Abstract
Introduction
In household laundry, fabric softeners are used to add softness to the fabric. The main component of fabric softeners is a cationic surfactant, comprising hydrophilic and lipophilic groups compatible with water and oil, respectively. Cationic surfactants are those whose hydrophilic groups ionize to cations when dissolved in water. In general, cationic surfactants are strongly adsorbed on a negatively charged solid surface (Figure 1). In particular, when adhered to the surface of a fiber, the cationic surfactant imparts softness and flexibility to the fiber.1,2 Previous works have reported the mechanism whereby fabric softeners give flexibility to the fabric by investigating the effect fabric softener has on the mechanical properties and feel of the fabric.3–6

Cationic surfactant components (a) and alignment on a fiber (b). 13
The conventional explanation for the softening mechanism of fabric softener is a reduction of the fiber-to-fiber friction on the basis of the hypothesis. 1 The changes in friction of fibers and yarns by softeners have been investigated by many researchers. Ollofsson and Gralén 7 and Röder 8 investigated the effect of finishing agents on the friction of fibers by measuring the static and dynamic friction coefficient of fibers. Azuma 9 discussed various models for the friction of fibers, describing the possibility that physicochemical (i.e., electrical) phenomena at the interface between the fibers would cause adhesion between the fibers, inducing frictional forces. Noda and Nakamura 10 described fabric softener as a special surfactant that provides a flexible feel by reducing the friction coefficient between fibers. Sebastian et al. 11 examined the changes of the pull-out resistance of yarns pulled from fabric treated with cationic softener. They reported that the treatment reduces the tensile modulus of the yarn in the weave, the inter-yarn adhesion, and the inter-yarn sliding friction, while also increasing the deformability and recoverability of the fabric. Fujii et al. 12 experimentally investigated the relationship among torsional rigidity, compressional elasticity, the friction coefficient, and the softness of nylon and wool fibers treated with various surfactants. In their results, although the friction coefficient was reduced, a correlation between the feel and the friction coefficient was not observed. As described above, although the friction of fibers or yarns is reduced by fabric softener treatment, it is not directly related to the softness and flexibility of the fabric.
Igarashi et al. 13 introduced a new softening mechanism for cotton yarn via a softener comprising cationic surfactants. They measured the bending rigidities of cotton yarn with and without a softener treatment under different drying conditions; that is, drying under 50% relative humidity (RH) and drying in vacuo. In their results, the bending rigidity of cotton yarn without softener treatment increases after natural drying compared to that dried in vacuo, which is similar to the trend observed for yarn treated with softener. They showed that the rigidity increase was caused by cross-linking among inner fibers aided by hydrogen bonding (Figure 2). The softener treatment prevents hydrogen bonding and imparts a soft feel to the fabric. Igarashi et al. 14 also investigated the relationship between shear properties and softener treatments, taking into account adsorption of fabric softener agents to yarns. In their results, both the shear stiffness and shear hysteresis decreased with increasing amounts of softener agent, which is the same result found by Inoue et al. 6 Igarashi et al. 14 showed the uneven adsorption phenomenon of fabric softener agents to yarns by using bromophenol blue coloring reaction. The unevenness was happening gradationally more on the outer surface and less on the inner surface of yarns.

Schematic of the inter-fiber cross-linking of cotton fibers by hydrogen bonding (a) and its prevention in the presence of a surfactant (b). 13
In their mechanism, the existence of water at the cotton fiber is important but it was not verified experimentally. Recently, Igarashi et al. 15 reported that bound water, which has a pseudo liquid property, exists at the surface of a cotton fiber by using atomic force microscopy (AFM) and atomic force microscopy-based infrared spectroscopy (AFM-IR). The water with the pseudo liquid property has been known to have higher viscosity, dependent on the condition 16 (the viscosity of bulk water is 0.89 mPa·s), and that of pseudo liquid water is known to be greater. This can be the cause of friction caused by the capillary adhesion by this highly viscous water. Thus, the results prove the softening model of preventing inter-fiber cross-linking of cotton fibers by hydrogen bonding.
Although it was shown that shear stiffness and shear hysteresis of fabric decreased by softener treatments, the effect of a softener on the yarn crossing torque and compressional properties of yarns, which are related to the shear properties of fabric, remains unclear. Yarn crossing torque is the resistance torque of crossing yarns against the change in the crossing angle. Niwa et al. 17 reported that the shear stiffness of fabric is partially dependent on the crossing torques required to change the intersecting angle between warp and weft yarns. In addition to the crossing torque, it is assumed that the transverse compression of yarns can also have an effect on the shear stiffness, because the fabric yarns become closer while shear deforming. Thus, it is necessary to investigate the effect of fabric softener on the crossing torque and compressional properties of yarns in terms of Igarashi’s model. By measuring the properties of the crossing torque–intersecting angle behavior and the transverse compression properties of yarns, the effect of softener on these mechanical properties can be clarified.
In this study, the effect of a fabric softener on the crossing torque properties and compression properties of cotton yarns, where both could be related to shear properties, was investigated. After the cotton samples underwent a water treatment with and without softener, the shear properties of cotton fabric via the crossing torque and hysteresis in the crossing torque–intersecting angle curve (i.e., crossing torque hysteresis) were measured, and the compression properties of cotton yarns were also measured. When cotton fabric dries, hydrogen bonds can occur at yarn intersections in the fabric. 15 Therefore, the properties of dried yarn whose fibers are in parallel strands and whose fibers are woven could be different. For this reason, the results for the two different yarn conditions of drying in parallel and in a crossed state that assumed an intersection between the warp and weft were compared. Then, the changes of those properties as a function of softener treatment and drying conditions were discussed. The applicability of Igarashi’s model of the softening mechanism was also discussed.
Experimental details
Fabric shear property measurements
To verify the effect of a softener on the shear properties of a fabric, 100% plain cotton fabric (single yarn, yarn count (warp × weft) 14.5 tex × 14.5 tex, weave density 60.0 ends·cm−1 × 28.6 pics·cm−1, mass per unit area 12.9 mg/cm2, thickness 0.245 mm) was used. To remove the sizing compound remaining on the yarns of the fabric, we washed the fabric samples in a fully automatic washing machine (NW-6CD, Hitachi, Ltd) using detergent (Emulgen 108, Kao Corp., Japan) and tap water (detergent:water ratio of 1 g:10 L). The fabric underwent five cycles of the laundering process, where each cycle comprised one washing (9 min); one rinse; and wringer water removal (3 min). In addition, two cycles of the laundering process were used without the detergent. Then, the fabric was stretched and dried naturally for 24 h.
The washed fabric samples then underwent a water treatment with and without softener, as shown in Figure 3. Using an immersion shaker (exhibiting lateral rolling at a rate of 190–200 min−1), a flask containing the fabric samples and a dispersion of the liquid softener (0.1% o.w.f. distearyl dimethyl ammonium chloride (DSDMAC) cationic surfactant, Tokyo Kasei Corp., Japan) was shaken for 30 min, which is enough time for 100% absorption of the softener 13 (Figure 3(a)). Most of the liquid was then removed from the fabric using a roller-type wringer (Figure 3(b)), whereupon the fabric was dried for 24 h under the standard condition of 20 ± 1°C and 65 ± 5% RH (Figure 3(c)). For the treatment without softener, the same process while adding water instead of a softener in the dispersion was carried out.

Shear properties sample treatment. Softener and water treatment process of fabric, including (a) fabric treatment, (b) wringer water removal, and (c) drying.
The shear properties of the samples were measured using the KES-FB1 system. 18 The number of measurements was five per sample and the average was used as the result. The measurement environment was under the standard condition (20 ± 1°C and 65 ± 5% RH).
Measurement of the crossing torque properties
To measure the yarn torque properties, 100% cotton two plied yarns (14.5 tex × 2 tex, 40/2 NeC, yarn twist 497 turns/m) were used. The yarn bundle samples (each ∼3 g) were prepared using a yarn winding device, as shown in Figure 4. To remove the sizing compound present on the yarns, the yarn samples were washed in a dispersion of detergent (Emulgen 108, Kao Corp., Japan) and tap water (1 g:10 L). For the washing process, each yarn bundle was placed in a flask, whereupon enough dispersion liquid was poured into the flask to immerse the yarn bundle (400–500 mL dispersion total for 11 yarn bundles). Using an immersion shaker, the yarn bundles underwent five washing cycles, where each cycle included a 9 min washing; 5 min rinse; and one rotation through the roller-type wringer. After the five washing cycles, the bundles underwent two of the washing cycles excluding the detergent, whereupon the bundles were dried.

Crossing torque properties sample yarn bundle.
Next, the washed yarn bundles underwent a water treatment with and without softener, as shown in Figure 5. Using an immersion shaker, a flask containing the yarn samples and a dispersion of the liquid softener (0.1% o.w.f. DSDMAC cationic surfactant) was shaken for 5 min, which is enough time for 100% absorption of the softener. 13 In this treatment, the ratio of softener dispersion to sample was 328 g softener dispersion:15 g yarn. Then, the liquid in the treated yarn bundles was removed using the roller-type wringer and the treated bundles were dried for 24 h under the standard condition. For the treatment without softener, we carried out the same process while adding water instead of a softener in the dispersion.

Crossing torque properties sample treatment. Softener and water treatment process of yarn bundles, including (a) yarn bundle treatment, (b) wringer water removal, and (c) drying.
Crossing torque properties can be measured by turning the two interlaced yarns by applying a contact load (1.83 g) using a dead weight hung forward up to the crossing angle of 30 degrees and then turning it backward up to the same angle, as shown in Figure 6.19–21 The yarn inclination angle between the yarn axis and the horizontal axis is 30°. The slope between 10° and 20° and hysteresis at 15° in the crossing torque–intersecting angle curve can be used as the crossing torque and hysteresis. To measure the crossing torque properties, the treated yarns were intersected before drying (Figure 7(a)) by crossing two undried yarns and applying a load (1.83 g). The intersected yarn samples were then dried for 24 h (Figure 7(b)), and the resulting samples were referred to as crossed-yarn dried (CD).


Crossing torque properties sample drying: (a) intersected yarns with applied load; (b) drying method for the crossed-yarn dried samples.
The torque properties of the yarns were measured using a multipurpose torsion and crossing torque tester (KES-YN1, Katotech Co. Ltd, Kyoto). Thirty samples were measured for each treatment type, where the measurement environment was under the standard condition (20 ± 1°C and 65 ± 5% RH).
Measurement of yarn compression properties
To measure the compression properties of yarn, two types of samples with different post-treatment drying conditions were made, as shown in Figure 8. In the first, a yarn was dried in a straight configuration, referred to as a straight-yarn dried (YD) sample. In the second, two yarns were crossed and then dried, referred to as CD samples, which imitated the crossed yarns in a woven fabric. For the YD, a constant tensile load of 7.5 gf was applied to each yarn to straighten it, whereupon both ends of the yarn length were fixed to a plastic frame using adhesive tape (Figure 9). The CD samples were also prepared using the plastic frame with the same load (Figure 8(c)).

Yarn compression properties sample preparation. (a) Schematic of the plastic frame. (b, c) Photographs of the sample setup for the (b) straight-yarn dried and (c) crossed-yarn dried.

Yarn compression properties sample preparation setup for applying an initial load.
Both types of yarn samples were treated with and without softener while mounted on the plastic frame. For the treatment, the frames were placed into a tub with the softener dispersion and treated using the same process used for the yarn bundles. Owing to the small amount of yarn in these samples, the yarn samples were treated together with extra fabric to ensure the same ratio of dispersion liquid to cotton. Instead of using the immersion shaker during the treatments, however, the tub was manually shaken up and down for enough time to absorb the softener. After drying, the yarns were removed from the frames.
The compression properties, including the compressional linearity (LC), compressional energy (WC), and compressional resilience (RC) of the yarn samples, were measured using a compression tester (KES-FB3, Katotech Co. Ltd, Kyoto) with a maximum load of 200 gf and a circular compressor with an area 18 of 2 cm2. The measurement speed was 0.02 (mm/s). Fifteen samples were measured for each treatment type, where the measurement environment was under the standard condition (20 ± 1°C and 65 ± 5% RH).
Results and discussion
Softener treatment effect on shear properties of fabric
Figure 10 shows a comparison of shear stiffness of fabrics treated with and without softener. Figures 11 and 12 show a comparison of the shear hysteresis (at 0.5° and at 5° of the shear curve). All of the values of the samples with a softener treatment were lower than those with a water treatment with a significant difference at the 1% level. Therefore, it is clear that it is easier for the fabrics to shear and to recover from shear deformation after softener treatment than after water treatment. These results are consistent with those of Inoue et al. 6 and Igarashi et al. 14

Comparison of shear stiffness of fabrics treated with water (WF) and softener (SF) (** significant difference at the 1% level).

Comparison of shear hysteresis (at 0.5°) of fabrics treated with water (WF) and softener (SF) (** significant difference at the 1% level).

Comparison of shear hysteresis (at 5°) of fabrics treated with water (WF) and softener (SF) (** significant difference at the 1% level).
Effect of the treatments and drying conditions on the crossing torque properties
Figure 13 shows typical crossing torque versus intersecting angle curves for the YD and CD samples with and without softener treatment (labeled S and W, respectively). Figure 14 shows comparisons of the crossing torque and crossing torque hysteresis values of the YD and CD samples for S and W conditions. The crossing torque value of the S-YD sample was lower than that of the W-YD sample, with a significant difference at the 5% level. Therefore, it was found that the crossing torque of yarns became lower after softener treatment. When comparing crossing torque value of the samples without softener (i.e., W-YD and W-CD), the crossing torque value of W-CD is higher than that of W-YD. Figures 15 and 16 show the crossing points of the W-CD and W-YD samples, respectively, where a sharper dented surface is seen on the crossing point of W-CD than on the crossing point of W-YD. A sharper dented surface will cause higher friction and difficulty of moving between yarns while crossing yarns. Therefore, it is conceivable that the crossing torque value is affected by the changes of the yarn shape at the crossing point.

Crossing torque versus intersecting angle for (a) straight-yarn dried and (b) crossed-yarn dried samples with (S) and without (W) softener treatment.

(a) Crossing torque and (b) crossing torque hysteresis of straight-yarn dried (YD) and crossed-yarn dried (CD) samples with (S) and without (W) softener treatment (** significant difference at the 1% level, * significant difference at the 5% level).

Photographs of crossing points of the crossed-yarn dried sample without softener.

Photographs of crossing points of the straight-yarn dried sample without softener, after drying for 15 min while applying a constant 1.83 g load on the yarn.
The crossing torque hysteresis values of the W-YD and W-CD samples were higher than those of the S-YD and S-CD samples (with a significant difference at the 1% level) owing to the softener adhered to the yarn surface. It is considered that the softener reduces the friction at the yarn crossing point. Furthermore, the torque hysteresis value of the W-CD sample is larger than that of the W-YD sample, although the torque hysteresis values of the S-YD and S-CD samples are similar to each other. Because the same softening agent was used in the treatments, it is assumed that the amount of softening agent attached on the surfaces of the samples was the same. Accordingly, the surface friction should be the same, which correlates well with the similarity of the torque hysteresis values of the S-YD and S-CD samples.
Effect of the treatments and drying conditions on the compression properties
Figure 17 shows typical compression curves of the W-YD, S-YD, W-CD, and S-CD samples. Figures 18–20 show the compression property values of LC, WC, and RC, respectively, exhibited by the four sample types. The LC and WC values of the W-YD and S-YD samples were similar to each other, as were those of the W-CD and S-CD samples. Comparing the WC values of the W-YD and W-CD samples, that of the W-CD sample is larger than that of the W-YD sample. However, the WC value of the W-CD sample is not twice that of the W-YD sample, as one would expect because of the doubling of the number of yarns in the sample. This result is due to the deformation of the yarn at the cross-point varying between the samples with the crossed yarn and the straight yarn. For the RC of the samples, the RC value of the S-CD sample is larger than that of the W-CD sample, with a significant difference at the 1% level (Figure 20(b)), although the WC values of the S-YD and W-YD samples are similar. This indicates that the S-CD sample has a higher recovery of the compression than the W-CD sample, which could be due to two reasons: the reduction of friction and the reduction of hydrogen bonding at the yarn intersection. These results could support the model proposed by Igarashi et al., 15 because the friction is caused by the pseudo liquid type water originated from hydrogen bonding.

Comparison of the compression curve of (a) straight-yarn dried and (b) crossed-yarn dried samples with water (W) and softener (S) treatment.

Comparison of the compressional linearity (LC) of the (a) straight-yarn dried (YD) and (b) crossed-yarn dried (CD) samples with water (W) and softener (S) treatment.

Comparison of the compressional energy (WC) of the (a) straight-yarn dried (YD) and (b) crossed-yarn dried (CD) samples with water (W) and softener (S) treatment.

Comparison of the compressional resilience (RC) of the (a) straight-yarn dried (YD) and (b) crossed-yarn dried (CD) samples with water (W) and softener (S) treatment (** significant difference at the 1% level).
Conclusion
The effect that a fabric softener comprising a cationic surfactant had on the yarn crossing torque properties and compression properties of cotton yarn was investigated. These properties are related to shear properties, and are revealed by measuring the crossing torque, crossing torque hysteresis, and compression properties of cotton yarns treated with and without softener. The effect of the drying condition was also investigated by varying the drying state of the yarns: placing them in either a straight configuration with a single yarn or crossed configuration with two yarns crossing.
The results of the crossing torque measurements showed that the fabric softener treatment reduced the crossing torque in yarn dried in a straight configuration but not in a crossed configuration, and reduced crossing torque hysteresis in both configurations. This could be due to the reduced friction between yarns induced by the softener treatment and/or to the shape change at the yarn intersection point. In the results of the compression properties measurements, the RC value of the crossed and dried yarns increased with softener treatment, signifying that the yarns were more recoverable. Therefore, it was found that softener treatment makes the yarn more recoverable at the yarn intersection point. This could be due to a reduction of friction and a reduction of hydrogen bonding at the yarn intersection via the softener. This study will aid in making the use and development of fabric softener more effective.
This study revealed the effect of softener treatment on yarn crossing torque properties and compressional properties. A limitation of the study was that the effects of the softener on the friction and on the hydrogen bonding were not able to be separated. However, the non-separation of the results from two different methodologies conversely supports the model proposed by Igarashi et al., 15 because the friction itself is caused by the highly viscous bound water 19 at the surface of cotton single fibers. By interpreting the measurement results with Igarashi’s model, new softening technology, such as softening agents, formulations, and systems, can be constructed.
