Dry cleaning of electroconductive layers screen printed on flexible substrates

Abstract

Smart textiles and wearable textile systems are up to now a growing research field, with a view to applications such as monitoring vital signs or environmental parameters through garments. These developments require new materials such as electroconductive textiles. As a result, research into screen printing of conductive silver inks onto textiles has emerged.

In this study, screen printing is used to print four kinds of silver-based conductive inks on a flexible foam and a nonwoven substrate. The screen printing method is chosen because it is a low cost and user friendly technique to obtain flexible and lightweight conductive fabrics. For the evaluation of the electrical properties the square resistance of the printed fabrics is measured after repeated dry cleaning cycles. The printed textiles studied here show good electrical properties after printing (< 0.05 Ω/□). However, after 60 dry cleaning cycles, the conductivity decreased considerably. Consequently, in order to improve washability, a protective polyurethane layer was put on top of the printed samples. In this case, the resistivity remained below 2.3 Ω/□ after 60 dry cleaning cycles.

Keywords

ink printing properties

Textiles with electroconductive characteristics have been found to be useful in several fields of application, including military, electronic, medical and automotive.^1–3 Textiles are preferred to other types of material when the end product needs to be flexible, lightweight and washable. To obtain these electroconductive textiles several methods can be applied. These range from conductive fibers or yarns to coating or to printing with conductive inks.^4,5 Several studies of screen printing with conductive ink on various woven and nonwoven textiles have already been published.^5–10 However, textiles for clothing need to withstand a maintenance procedure: laundering or dry cleaning. To our knowledge, this is the first paper to report on the dry cleaning durability of screen-printed electroconductive flexible substrates. The substrates used in this study are flexible foam and nonwoven polyester (PES). Karaguzel et al.⁷ have already proved that laundering printed electroconductive textiles in a washing machine results in a loss of conductivity. This study, therefore, concentrates on dry cleaning. A further motivation is the fact that dry cleaning is the preferred maintenance method for garments in which these substrate materials will be integrated, for example, protective clothing, such as fire-fighting suits, or casual jackets and outdoor sportswear. The procedure comprised the dry cleaning of printed samples 5, 10, 15, 20 and 60 times. At first, the samples were measured after every single dry cleaning cycle. However, since the differences between the values after each cycle were negligible, the frequency of measuring was decreased to every five cycles.

In this study, we worked with electroconductive flexible substrates screen printed with four different silver-based inks on foam and nonwoven PES. The direct current (DC) resistance of the printed patterns was measured before and after dry cleaning in order to evaluate these four inks. To preserve electroconductive properties after dry cleaning, a protective non-conductive layer was put on top of the printed patterns. The final result was that the values of square resistance R_□ only slightly increased after dry cleaning, varying between 0.025 and 2.263 Ω/□.

Materials and methods

Flexible substrate materials

For this study two flexible substrates were selected: a polyurethane flexible foam (Urecom) and a nonwoven PES (Table 1). These particular two substrates were chosen, firstly, because they are clearly different, making it possible to observe potentially different interaction with the conductive inks and, secondly, because they both have the thickness (>2 mm) required by some applications, such as planar antennas for wireless off-body communication.¹¹

Table 1.

Properties of applied foam and nonwoven polyester (PES)

Printed flexible substrate	Thickness^a (mm)	Basic weight^b (g/m²)	Density^c (g/cm³)
Urecom	3.43	225.18	0.15
PES	3.71	526.36	1.38

ISO 9073-2.

ISO 9073-1.

The density (g/cm³) is indicated from the producers.

Conductive inks

Four different commercial silver-based inks were applied, ink 1 produced by DuPont, ink 2 by Acheson and inks 3 and 4 by Sun Chemical. These conductive inks consist of a dispersion of silver particles and a suitable resin in an organic solvent. Table 2 shows their properties.

Table 2.

Properties of applied silver-based conductive inks

Ink type	Ink 1 (DuPont)	Ink 2 (Acheson)	Ink 3 (Sun Chemical)	Ink 4 (Sun Chemical)
Solid content (%)^a	68–72	73.5–76	69–71	84–86
Cure conditions (120℃)^a	5–6 minutes	15 minutes	2–10 minutes	2–15 minutes
Sheet resistance (Ω/□ at 25 µm)^a	0.012–0.015	<0.025	0.011	≤0.025
Viscosity (Pa.s)^a	20–30	10–25	2–3	7–9
Resin^b	Epoxy	PES	PES	Epoxy

Given by the producers.

Determined by the authors.

PES: polyester.

Screen printing

Screen printing is preferred as a method because it is a flexible, economical and fast process. The screen printing in this study was performed by a semiautomatic machine, Johannes Zimmer Klagenfurt - type Mini MDF 482. In order to achieve an even coverage of ink over the entire pattern (square of 6 cm by 6 cm) a PES mesh was used, monofilament 90 T, with a sieve thickness of 110 µm and a sieve opening of 45%. In order to bond and fix the ink onto the substrate, the samples were cured in an oven at the temperature recommended by the ink producers.

Dry cleaning

Dry cleaning was chosen over laundering because the electroconductive materials examined in this study are meant for integration into garments that are normally dry cleaned. Foam material, for example, is integrated in the shoulder pads of fire-fighter outer jackets to support the weight of the compressed air bottle and nonwovens are commonly used as inlayers in collars and lapels for jackets. In addition, a laundering study is already available: Karaguzel et al.⁷ printed transmission lines on nonwoven substrates and reported that after 25 washing cycles in a home laundry machine conforming to the ISO 6330 standard, the printed inks began to degrade and showed lower conductivity.

The dry cleaning machine used was a UNION XL 835 E, the drum capacity of which is 15 kg. The solvent perchloroethylene (PERC), also known as tetrachloroethylene (PCE), is used as cleaning agent. This is a synthetically produced organic compound.

The dry cleaning process lasts 45 minutes and consists of three steps: washing, extracting and drying. At the beginning of the washing process, the chamber is automatically filled with solvent to approximately one-third full and is heated to a temperature of 30℃. The drum is agitated, in order to allow the solvent to clean the substrate material. The machine has two solvent tanks: one tank contains pure solvent and the other contains used solvent. Generally, clothes are first dry cleaned with filtrated solvent and then rinsed with pure solvent.

Following washing, the extraction cycle begins by draining the solvent from the drum, causing much of the solvent to spin free of the fabric. The used solvent is passed through a filtration chamber and part of the filtered solvent is returned to the second tank to be used in another washing cycle. The residual solvent is then distilled to remove impurities (such as oils, fats or grease) and is returned to the pure solvent tank. The filter often needs to be cleaned of impurities. After washing and extracting, the drying process starts. The printed samples are tumbled in a warm air flow of around 60℃.^12–14 The printed samples were dry cleaned 60 times in total.

Electrical characterization

Determination of amount of ink

The amount of ink on the flexible substrates screen printed with the four inks was determined by weighing the samples before and after the screen printing (Table 3).

Table 3.

Amount of ink for screen printed flexible substrates

Flexible substrates	Ink 1 (DuPont) (g)	Ink 2 (Acheson) (g)	Ink 3 (Sun Chemical) (g)	Ink 4 (Sun Chemical) (g)
Urecom	2.67	2.62	2.22	3.87
PES	1.26	4.13	1.18	4.12

PES: polyester.

Resistance measured with the four-point method

The electroconductive properties of the printed flexible substrates were studied by measuring their DC resistance. At the first stage of the study (before lamination), a four-point method by means of an MR-1 Instrument¹⁵ was used before and after dry cleaning (5, 10, 15 and 20 times). The values of the square resistance R_□ for Urecom before dry cleaning were between 0.005 and 0.030 Ω/□ and after 20 dry cleaning cycles they increased to 0.135–0.280 Ω/□. However, the samples with inks 1 and 4 lost their conductivity completely, as can be seen from Figure 1. Regarding the nonwoven PES, the values of square resistance R_□ before dry cleaning varied from 0.002 to 0.020 Ω/□, slightly lower than those of Urecom. After 20 dry-cleaning cycles, the values increased to 0.050–0.122 Ω/□, still lower than with Urecom. Again, though, as with Urecom, the sample with ink 1 entirely lost its conductivity.

Figure 1.

Square resistance R_□ of printed samples (before and after 20 times dry cleaning). PES: polyester.

Thus, the values of the square resistance R_□ after 20 dry cleaning cycles were very high compared to those before dry cleaning and some samples lost their conductivity entirely (Urecom samples printed with inks 1 and 4 and PES samples printed with ink 1, see Figure 1 and 2).

Figure 2.

Printed samples (a) before and (b) after 20 dry-cleaning cycles (printed with ink 2).

Figure 2 illustrates that the conductive layer has disappeared from the surface of both printed flexible substrates.

Resistance measured with the Van Der Pauw method

In order to avoid the ink disappearing, some kind of pre-treatment may be given to the flexible substrate surface. However, in this study we chose to laminate the electroconductive layer with a thermoplastic polyurethane (TPU) layer from Epurex (LPT, thickness 80 µm). The lamination temperature was set to 150℃ without the application of any extra pressure.¹⁶ However, covering the printed square with this non-conductive layer meant that the four-point probe method was no longer suitable for evaluating the resistance. Therefore, the DC resistance was measured with the Van Der Pauw method.¹⁷ For this purpose, four electroconductive copper foil contacts were attached to the vertices of each conductive sample and fixed with conductive glue (Figure 3). After the four contacts were glued, the TPU was laminated on the printed area.¹⁸ The current I was supplied, by a power supply PL601, on two contacts 1 and 2 and the resulting voltage V across the contacts 3 and 4 was measured with a voltmeter Digi-tool, digi-16. Then the current I was supplied in another direction through two other contacts 2 and 3 and the voltage V was measured across the contacts 1 and 4. The currents applied for our measurements were 200, 400 and 1000 mA and the square resistance R_□ was then calculated as the average value. The measurements were performed at room temperature.

Figure 3.

Square resistance R_□ of printed samples measured with the Van Der Pauw method.

The R_□ was calculated:

R_{□} = \frac{π V}{\ln 2 I}

where V is the voltage, I is the current and

\frac{π}{\ln 2}

is the correction factor.¹⁹

Results and discussion of laminated printed samples

The square resistance R_□ of the printed samples before and after dry cleaning, with four inks and covered with a TPU layer, is shown in Figure 4.

Figure 4.

Square resistance R_□ of printed samples covered with a thermoplastic polyurethane layer (before and 60 times after dry cleaning). PES: polyester.

Figure 5 shows how the resistance gradually increases after five dry cleaning cycles, up to a maximum of 60 cycles. Sixty dry cleaning cycles is a worst case scenario, because that would mean cleaning the garment every two months during a life cycle of 10 years. A paper by Kalliala and Nousiainen²⁰ also fixes on a maximum of 60 laundries for cotton/polyester bed sheets, and it was this paper that inspired us to take dry cleaning to the same lengths in our research.

Figure 5.

Square resistance R_□ of printed samples measured after every five dry cleaning cycles. PES: polyester.

Again we observe that both substrates printed with ink 1 completely lose their conductivity after dry cleaning. For nonwoven PES this trend was observed after as few as 10 dry cleaning cycles. Figure 6 clearly shows what is happening: the conductive surface has several troughs, as seen from the cross-section. These discontinuities in the conductive layer prevent the current from passing through. They are caused by creases in the fabric during the mechanical movements in the drums, in combination with a low amount of ink on the substrate and bad adhesion of the epoxy resin on the PES substrate. For this reason, this ink cannot be recommended for use in combination with these substrate materials.

Figure 6.

Cross-section of the conductive printed layer on nonwoven polyester (PES) (printed with ink 1). TPU: thermoplastic polyurethane.

Ink 2 gives better values of square resistance R_□ in combination with nonwoven PES than with polyurethane foam. The other two inks, 3 and 4, give the best results for both substrate materials. Furthermore, after dry cleaning the values show almost no difference whether printed on polyurethane foam or on nonwoven PES.

After 60 dry cleaning cycles, Urecom shows the lowest square resistance R_□, 0.129 Ω/□ when printed with ink 4 (with epoxy resin – Table 2) and nonwoven PES shows the lowest square resistance R_□, 0.025 Ω/□ when printed with ink 2 (with PES resin); hence, for these substrates, the inks have better adhesion after 60 cycles.

Conductive ink 4 generally (before and after dry cleaning) shows lower square resistance R_□ than the other inks. This can be explained by the amount of ink (Figure 7) and the value of solid content (84–86% – see Table 2), which is the highest for this ink.

Figure 7.

Square resistance R_□ versus the amount of ink before and after dry cleaning on a double logarithmic scale. PES: polyester.

In Figure 7, trendlines with a negative slope −1 have been added representing the relation R_□ ∼ 1/amount of ink. For Urecom and PES (before and after dry cleaning) the experimental data follow the trendlines.

Before dry cleaning, it was observed that the amount of ink deposited on the substrate surface and the solid content of the ink (Table 2) plays an important role in the conductivity: the higher these values, the lower the square resistance R_□ is. This can be seen for Urecom printed with ink 4 (Figure 7), which has the largest amount of ink (3.87 g) and the ink has the highest percentage of solid content (84-86% – Table 2), resulting in the lowest square resistance R_□. However, the amount of ink and solid content are not enough to explain the conductivity of inks 1, 2 and 3, because the percentage of solid content of these three inks is very similar (around 70%). It is possible, therefore, that the nominal sheet resistance of the inks may affect the resistance of the conductive flexible surfaces. The nominal sheet resistances of the inks, as indicated by the producers, are different: 0.012–0.015 Ω/□ for ink 1, <0.025 Ω/□ for ink 2 and 0.011 Ω/□ for ink 3 at 25 µm dry coating thickness. In the case of ink 3, the smallest amount resulting in the lowest square resistance R_□ can be explained by the lowest nominal sheet resistance. A closer look at inks 2 and 3 reveals that they have different amounts of ink (2.62 and 2.22 g, respectively), which should make it more likely that ink 2 has lower square resistance R_□, but this does not turn out to be the case. Thus, this can only be assigned to the difference in nominal sheet resistance of inks 2 and 3.

The same situation pertains for nonwoven PES, where inks 2 and 4 have almost the same amount of ink (4.13 g for ink 2 and 4.12 g for ink 4) and almost the same nominal sheet resistance as indicated by the producer (<0.025 Ω/□ for both inks at 25 µm dry coating thickness). However, the square resistance R_□ before dry cleaning is different: 0.003 Ω/□ for ink 2 and 0.001 Ω/□ for ink 4. Here, the properties of the inks are different. The solid content is 73.5–76% for ink 2 and 84–86% for ink 4, and the resins are PES for ink 2 and epoxy for ink 4. So, the higher the solid content is, the lower the square resistance R_□ will be.

PES printed with ink 3, the same as Urecom printed with ink 3, has the lowest amount of ink (1.18 g), before dry cleaning, compared with ink 1 (1.26 g) and the lowest square resistance R_□ (before dry cleaning). The percentage of solid content here is approximately the same (around 70%), but the nominal sheet resistance indicated by the producer, 0.012–0.015 Ω/□ for ink 1 and 0.011 Ω/□ for ink 3 at 25 µm dry-coating thickness, are different.

So, as discussed above, the amount of ink, the solid content and the sheet resistance are very important and affect the conductivity level: the thicker the layer, the lower the square resistance.^21,22

After dry cleaning through the 60 cycles, it was observed that the type of resin also has a significant effect on electroconductive flexible substrate. Urecom printed with ink 4, which has epoxy resin, proved to have the lowest square resistance R_□.

This also explains the behavior of the nonwoven PES substrate printed with inks 2 (PES resin) and 4 (epoxy resin). Before dry cleaning, the flexible substrate printed with ink 4 (epoxy resin) showed the lower resistance (0.001 Ω/□ versus 0.010 Ω/□ for ink 2. After 60 dry cleaning cycles, however, the lower resistance was shown by the sample printed with ink 2 (PES resin – 0.025 Ω/□ versus 0.125 Ω/□ for ink 4. Before dry cleaning, the amount of ink was approximately the same (4.12 g for ink 4 and 4.13 g for ink 2), and the nominal sheet resistance indicated by the producer was the same (≤0.025 Ω/□ at 25 µm dry coating thickness). It is thus clear that, as well as the amount of ink and properties of the ink, such as solid content and sheet resistance, the type of resin influences the resistance after dry cleaning.

Conclusions

In this study, we screen printed with four different silver-based inks (two with PES resin and two with Epoxy resin) on polyurethane foam and nonwoven PES. The square resistance R_□ was measured after each five cycles of dry cleaning up to 60 cycles in order to provide an evaluation of the conductivity of the four inks with respect to each substrate. The values of the square resistance R_□ ranged from 0.025 to 2.263 Ω/□.

We can conclude that the silver conductive inks available on the market for electronic applications reach broadly square resistance R_□ values on the flexible substrates. However, it is necessary to cover them with a protective layer (e.g. a polyurethane layer) to maintain electrical properties after dry cleaning. PES-based ink combines best with a PES substrate, while epoxy-based ink gives better results with polyurethane foam, so a good combination between ink and substrate needs to be found in order to obtain a high-quality electroconductive substrate that can be integrated into garments and easily maintained.

Footnotes

Acknowledgements

The authors want to express their acknowledgement to Nicole Hargarter from Epurex Company for supporting this research with providing the thermoplastic polyurethane (TPU) layer, Huys Christian from Vita shop for supporting with the dry cleaning tests and Mr Bob Smith and Mr Frank Eirmbter from SunChemical Company for making the conductive ink available and for their interest in this research.

Funding

I Kazani, on leave from the Polytechnic University of Tirana, Albania, was supported by the BASILEUS project [2008-1799/001 – 001 MUN ECW].

References

Meoli

May-Plumlee

. Interactive electronic textile development: A review of technologies. J Text Apparel Technol Manag 2002; 2: 39–49.

Drean

Schacher

Bauer

. A smart sensor for induced stress measurement in automotive textiles. J Text Inst 2007; 98: 523–531.

Carpi

De Rossi

. Electroactive polymer-based devices for e-textiles in biomedicine. IEEE Trans Inf Technol Biomed 2005; 9: 295–318.

Xue P, Tao X, Leun M-Y, et al. Electromechanical properties of conductive fibres, yarns and fabrics. X. Tao (ed.), 2005, pp. 81–104.

Gasana

Westbroek

Hakuzimana

. Electroconductive textile structures through electroless deposition of polypyrrole and copper at polyaramide surfaces. Surf Coat Technol 2006; 201: 3547–3551.

Merritt

Karaguzel

Kang

T-H

. Electrical characterization of transmission lines on nonwoven textile substrates. Mater Res Soc 2005; 870: 87–95.

Karaguzel

Merritt

Kang

. Flexible, durable printed electrical circuits. J Text Inst 2009; 100: 1–9.

Kim

Yoo

. Electrical charachterisation of screen – printed circuits on the fabric. IEEE Trans Adv Packag 2010; 33: 196–205.

Yang

Chuang

Lou

. Thick-film textile-based amperometric sensors and biosensors. R Soc Chem 2010; 135: 1230–1234.

10.

Merilampi

Laine-Ma

Ruuskanen

. The characterization of electrically conductive silver ink patterns on flexible substrates. Microelectron Reliab 2009; 49: 782–790.

11.

Hertleer

Van Langenhove

Rogier

. Printed textile antennas for off-body communication. Adv Sci Technol 2008; 60: 64–66.

12.

Earnest

. A control technology evaluation of state-of-the art, perchloroethylene dry-cleaning machines. Appl Occup Environ Hyg 2002; 17: 352–359.

13.

TSA. http://www.tsa-uk.org/ (Accessed 6 July 2010).

14.

Lohman

. A history of dry cleaners and sources of solvent releases from dry cleaning equipment. Environ Forensics 2002; 3: 35–58.

15.

MR-1 surface resistance meter, http://www.schuetzmesstechnik.de/de/pdf/Ken-MR1.pdf (accessed 6 July 2010).

16.

Vieroth

Löher Th, Seckel

. Stretchable circuit board technology and application, 2009. IEEE Computer Society. 2009, ISWC 2009, international symposium on wearable computers33–36.

17.

Van Der Pauw

. A method of measuring specific resistivity and Hall effect of disc of arbitrary shape. Philips Res Rep 1958; 13: 1–9.

18.

Banaszczyk

Mey

De, Schwarz

. The Van Der Pauw method for sheet resistance measurements of polypyrrole coated para aramide woven fabrics. J Appl Polym Sci 2010; 117: 2553–2558.

19.

Kalavagunta A and Weller RA. Accurate geometry factor estimation for the four point probe method using COMSOL multiphysics. 2005. COMSOL Multiphysics.

20.

Kalliala

Nousiainen

. Life cycle assessment environmental profile of cotton and polyester-cotton fabrics. AUTEX Res J 1999; 1: 8–20.

21.

Bidoki

McGorman

Lewis

. Inkjet printing of conductive patterns on textile fabrics. AATCC Rev 2005; 5: 11–11.

22.

Locher

. Technologies for system-on-textile, 1: Hartung-Gorre Verlag2006, pp. 165–165.