 161   Comparing the different dyes, the color depth of the cotton fabric dyed with reactive 195 was deeper than that of the fabric dyed with reactive blue 19. This was mainly because the reactive red 195 molecular structure contains two different active groups; namely, the homogeneous chlorotriazine and vinyl sulfone active groups, with the result that the reaction between fiber and dye was higher than that of a single functional group (blue 19) [24][25][26]. Furthermore, the molar extinction coefficients of dye might be a reasonable guide to the color depth of dyed fabric [27]. Reactive red 195 may absorb more light and reflect less light. Therefore, the color depth of fabric after dyeing with reactive red 195 was significantly deeper than that of cotton fabric dyed with reactive blue 19.

Results and Discussion
As shown in Figure 4, the uptake of reactive dye in non-aqueous media was above 95%, indicating that almost all of the dye could diffuse to the cotton fabric without adding any electrolyte at 25 • C during dyeing. However, in the conventional aqueous dyeing system, the final uptake of these two reactive dyes was only 70%~75% when using an accelerating agent (sodium chloride) at 60 • C. Considering removal of these salts is a major challenge for the treatment of dyeing wastewater [25]; however, there is a little wastewater after dyeing in the non-aqueous media dyeing systems. Therefore, non-aqueous dyeing systems are green and recyclable dyeing systems. Comparing the different dyes, the color depth of the cotton fabric dyed with reactive 195 was 174 deeper than that of the fabric dyed with reactive blue 19. This was mainly because the reactive red 175 195 molecular structure contains two different active groups; namely, the homogeneous 176 chlorotriazine and vinyl sulfone active groups, with the result that the reaction between fiber and dye 177 was higher than that of a single functional group (blue 19) [24][25][26]. Furthermore, the molar extinction 178 coefficients of dye might be a reasonable guide to the color depth of dyed fabric [27]. Reactive red 179 195 may absorb more light and reflect less light. Therefore, the color depth of fabric after dyeing with 180 reactive red 195 was significantly deeper than that of cotton fabric dyed with reactive blue 19.

181
As shown in Figure 4, the uptake of reactive dye in non-aqueous media was above 95%,

182
indicating that almost all of the dye could diffuse to the cotton fabric without adding any electrolyte 183 at 25 °C during dyeing. However, in the conventional aqueous dyeing system, the final uptake of 184 these two reactive dyes was only 70%~75% when using an accelerating agent (sodium chloride) at 185 60 °C. Considering removal of these salts is a major challenge for the treatment of dyeing wastewater 186 [25]; however, there is a little wastewater after dyeing in the non-aqueous media dyeing systems.

187
Therefore, non-aqueous dyeing systems are green and recyclable dyeing systems.

191
The effect of the non-aqueous media on the fixation of the reactive dye was also investigated, as 192 shown in Figure 4.The fixation rates for the non-aqueous dyeing systems were higher than for the 193 traditional process (80%~90% for siloxane non-aqueous dyeing system vs. 40%~50% for traditional).

194
Fixation is also affected by the sorption of dye, dyeing time, etc. [10,28]. The apparent sorption of 195 reactive dye was over 95% in non-aqueous dyeing systems, indicating that a lot of the dyes could 196 react with cotton fabrics. Furthermore, the dyeing temperature is lower (25 °C in non-aqueous media 197 vs. 60 °C in water base), and the hydrolysis rate of reactive dye in non-aqueous dyeing systems is 198 slower than that in the traditional water base [15,16]. As a result, a lot of reactive dyes can bond with 199 cotton textiles, and a higher fixation rate can be obtained in the non-aqueous dyeing systems [29,30].

200
As can be seen, this dyeing technology can solve some of major problems of a traditional water bath, 201 such as difficult wastewater treatment due to the high residue amounts of dye and electrolytes, low 202 dye efficiency, etc.

204
The degree of dye adsorption (qt) can be calculated using the following equation:  The effect of the non-aqueous media on the fixation of the reactive dye was also investigated, as shown in Figure 4. The fixation rates for the non-aqueous dyeing systems were higher than for the traditional process (80%~90% for siloxane non-aqueous dyeing system vs. 40%~50% for traditional). Fixation is also affected by the sorption of dye, dyeing time, etc. [10,28]. The apparent sorption of reactive dye was over 95% in non-aqueous dyeing systems, indicating that a lot of the dyes could react with cotton fabrics. Furthermore, the dyeing temperature is lower (25 • C in non-aqueous media vs. 60 • C in water base), and the hydrolysis rate of reactive dye in non-aqueous dyeing systems is slower than that in the traditional water base [15,16]. As a result, a lot of reactive dyes can bond with cotton textiles, and a higher fixation rate can be obtained in the non-aqueous dyeing systems [29,30]. As can be seen, this dyeing technology can solve some of major problems of a traditional water bath, such as difficult wastewater treatment due to the high residue amounts of dye and electrolytes, low dye efficiency, etc.

Dye Adsorption in Different Non-Aqueous Dyeing Systems
The degree of dye adsorption (q t ) can be calculated using the following equation: where q t (mg/g) refers to the mass of dye adsorbed on cotton fabric at given times during the dyeing process, C 0 refers to the initial dye concentration (g/L) in the dye bath, C t refers to the residue dye concentration (g/L) at given times, V is the volume of dye solution (mL), and W is the mass of cotton fabric (g). As shown in Figure 5, the fastest equilibrium was observed at 25 • C in the siloxane non-aqueous dyeing system. In other non-aqueous dyeing systems, adsorption equilibrium was established within 20-30 min. This is mainly because the surface tension of non-aqueous media is low (18.13 mN/m), which makes them highly hydrophobic, and reactive dye solution is easily adsorbed on the cotton fabric surface under mechanical force [31,32]. Therefore, salt-free adsorption could be realized by non-aqueous media dyeing.
Polymers 2017, 9, x FOR PEER REVIEW 7 of 16 concentration (g/L) at given times, V is the volume of dye solution (mL), and W is the mass of cotton 208 fabric (g).
which makes them highly hydrophobic, and reactive dye solution is easily adsorbed on the cotton 213 fabric surface under mechanical force [31,32]. Therefore, salt-free adsorption could be realized by

218
In addition to the faster equilibrium, the highest uptake of dye (20 mg/g) was achieved after 219 dyeing for 10 min in the siloxane non-aqueous dyeing system. For other non-aqueous dyeing media, 220 the uptake of dye was also above 94%, which was significantly higher than that of the traditional   where qe (mg/g) and qt (mg/g) denote the masses of dye adsorbed on cotton fabric at equilibrium sate 232 and at given times t (min) respectively, andk1 refers to the rate constant of the pseudo-first-order 233 kinetic model.

236
The fitted line plots of equations in non-aqueous dyeing systems and conventional water are 237 shown in Figure 6. Table 2 shows the values of R 2 for cotton fabric dyed with red 195 and blue 19.

238
Both non-aqueous dyeing systems and the aqueous system showed a low correlation coefficient,

239
indicating that the pseudo-first-order kinetic model did not fit well for the adsorption or reactive dye  In addition to the faster equilibrium, the highest uptake of dye (20 mg/g) was achieved after dyeing for 10 min in the siloxane non-aqueous dyeing system. For other non-aqueous dyeing media, the uptake of dye was also above 94%, which was significantly higher than that of the traditional water base. The distribution of reactive dye solution between the dyeing bath and the cotton fabric depends on its relative solubility in the dyeing media and fabric. Since reactive dye solution is totally incompatible with non-aqueous media, there is a strong affinity between fiber and dye, with the result that nearly 100% of reactive dye can diffuse to the surface of cotton textile under mechanical force [18,33].

Fitting of Pseudo-First-Order Kinetic Model
In order to assess the adsorption performance of reactive dye in different non-aqueous dyeing systems, the adsorption dynamics of the reactive dye were evaluated using Lagergren equations [34][35][36][37].
ln(q e − q t ) = ln q e − k 1 t where q e (mg/g) and q t (mg/g) denote the masses of dye adsorbed on cotton fabric at equilibrium sate and at given times t (min) respectively, and k 1 refers to the rate constant of the pseudo-first-order kinetic model. According to Equation (6), R 2 (correlation coefficient) was calculated using plots of ln(q e − q t ) versus adsorption time (t).
The fitted line plots of equations in non-aqueous dyeing systems and conventional water are shown in Figure 6. Table 2 shows the values of R 2 for cotton fabric dyed with red 195 and blue 19. Both non-aqueous dyeing systems and the aqueous system showed a low correlation coefficient, indicating that the pseudo-first-order kinetic model did not fit well for the adsorption or reactive dye in different non-aqueous dyeing systems.
where k2 (g/mg·min) denotes the adsorption rate constant of reactive dye, and qe and qt denote the 248 masses of dye which react with cotton fabric samples at equilibrium state and at given times t, 249 respectively.

250
As shown in Equation (7), R 2 was calculated using plots of t/qt versus adsorption time (t).We can 251 use the slopes and intercepts of these plots to calculate the adsorption rate of reactive dye (k2) and the 252 adsorption equilibrium (qe), as shown in Equation (8): where tan α and b denote the slope and intercept of the linear equation, respectively.

254
The dyeing rate can also be expressed by the half-dyeing time (t0.5) which is the time required 255 for half of the dye to reach equilibrium adsorption amount. The t0.5 was calculated using Equation (

Fitting of Pseudo-Second-Order Kinetic Model
The pseudo-second-order kinetic model is expressed as the following equation [38][39][40][41]: where k 2 (g/mg·min) denotes the adsorption rate constant of reactive dye, and q e and q t denote the masses of dye which react with cotton fabric samples at equilibrium state and at given times t, respectively. As shown in Equation (7), R 2 was calculated using plots of t/q t versus adsorption time (t).We can use the slopes and intercepts of these plots to calculate the adsorption rate of reactive dye (k 2 ) and the adsorption equilibrium (q e ), as shown in Equation (8): where tan α and b denote the slope and intercept of the linear equation, respectively. The dyeing rate can also be expressed by the half-dyeing time (t 0.5 ) which is the time required for half of the dye to reach equilibrium adsorption amount. The t 0.5 was calculated using Equation (9): Figure 7 shows the fitting curves of the pseudo-second-order model and Table 3 lists the results of rate constant studies for different non-aqueous dyeing systems and the conventional water dyeing system with the pseudo-second-order model. Both non-aqueous dyeing systems and the aqueous system showed a high correlation coefficient value (>0.991). Furthermore, the equilibrium adsorption capacities, q e,cal , fit well with the experimental adsorption capacities, q e,exp , in the non-aqueous dyeing systems and the traditional aqueous system. These results suggest that the adsorption kinetics of these two reactive dyes on a cotton fabric surface can be described by the pseudo-second-order adsorption mechanism in non-aqueous dyeing systems and a traditional aqueous system. Polymers 2017, 9, x FOR PEER REVIEW 9 of 16 these two reactive dyes on a cotton fabric surface can be described by the pseudo-second-order 263 adsorption mechanism in non-aqueous dyeing systems and a traditional aqueous system.

267
The results in Table 3 also show k2, t0.5, as a main factor in the adsorption rate. For the adsorption 268 rate, the rate constant in non-aqueous dyeing systems was significantly faster than that of the dyeing time, t0.5, was significantly lower in non-aqueous dyeing systems than in the aqueous dyeing 272 system. These results indicated that the adsorption of reactive dye in non-aqueous dyeing systems 273 was faster than in the conventional water system.  The results in Table 3 also show k 2 , t 0.5 , as a main factor in the adsorption rate. For the adsorption rate, the rate constant in non-aqueous dyeing systems was significantly faster than that of the traditional water bath. Especially in the siloxane non-aqueous dyeing system, the rate constants of these two reactive dyes were 5~6 times those of the conventional aqueous bath. In addition, the half-dyeing time, t 0.5 , was significantly lower in non-aqueous dyeing systems than in the aqueous dyeing system. These results indicated that the adsorption of reactive dye in non-aqueous dyeing systems was faster than in the conventional water system. Based on the amount of reactive dye (2%, o.w.f) added to the dyeing system, if all of the dye was adsorbed on the cotton fabric, the adsorption capacity was 20 mg/g. For reactive red 195, the equilibrium adsorption, q e,exp , at 10 min reached above 18.81 mg/g in the non-aqueous dyeing systems, indicating that at least 94.05% of dye had been adsorbed on the cotton fabric. However, only 13.75 mg/g (68.75%) of dye was adsorbed in the water bath after dyeing for 70 min. We also observed a similar phenomenon in the adsorption of reactive blue 19. These results suggest that the equilibrium time of reactive dye in non-aqueous dyeing systems was shorter than in the water bath, and almost all the dye could be adsorbed on the fabric in the non-aqueous dyeing systems. For different dyes, the rate constant value of red 195 was lower than that of blue 19 under the same dyeing conditions. The reason may be that the molecular structure of red 195 is complex and its molecular weight is higher than that of blue 19 [42][43][44], so the adsorption rate of the dye is influenced by the molecular structure.

Effect of Non-Aqueous Media on the Dyeing Level Property
Compared with a conventional aqueous bath, reactive dye has a faster adsorption rate in non-aqueous dyeing systems, which might influence their dyeing property. As shown in Figures 8 and 9, for the dyeing property of reactive red 195 and blue 19, their values of S γ (λ) were higher in non-aqueous media than in the traditional water bath, but were less than 1, indicating that reactive dye can obtain a good level dyeing property by mechanical force during the dyeing process.
time of reactive dye in non-aqueous dyeing systems was shorter than in the water bath, and almost 283 all the dye could be adsorbed on the fabric in the non-aqueous dyeing systems. For different dyes, 284 the rate constant value of red 195 was lower than that of blue 19 under the same dyeing conditions.

285
The reason may be that the molecular structure of red 195 is complex and its molecular weight is 286 higher than that of blue 19 [42][43][44], so the adsorption rate of the dye is influenced by the molecular 287 structure.     all the dye could be adsorbed on the fabric in the non-aqueous dyeing systems. For different dyes, 284 the rate constant value of red 195 was lower than that of blue 19 under the same dyeing conditions.

285
The reason may be that the molecular structure of red 195 is complex and its molecular weight is 286 higher than that of blue 19 [42][43][44], so the adsorption rate of the dye is influenced by the molecular 287 structure.

289
Compared with a conventional aqueous bath, reactive dye has a faster adsorption rate in non-290 aqueous dyeing systems, which might influence their dyeing property. As shown in Figures 8 and 9,

291
for the dyeing property of reactive red 195 and blue 19, their values of Sγ(λ) were higher in non-292 aqueous media than in the traditional water bath, but were less than 1, indicating that reactive dye 293 can obtain a good level dyeing property by mechanical force during the dyeing process.

325
Thus, the above results reveal that, for dyeing with reactive dye in non-aqueous media, the 326 higher surface tension of non-aqueous media is a possible reason for the longer adsorption time. This 327 is due to the insolubility of reactive dye in non-aqueous media, and the strong affinity of reactive dye 328 to cotton fabric [45,46]. Therefore, reactive dye can quickly diffuse to the fabric surface and complete 329 the adsorption process [47][48][49]. When the surface tension of non-aqueous media is increased, the 330 repulsive force between the reactive dye and non-aqueous media may decrease, which will influence 331 the adsorption rate of the dye. As shown in Figure 10, the surface tension of non-aqueous media increased with the increase of M-oil or paraffin. For example, when the mass ratio of siloxane to M-oil was 4:1, 1:1, and 1:4, the surface tensions of the mixed media were 22.83, 25.73, and 28.16 mN/m respectively. To study the influence of surface tension on reactive dye adsorption in non-aqueous dyeing systems, several mass ratios of siloxane to M-oil (4:1, 1:1, 1:4) were chosen in this investigation.

Effect of Surface Tension on Reactive Dye Adsorption
As shown in Figure 11, in different surface tension non-aqueous dyeing systems, the uptake of reactive dyes was almost the same and was close to 100%, indicating that the surface tension of non-aqueous media had little influence on the final adsorption capacity of reactive dye. In the siloxane non-aqueous dyeing system, the reactive dye completes the adsorption process in 5-10 min. If M-oil was added to siloxane to increase the media surface tension, the adsorption rate of the reactive dye was decreased which meant it needed 10-15 min to reach the adsorption equilibrium. Moreover, reactive dye took more time (close to 20 min) to accomplish the adsorption process in M-oil non-aqueous dyeing media, in which the surface tension was 31.23 mN/m.

308
As shown in Figure 10, the surface tension of non-aqueous media increased with the increase of

325
Thus, the above results reveal that, for dyeing with reactive dye in non-aqueous media, the 326 higher surface tension of non-aqueous media is a possible reason for the longer adsorption time. This 327 is due to the insolubility of reactive dye in non-aqueous media, and the strong affinity of reactive dye 328 to cotton fabric [45,46]. Therefore, reactive dye can quickly diffuse to the fabric surface and complete 329 the adsorption process [47][48][49]. When the surface tension of non-aqueous media is increased, the 330 repulsive force between the reactive dye and non-aqueous media may decrease, which will influence 331 the adsorption rate of the dye. Thus, the above results reveal that, for dyeing with reactive dye in non-aqueous media, the higher surface tension of non-aqueous media is a possible reason for the longer adsorption time. This is due to the insolubility of reactive dye in non-aqueous media, and the strong affinity of reactive dye to cotton fabric [45,46]. Therefore, reactive dye can quickly diffuse to the fabric surface and complete the adsorption process [47][48][49]. When the surface tension of non-aqueous media is increased, the repulsive force between the reactive dye and non-aqueous media may decrease, which will influence the adsorption rate of the dye.

Fitting of Pseudo-First-Order Kinetic Model
It can be seen from the results obtained in the above section, the salts can be eliminated and reactive dye can quickly diffuse to the cotton fabric surface in different surface tension non-aqueous dyeing processes. In order to thoroughly understand the effect of non-aqueous surface tension on the adsorption kinetics, the fitted line plots of equations for different surface tension non-aqueous dyeing media provide the values of R 2 . As shown in Figure 12 and Table 4, the correlation coefficients for different surface tension non-aqueous dyeing systems were above 0.744. However, the linearity was not good, indicating that the pseudo-first-order kinetic model did not fit well to the adsorption or reactive dye in different non-aqueous dyeing systems.

333
It can be seen from the results obtained in the above section, the salts can be eliminated and 334 reactive dye can quickly diffuse to the cotton fabric surface in different surface tension non-aqueous 335 dyeing processes. In order to thoroughly understand the effect of non-aqueous surface tension on the 336 adsorption kinetics, the fitted line plots of equations for different surface tension non-aqueous dyeing 337 media provide the values of R 2 . As shown in Figure 12 and Table 4, the correlation coefficients for 338 different surface tension non-aqueous dyeing systems were above 0.744. However, the linearity was  From the data in Table 5 and as shown in Figure 13, the plots provide higher correlation   From the data in Table 5 and as shown in Figure 13, the plots provide higher correlation coefficients of more than 0.998, and excellent linearity for different surface tension non-aqueous dyeing systems. Moreover, the equilibrium adsorption capacities, q e,cal , fit well with the experimental adsorption capacities, q e,exp , in non-aqueous dyeing systems and in the traditional aqueous system. Therefore, the pseudo-second-order adsorption kinetic model was suitable to describe the adsorption kinetics of reactive dyes on cotton fabrics in different surface tension non-aqueous dyeing systems [50][51][52].    For the adsorption rate, the rate constant of reactive dye in lower surface tension non-aqueous dyeing systems was significantly faster than in higher surface tension systems. For example, the adsorption rate of reactive red 195 decreased from 3.18×10 −2 to 2.11×10 −2 /(g/mg·min) when the non-aqueous surface tension increased from 19.74 to 28.16 mN/m. In addition, the half-dyeing time, t 0.5, was significantly lower in lower surface tension non-aqueous dyeing systems. These results indicated that the lower the surface tension of non-aqueous media, the faster was the adsorption rate of reactive dye, as well as the higher the final uptake of dye.

Conclusions
Reactive dyeing for cotton fabric and the adsorption kinetics of dye in non-aqueous dyeing systems, and a traditional water base, were investigated in this study. Salts were eliminated throughout the dyeing process because reactive dye was completely non-miscible with non-aqueous media, but had strong affinity to cotton fabric. Compared with reactive dyeing in a traditional water base, a higher (>94%) uptake of dye rendered the entire media recyclable. Moreover, the fixation of reactive dye was also higher (80%~90% for non-aqueous dyeing vs. 40%~50% for traditional dyeing), resulting in cotton fabric achieving a deeper shade after dyeing. The pseudo-second-order kinetic model can adequately describe the adsorption and equilibrium of reactive dyes in non-aqueous dyeing systems as well as in a traditional water dyeing system. In non-aqueous dyeing systems, the adsorption rate of reactive dye was significantly faster than in a traditional water bath. In the siloxane non-aqueous dyeing system, the adsorption equilibrium time of reactive dye was only 5-10 min at 25 • C, whereas it needed more time at 60 • C in the water dyeing system. The surface tension of non-aqueous media affected the adsorption rate of the dye. The lower the surface tension of non-aqueous media, the faster was the adsorption rate of reactive dye, and the higher the final uptake of dye. These favorable results have implications for reducing the environmental pressure from reactive dyeing for cotton textiles in a source-control manner through being salts-free, with higher dye uptake, lower dyeing temperature, and large water savings.
Author Contributions: Y.G. and L.Z. designed the experiments; J.W. and L.P. performed the experiments and analyzed the data; L.P. and X.G. wrote the paper; L.P. and H.D. revised the paper. All authors discussed the results and improved the final text of the paper.