Machine-Washable PEDOT:PSS Dyed Silk Yarns for Electronic Textiles.

Durable, electrically conducting yarns are a critical component of electronic textiles (e-textiles). Here, such yarns with exceptional wear and wash resistance are realized through dyeing silk from the silkworm Bombyx mori with the conjugated polymer:polyelectrolyte complex PEDOT:PSS. A high Young's modulus of approximately 2 GPa combined with a robust and scalable dyeing process results in up to 40 m long yarns that maintain their bulk electrical conductivity of approximately 14 S cm-1 when experiencing repeated bending stress as well as mechanical wear during sewing. Moreover, a high degree of ambient stability is paired with the ability to withstand both machine washing and dry cleaning. For the potential use for e-textile applications to be illustrated, an in-plane thermoelectric module that comprises 26 p-type legs is demonstrated by embroidery of dyed silk yarns onto a piece of felted wool fabric.

Introduction

Electronic textiles (e-textiles) currently receive tremendous attention for a range of application areas from medical care[1,2] to energy harvesting and storage.[3−7] A critical component is electrically conducting fibers that can be incorporated through weaving, sewing, knitting, or embroidery.[8] A prerequisite for functional fibers is wash and wear resistance, which is necessary to withstand both textile manufacture and daily use. Integration of the conducting material into the fiber, which can be achieved through either incorporation during fiber spinning or impregnation of an existing fiber, is likely to result in a more durable functionalization because a surface coating is prone to delamination during bending and abrasion during wear.

The use of natural materials such as polysaccharide-based cotton and protein-based silk offer a number of advantages because they can be derived from renewable sources, are biocompatible, and, being traditional textile materials, are readily compatible with existing manufacturing routines.[9] A lasting wear and wash resistant finish can be realized through functionalization with materials that carry ionizable groups, which is a common feature of acid dyes that are used by the textile industry to color natural fibers such as wool, mohair, and silk.[10] Similarly, water-soluble conjugated polyelectrolytes that carry negatively charged acetate or sulfonate counterions can tightly bind to silk through electrostatic interactions with cationic sites, provided that the pH is adjusted during the dyeing process to ensure polypeptides with a net positive charge.[11,12]

One of the most versatile classes of conjugated materials is based on poly(3,4-ethylenedioxythiophene) (PEDOT),[13,14] which is typically processed as a water-dispersible complex with the polyelectrolyte poly(styrenesulfonate) (PSS). Coating of synthetic and natural materials with PEDOT:PSS has been used to realize a wide range of fiber- and textile-based electronic components, recent examples of which include electrical interconnects,[15] electrodes for electrocardiography,[2] electrochemical transistors,[16,17] electrochromic pixels,[18] electrodes for organic solar cells,[19] organic light-emitting diodes (OLEDs)[20] and piezoelectric sensors,[21] pressure sensors for stretchable keyboards,[22] and thermoelectric elements.[23] The wash and wear resistance of PEDOT:PSS-coated fibers and textiles remains a challenge,[21,24] though recent progress has been reported by Guo et al., who reported wash-resistant PEDOT:PSS patterns on top of nonwoven polyethylene terephthalate (PET) fabrics.[25]

Here, we demonstrate that PEDOT:PSS can be used akin to an acid dye. Impregnation of silk from the silkworm Bombyx mori allows us to realize durable high-modulus yarns with a bulk electrical conductivity of 14 S cm–1. A high Young’s modulus of approximately 2 GPa combined with a robust and scalable dyeing process results in up to 40 m long yarns that maintain their electrical properties when experiencing repeated bending stress as well as mechanical wear during sewing. Moreover, a high degree of ambient stability is paired with the ability to withstand both machine washing and dry cleaning. To illustrate the potential use for e-textile applications, we realize an in-plane thermoelectric module that comprises 26 p-type legs by embroidery of dyed silk yarns onto a piece of felted wool fabric.

Experimental Section

Materials

Degummed silk yarn was obtained from Aurora Silk, USA and used as received (Ahisma sewing thread; ⌀yarn ≈ 0.25 mm as measured with an optical microscope). Cotton yarn was obtained from Gütterman, Germany and used as received (CA02776; ⌀ ≈ 0.18 mm). Water dispersion of PEDOT:PSS was purchased from Heraeus, Germany (PH1000; ratio 1:2.5; solid content ≈ 1.1–1.3 wt %; pH ∼2). PEDOT-S was synthesized according to a previously published procedure.[26] Ethylene glycol (EG), dimethyl sulfoxide (DMSO), methanol 99% (MeOH), ethanol 95%, and fuming hydrochloric acid (HCl) were purchased from Sigma-Aldrich and used as received. Zonyl FS-300 was purchased from Dupont. Silver paint was purchased from Agar Scientific, UK. For the textile thermoelectric devices, a stretchable silver paste (PE872 Conductor paste, DuPont, UK) was used. After application, this silver paste was dried for 1 h at room temperature and then cured at 100 °C for 10 min. The silver wire was 0.3 mm in diameter.

Sample Preparation

The PEDOT:PSS dispersion (pH 2) was mixed with 0.2 vol % Zonyl and 5 vol % EG or 5 vol % DMSO as indicated, vortexed, and allowed to settle for several hours. Then, silk or cotton yarns were submerged twice, sonicated for 1 h (maximum power = 175 W; 300 Ultrasonik, NEY, USA) and dried at 130 °C for 30 min. Optionally, silk fibers prepared without EG or DMSO were subsequently treated with MeOH and again dried at 130 °C. Dyeing with PEDOT-S was performed by submerging silk yarns for 1 h at 90 °C in PEDOT-S solution (pH 1; ∼5 g L–1 HCl), followed by removal of excess PEDOT-S in ethanol and finally drying under N2 flow.

Mechanical Characterization

Tensile testing of 50 cm long yarns was carried out with an Instron tensile tester (model 5565A) at a cross-head speed of 10 mm min–1.

Electrical Characterization

The electrical resistance of yarns was determined with a Keithley 2400 sourcemeter in two-point probe configuration by contacting ∼1 cm long segments with silver paint (see Table for number of measured segments, n). The Seebeck coefficient of ∼5 mm long segments was measured at 300 K with an SB1000 instrument equipped with a K2000 temperature controller from MMR Technologies using a thermal load of approximately 1–2 K and a constantan wire as an internal reference. Samples were mounted with silver paint. The change of resistance of yarns during stretching with a custom-built setup was measured with a Keithley 2400 sourcemeter by connecting yarns with flat crocodile clips and silver paste. The resistance during cyclic bending around Teflon bars (⌀ ≈ 4.5 or 52 mm) was measured with a Keithley 2400 sourcemeter using a home-built setup constructed with LEGO. Yarns were contacted with silver paste, followed with copper tape (to prevent silver paste rubbing off) and then toothless crocodile clips.

Table 1

Mechanical and Electrical Properties of Silk and Cotton Yarns Dyed with PEDOT-S and PEDOT:PSS: Stress at Break σbreak, Strain at Break εbreak, Young’s Modulus E, Electrical Conductivity cyarn Relative to the Fiber Cross Section, and Seebeck Coefficient α, Measured for n Yarn Segments

			mechanical properties				electrical properties
yarn	coating	processing agent	σbreak (MPa)	εbreak (%)	E (GPa)	n	cyarn (S cm–1)	n	α (μV K–1)	n
silk			157 ± 12	11 ± 1	3.2 ± 0.3	45
	PEDOT-S		93 ± 16	13 ± 1	2.4 ± 0.5	2	(3 ± 2) × 10–2	12	8 ± 1	2
	PEDOT:PSS		109 ± 31	8 ± 2	2.6 ± 0.5	3	(1 ± 1) × 10–2	15	18 ± 2	3
	PEDOT:PSS	EG	139 ± 22	12 ± 2	2.0 ± 0.2	21	15 ± 6	84	14 ± 1	4
	PEDOT:PSS	DMSO	136 ± 24	12 ± 3	1.8 ± 0.3	23	14 ± 4	84	15 ± 1	4
	PEDOT:PSS	MeOH	159 ± 20	12 ± 2	1.9 ± 0.2	18	14 ± 6	53	17 ± 3	4
cotton			305 ± 27	7 ± 1	5.3 ± 0.6	20
	PEDOT:PSS	EG	261 ± 35	6 ± 1	5.8 ± 1.8	8	12 ± 3	12	14 ± 1	2
	PEDOT:PSS	DMSO	260 ± 28	6 ± 1	7.0 ± 2.5	8	15 ± 6	18	14 ± 1	2
	PEDOT:PSS	MeOH	a	a	a	a	15 ± 4	18	16 ± 3	2

Samples too brittle to be measured as the methanol may have further degraded the cotton.

Light Microscopy

Light microscopy was carried out in bright field reflected light mode with a Carl Zeiss A1 optical microscope.

Scanning Electron Microscopy (SEM)

Measurements were taken with a Leo Ultra 55 SEM equipped with a field emission gun (LEO Electron Microscopy Group, Germany) and a secondary electron detector. The acceleration voltage was 3 kV. Samples were freeze-fractured by hand under liquid nitrogen or cut by razor blade. Some of the samples were sputtered with gold.

Results and Discussion

Initially, we explored dyeing of silk with two types of PEDOT-based materials: the conjugated polyelectrolyte PEDOT-S and a commercial PEDOT:PSS formulation. Drop-cast films displayed an electrical conductivity of approximately 1 S cm–1 [26] and up to 500 S cm–1, respectively (Table S1). Both materials feature sulfonate groups in aqueous media, which due to the low pKa < 0 of sulfonic acids, is always dissociated.[11] Degummed silk is composed of fibroin, which has an isoelectric point around pH 4,[27] meaning that the fiber will carry a net positive charge at sufficiently acidic conditions. We therefore chose to work at pH 1 for PEDOT-S experiments and the inherent pH 2 of the PEDOT:PSS dispersion used here (PH1000 from Heraeus). We find that both PEDOT-S and PEDOT:PSS readily bind to silk at low pH. In contrast, at less acidic conditions, e.g., pH 2 for PEDOT-S and pH 9.5 for ammonium hydroxide-neutralized PEDOT:PSS, dyeing was less successful as evidenced by the lack of coloring of silk yarns by the conducting material, which confirms the importance of favorable electrostatic interactions.[11] We explored a wide range of conditions to select a dyeing process akin to industrially used exhaust dyeing where the textile material gradually takes up dye molecules from a large dye bath (see the Experimental Section; weight uptake of approximately 8–12% in the case of PEDOT:PSS). The scalability of the employed process allowed us to dye up to 40 m long yarns (diameter ⌀yarn ≈ 0.25 mm), which featured a uniform blue color and hence similar resistivity along their whole length (Figure a).

Figure 1

(a) Bundle of PEDOT:PSS dyed silk yarn (left) and optical microscopy image of neat and PEDOT:PSS dyed silk yarns with a diameter ⌀yarn ≈ 0.25 mm (right); (b) SEM images of freeze-fractured PEDOT:PSS dyed silk yarns prepared with EG (top) or DMSO (bottom); right row of SEM images recorded after sputtering yarns with gold; scale bars = 2 μm (see Figure S1 for SEM image of pristine silk). Arrows indicate delaminated PEDOT:PSS.

Closer inspection of dyed silk yarns with scanning electron microscopy (SEM) confirmed the strong interaction between PEDOT:PSS and silk. SEM images of freeze-fractured dyed yarns revealed that individual fibers (⌀fiber ≈ 10 μm) are composed of a core–shell structure (Figure b). Charging artifacts, which were absent for sputtered samples, indicate a poorly conducting core that is surrounded by an approximately 1 μm thick conducting outer layer from which an approximate 100 nm thin film detaches at certain places (see arrows in Figure ). It appears that the dyeing process resulted in impregnation of the outer layer of individual silk fibers with PEDOT:PSS, which suggests a high degree of interaction.

Dyeing with both PEDOT-S and neat PEDOT:PSS yielded silk yarns with a bulk electrical conductivity of cyarn ≈ 10–2 S cm–1 (Table ; note that cyarn is given relative to the cross sectional area A = π⌀yarn2/4 according to cyarn = R–1 · L/A, where R is the resistance of yarn segments with length L ≈ 1 cm). We also measured the Seebeck coefficient αyarn of the yarns prepared here. Values of αyarn ≈ 8 and 20 μV K–1 for PEDOT-S and PEDOT:PSS dyed yarns are in agreement with a previous report.[28] Several methods exist that enhance the electrical conductivity of PEDOT:PSS, including the addition of ethylene glycol (EG) or dimethyl sulfoxide (DMSO) to the dispersion or post-treatment of solidified material with methanol (MeOH).[29] Regardless of the treatment method, we obtained silk yarns with a significantly increased bulk electrical conductivity between cyarn ≈ 4 and 32 S cm–1 (Figure a). A mean value of cyarn ≈ 14 S cm–1 (Table ) rivals results obtained by oxidative chemical vapor deposition of PEDOT on viscose and dip-coating of silk in PEDOT:PSS.[15,30] The Seebeck coefficient only slightly decreased to, e.g., αyarn ≈ 15 μV K–1 in the case of DMSO, indicating that these additives predominately modify the nanostructure of PEDOT:PSS and not the doping level. We also attempted to treat cotton yarns in the same way as silk, which yielded comparable results (Table ).

Figure 2

(a) Distribution of electrical conductivity cyarn of 221 segments of PEDOT:PSS dyed silk yarns prepared with DMSO (red), EG (blue), and treated with MeOH (green) and corresponding log-normal distribution with a peak at 11.3 S cm–1 (solid line); (b) resistance R of 32 cm long yarns bent repeatedly around Teflon bars with a diameter ⌀ of 4.5 (•) and 52 mm (◊).

We noted that the acidic processing conditions affected silk and cotton yarns to a different extent. Neat silk yarns displayed a Young’s modulus of E ≈ 3.2 GPa and strain at break εbreak ≈ 11% (Table ). After treatment with acidic PEDOT:PSS dispersion (or aqueous HCl at pH 2), we observed the same εbreak but lower modulus around E ≈ 2 GPa. The use of PEDOT-S gave rise to similar results. Instead, cotton yarns suffer from acidic conditions due to hydrolysis of cellulose[31] and, in the case of methanol treatment, were too fragile to be handled without fracture. To avoid acid hydrolysis, we neutralized PEDOT:PSS with ammonium hydroxide, which allowed for maintaining the mechanical strength of the yarns but resulted in significantly lower cyarn ≈ 1 and 10–1 S cm–1 in the cases of silk and cotton, respectively. Therefore, we chose to limit all subsequent studies to silk yarns dyed with PEDOT:PSS.

In a further set of experiments, we studied the impact of mechanical stress on the electrical resistance of dyed silk yarns. We note that stress–strain curves display close to linear behavior (Figure S2), which suggests that PEDOT:PSS dyed silk yarns are stiff and do not undergo plastic deformation before failure. Accordingly, the electrical resistance remains unaffected up to a strain of at least 10% (Figure S2). When incorporated into e-textiles, conducting yarns are likely to experience recurring mechanical stress. We therefore monitored the influence of repeated bending of yarns around Teflon bars with a diameter of 4.5 or 52 mm (Figure b). We find that after 1000 bending cycles the resistance had increased by not more than 50%, indicating a robust electrical performance of the yarns.

We studied the evolution of the electrical resistance and Seebeck coefficient over the course of 16 weeks to assess the long-term stability. Initially, the resistance of yarns doubled during the first 2 weeks after dyeing (Figure ). During the subsequent 14 weeks, the resistance changed by not more than 10%. The slight increase in resistance was accompanied by a minor increase in αyarn by approximately 1 μV K–1 after 2 weeks, which returned to its original value at longer times. Evidently, the PEDOT:PSS-coated silk yarns studied herein are characterized by a high degree of ambient stability.

Figure 3

Aging of 1 cm long segments of PEDOT:PSS dyed silk yarns under ambient conditions: (a) resistance R relative to initial resistance R0, (b) Seebeck coefficient αyarn (MeOH (◊), EG (blue triangle), DMSO (red circle)), and ambient humidity (stars).

The ability of conducting fibers to withstand prolonged exposure to aqueous environments and mechanical wear is critical for use in e-textile applications. We therefore studied the impact of washing on the electrical conductivity of PEDOT:PSS dyed silk yarns. Initially, we noted that PEDOT:PSS did not detach even after submerging dyed yarns in water for up to 4 days (Figure a). We then washed the conducting yarns in a standard household washing machine: the yarns were zigzag-stitched onto swatches of cotton fabric, which we placed in a laundry bag for machine washing using 20 mL of a common color detergent. We selected a “hand wash” program (30 °C, 50 min, spinning at 900 rpm) and added two towels for additional weight. The swatches were line dried at room temperature for at least 6 h prior to the next washing cycle (Figure b). We prepared four sets of two swatches that contained three dyed silk yarns each and washed them 1–4 times. For each set of swatches, we then measured the electrical conductivity by comparing 15 ∼1 cm long yarn segments. We found that PEDOT:PSS dyed silk yarns display, within the experimental error, no change in electrical conductivity after four washing cycles. It appears that the dyed silk yarns are characterized by an exceptional resistance against the wear experienced during a common machine washing program. We also investigated dry cleaning by submerging and gently shaking dyed silk yarns in the common solvent tetracholoroethylene for 1 h followed by drying at 130 °C for 10 min. After five consecutive dry cleaning cycles, the electrical conductivity had decreased by a factor of not more than 2, thus emphasizing the washable nature of dyed silk yarns (Figure S3).

Figure 4

(a) Neat and PEDOT:PSS dyed silk yarns submerged for 1 h in distilled water; (b) PEDOT:PSS dyed silk yarns were sewn onto cotton fabrics and washed in a household Electrolux washing machine at 30 °C for 50 min and a 900 rpm spin drying cycle; (c) using 20 mL Neutral COLOR WASH detergent; (d) electrical conductivity cyarn of PEDOT:PSS dyed silk yarns as a function of washing cycle (MeOH (◊), EG (blue triangle), DMSO (red circle)).

We chose to showcase the usefulness of PEDOT:PSS dyed silk in two types of e-textile applications: (1) as wires to connect electronic elements and (2) as p-type legs in fabric-based thermoelectric modules, which offer an alternative approach for the design of this type of device.[23,32,33] A circuit comprising a battery, conducting yarn, and LED was embroidered (couch stitch method) upon a swatch of felted wool fabric, which could be repeatedly bent around the wrist of the lead author (Figure a). Similarly, we devised a simple thermoelectric device that consisted of two ∼5 cm long p-type legs of dyed silk (10 yarns per leg) that we short-circuited with silver wire (Figure b). The yarns and wires were again embroidered onto a piece of felted wool fabric and connected with silver paste, applied using a rubber stamp. We then applied a temperature difference of up to ΔT ≈ 50 °C across this in-plane device, which resulted in an output voltage of Vout/ΔT ≈ 26 μV K–1, i.e., 13 μV K–1 per element. For ne elements of dyed yarn plus silver wire, we predict a Vout/ΔT ≈ ne(αyarn – αsilver), which for ne = 2, αyarn ≈ 15 μV K–1 (see Table ), and αAg ≈ 1.5 μV K–1 yields a value of Vout/ΔT ≈ 27 μV K–1 (see Figure b).

Figure 5

(a) Image of an LED connected with PEDOT:PSS dyed silk yarns to a battery (embroidered on felted wool fabric); (b) in-plane textile thermoelectric device with two ∼5 cm long p-type legs of PEDOT:PSS dyed silk yarns (10 yarns per leg) connected with silver wire (inset) and output voltage Vout as a function of applied temperature gradient ΔT (single element (black line), two elements (red line), calculated output (dotted line) Vout/ΔT = ne(αyarn – αsilver), where ne = 2, αyarn = 15 μV K–1, and αsilver = 1.5 μV K–1); (c) in-plane textile thermoelectric device with 26 p-type legs prepared in a similar way as the two leg module, and (d) the device in use, suspended between a hot and cold temperature reservoir (hot plate used for hot side, steel heat sink for cold side, and PET foil as electrical insulation); (e) electrical measurements of the 26 leg module, showing Vout as a function of ΔT (all elements (red line), calculated (dotted line) Vout/ΔT; see (b)) and power output P = VoutI as a function of measured current I for ΔT = 66 °C using load resistors ranging from Rload = 1 to 27 kΩ.

We went on to prepare a scaled-up version of a thermoelectric module that comprised 26 elements, again with ∼5 cm long p-type legs of dyed silk (10 yarns per leg) and a single silver wire as counterpart (Figure c). Contacts were made with silver paste on the backside of the fabric by sewing dyed silk yarns and silver wire through the fabric (Figure S4), which confirms the durability of the yarns despite rather vigorous handling. The module could be twisted repeatedly without compromising its total electrical resistance of Rin ≈ 8.7 kΩ, which highlights the inherent flexibility of the ensemble. We exposed the module to an in-plane temperature difference of up to ΔT ≈ 120 °C without an adverse effect on Rin (Figure d) and find an output voltage of Vout/ΔT ≈ 313 μV K–1 (i.e., 12 μV K–1 per element), which is close to the predicted value of Vout/ΔT ≈ 351 μV K–1 (Figure e). We then held the module at an in-plane temperature difference of ΔT ≈ 66 °C and connected a series of load resistors with Rload ranging from 1 to 27 kΩ. For load matching conditions Rload = Rin, we measured a current of 1.25 μA and hence a maximum power output of Pmax ≈ 12 nW (Figure e). The maximum power can be predicted according to Pmax = Vout/4Rin, which yields a value of approximately 12.3 nW that is in good agreement with our measurement.

Conclusions

In conclusion we have developed a highly durable conductive silk yarn through dyeing with PEDOT:PSS. These yarns, which could be produced with a length of up to 40 m using a scalable exhaust dyeing process, displayed a wide variety of properties that are desired for e-textile applications, including a high tensile modulus of approximately 2 GPa as well as resilience against bending stress and wear during sewing. Moreover, we were able to carry out repeated machine washing and, to some extent, dry cleaning without loss of conductivity. A prototype in-plane thermoelectric module with 26 p-type legs was fabricated to demonstrate the potential of dyed silk yarns for flexible and wearable applications.