Sensors and Actuators B 157 (2011) 473–481
Contents lists available at ScienceDirect
Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb
Low temperature gas sensing applications using copier grade transparency sheets as substrates G. Scandurra a , A. Arena a , C. Ciofi a , A. Gambadoro a , F. Barreca b , G. Saitta a , G. Neri c,∗ a b c
Dipartimento di Fisica della Materia e Ingegneria Elettronica, Università di Messina, Italy Advanced and Nano Materials Research s.r.l., Viale F. Stagno d’Alcontres, 31, 98166 Messina, Italy Dipartimento di Chimica Industriale e Ingegneria dei Materiali, Università di Messina, Italy
a r t i c l e
i n f o
Article history: Received 10 November 2010 Received in revised form 28 April 2011 Accepted 3 May 2011 Available online 11 May 2011 Keywords: Flexible sensors Ammonia sensors RH sensors
a b s t r a c t Transparency sheets, coated with copper on both sides by means of thermal evaporation in vacuum, are patterned by direct chemical etching to realize sensing platforms having copper heaters on the backside, and resistances having calibrated temperature coefficient on the topside. The mechanical and thermal stability of these structures was demonstrated up to 70 ◦ C. Bending tests also show that the metallic patterns do maintain unaltered performances after more than 104 bending cycles. Resistance measurements show that the resistance on the patterned copper structures linearly increases with the temperature in the range between room temperature and 70 ◦ C, while above this temperature an irreversible damage occurs. Experimental investigations demonstrate that the heaters on the backside of the sensing platforms allow to obtain a quite uniform temperature distribution on the top side over an area larger than 1 cm2 . Coating the flexible sensing platform by doped polyaniline and carbon nanotubes embedded in a polymer host, a chemoresistive system operating at low temperature is developed, which allows to perform tests at constant temperature, with the temperature being set and monitored by using the heater and the patterned resistance, respectively. The sensing performances of the films are evaluated by means of electrical measurements performed while exposing the samples to different relative humidity levels, and to calibrate ammonia pulses. © 2011 Elsevier B.V. All rights reserved.
1. Introduction During the last few years the rapid expansion experienced by flexible electronics and the unprecedented progresses in materials science have stimulated an increasing research activity focused on the development of novel electronic devices that, compared to the conventional ones, have simplified design, environmentally friendly processing, and inexpensive manufacturing [1,2]. The sensor field has taken advantage of the emerging technologies and materials, as it is evidenced by the advances in applications going from wearable healthcare devices [3] to flexible tactile sensors to be used as artificial skin in robotics [4]. Among the possible benefits of flexible electronics, conventional chemoresistive gas sensors, usually consisting of metal oxide layers deposited on alumina or silicon rigid substrates, can be replaced by lighter and cheaper flexible sensors. Recently, compact, lightweight and low power consumption sensing platforms have been developed by integrating miniaturised capacitive sensors and metal oxides gas sensors on polyimide flexible substrates,
∗ Corresponding author. E-mail address:
[email protected] (G. Neri). 0925-4005/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.snb.2011.05.003
equipped with platinum heaters and temperature sensors [5–7]. The processing technologies used to develop the polyimide sensing platforms require a number of steps including the deposition of layers to promote the adhesion of metal coatings to the plastic substrates, and a number of electron beam metallization followed by lift off. Embedded metallic heaters are developed by spin-coating a layer of soluble polyimide precursor on the top of platinum patterns applied onto the polyimide substrate. The integrated heaters allow the plastic platforms to reach the working temperature of metal oxides, typically higher than 300 ◦ C. However, it is worth mentioning that a number of sensing materials including nanostructured metal oxides [8], conjugated polymers [9], hybrid nanocomposites [10], carbon nanotubes grafted with metal nanoparticles [11], and metal nanoparticles [12], have been found to be able to detect a variety of gases at room temperature. On the other hand, even in the case of sensing materials able to work at low temperatures, substrate temperature control and stabilization are fundamental issues, because the sensing mechanism in itself, the properties of the sensing layers, the sensitivity towards the analyte, and the response/recovery times, are affected by the working temperature. Within such a framework, this paper explores the possibility of developing platforms for low temperature sensing applications by means of simple, environmentally friendly and inexpensive
474
G. Scandurra et al. / Sensors and Actuators B 157 (2011) 473–481
Fig. 1. Conventional lithography mask (a) and resulting patterned copper on the flexible substrate (b); direct etching pattern layouts (c and e) for obtaining the heather and a temperature sensing element on the top and bottom sides of the plastic substrate through local application of a fibre pen filled with FeCl3 aqueous solution. The resulting patterned system is shown in (d), with the top copper layer completely removed, and in (f), where both layers are present and the sample is illuminated from the bottom (the top copper layer is partially removed in order to allow to distinguish both patterns).
processing, using transparency sheets coated with metallic films as substrates. The platforms are required to support an integrated heater, that has the function to settle the substrate temperature slightly above the ambient temperature, and a temperature sensor for temperature monitoring/control. To meet these requirements, transparency sheets, coated on both sides with copper by means of thermal vacuum evaporation, are patterned on the top and
on the bottom sides. The thermal and mechanical stability of the evaporated copper films are investigated by means of resistance measurements vs. temperature, and by monitoring the electrical resistance of samples subjected to bending cycles at constant temperature. The structural modifications suffered by the samples when heated are investigated by means of scanning electron microscope (SEM) and X-ray Photoelectron Spectroscopy (XPS).
G. Scandurra et al. / Sensors and Actuators B 157 (2011) 473–481
α sample bending actuator temperature sensor sample holder heater Fig. 2. Structure of the stress chamber. The heater together with the temperature sensor are employed for maintaining the chamber temperature at 30 ◦ C.
2. Experimental 2.1. Substrates preparation and characterization Copier grade transparency sheets (Tartan TM 950) were used as flexible substrates. Thermogravimetric measurements indicate that the transparency sheets used do not undergo any sensitive weight loss in the range between room temperature and 100 ◦ C. Copper films were deposited by means of thermal evaporation in a vacuum onto copier grade transparency sheets (Tartan TM 950), previously cleaned and washed with ethanol and acetone. The substrates were placed at a distance of about 15 cm from the filament. Sheet resistances in the range between 0.2 / and 0.8 /, depending on the amount of copper used for evaporation, were obtained. The films were patterned either by means of photolithography, or by direct chemical etching, using a conventional pen plotter to provide local application of an etchant ink through fibre tip pens filled with FeCl3 aqueous solutions. This latter approach, useful for rapid prototyping applications that do not require high spatial resolution, is more environmentally friendly and simpler when compared to photolithography, as it allows the used quantity of FeCl3 and wastes to be minimized, and, in addition, it does not require the use of any photoresist layer. Fig. 1a shows the layout of a typical shadow photolithography mask used to develop copper patterns on transparency sheets, as the one shown in Fig. 1b; c and e shows the closed contours that are directly drawn onto the bottom and top sides of the copper metalized substrate by means of a fibre pen having 0.05 mm tip, filled with etchant ink; the results are shown in Fig. 1d (the top copper layer is not present) and in Fig. 1f where both top an bottom layers are present (the substrate is illuminated from the bottom and part of the top copper layer is removed for evidencing the patterns on both layers). The resistance of the investigated samples ranged from about 20 to about 400 depending on the thickness of the starting film and on the width w of the lines. A few samples were employed to investigate the temperature dependence of the resistances using an experimental set-up derived from the one described in [13]. The samples were placed flat and in close contact with the top face of an aluminium cylinder and thermally insulated from the environment in order to insure a uniform temperature. The temperature was made to change within a predefined range in such a way as the TCR (temperature coefficient of resistance) could be extracted. Another set of samples was used to investigate the mechanical stability of the patterned resistances vs. substrate bending. To this purpose we used the system depicted in Fig. 2 consisting of
475
a motorized lever that, by means of a couple of rolling cylinders placed just outside the patterned copper area causes the sample to undergo controlled cyclic bending (±50◦ with respect to the flat configuration). The bending device is enclosed in a thermally controlled chamber with the temperature set to 30 ◦ C (±0.2 ◦ C) in order to distinguish resistance changes due to stress from those arising from temperature fluctuations. The resistance vs. the number of bend cycles is recorded by the system. A series of bending cycles is performed (positive and negative angles) at a rate of 1 cycle per second, then the sample is set back in the rest position and the resistance is measured. In order to investigate the temperature distribution on the top side of the transparency foil when the heather patterned onto the bottom is active, a few samples were paced flat onto a supporting layer of thermally insulating material (the heather facing the supporting layer) and the temperature on the other side (in free air) was recorded by scanning a solid state temperature sensor over the surface. The solid state sensor was a LM35 in TO92 package. The flat surface of the sensor package in contact with the sample has an area of 5 × 5 mm2 , that therefore represents the spatial resolution of the sensing system. The sensor was mounted onto a pen plotter in such a way as to allow automatic scanning and data acquisition. During the scanning and for each programmed position, the sensor was held in contact with the sample surface until thermal equilibrium was reached. The morphological and structural properties of the films were investigated by means of optical and Scanning Electron Microscopy (SEM) and X-ray Photoelectron Spectroscopy (XPS). Scanning electron microscopy analyses were carried out with a JEOL 5600LV electron microscope, equipped with an Oxford EDX microprobe. Micrographs of the samples were recorded with a 10 kV accelerating voltage. The XPS spectra were acquired using a K-Alpha system of Thermo Scientific, equipped with a monochromatic Al K␣ source (1486.6 eV) and operating with an analyzer in CAE mode with a pass energy of 200 and 50 eV for survey and high resolution spectra, respectively. A spot size diameter on the samples of about 400 m was selected. The binding energy shifts were calibrated keeping the C1s position fixed at 285 eV. The chemical composition was determined using the Scofield’s sensitivity factors supplied with the Thermo Avantage analysis software. 2.2. Sensor fabrication and ammonia sensing test Conducting inks, to be used as humidity sensing material, were developed by dispersing Multi Walled Carbon Nanotubes (MWCNTs) (supplied by Bayer) in water, in the presence of poly3,4-ethylenedioxythiophene–poly(styrenesulfonate) (PEDOT:PSS) (supplied by Aldrich). After 5–6 h treatment in a sonication bath, well dispersed, highly conducting and stable for months black inks were obtained. Emeraldine base (EB) and dodecylbenzene sulfonate acid (DBSA, 70% 2-propanol solution), supplied by Aldrich, were used to prepare a doped polyaniline (PANI:DBSA) to be used as ammonia sensitive material. The clear green dispersions obtained following the procedure described in [14] were deposited onto the substrates. After washing with acetone, in order to remove the excess of DBSA, and heating at 50 ◦ C for a few seconds, highly conducting PANI:DBSA films with resistivity comparable with those obtained from commercial dispersions of conducting polyaniline (Aldrich), but with improved stability, and stronger adhesion to flexible substrate, were obtained.A solid state sensor (from Sensirion) capable of measuring and recording both the temperature and the relative humidity was used to monitor the environmental conditions into the measurement chamber. Electrical measurements were performed under stationary relative humidity and under exposure to calibrated ammonia flux by inserting both the flexible ammonia/humidity sensors and the
476
G. Scandurra et al. / Sensors and Actuators B 157 (2011) 473–481
Fig. 3. Schematic view of the experimental set-up used to evaluate the sensing response of the developed devices.
solid state reference sensor into a sealed aluminium measurement chamber. Sensing tests were carried out using the experimental setup schematized in Fig. 3. The relative humidity level inside the measurement chamber was settled and controlled by mixing dry and saturated air by employing a number of digital mass flow meter controllers connected to dry air and to a bubbler held at constant temperature. The measurements performed under dynamic conditions were carried out by exposing the sensor to abrupt changes in relative humidity between the ambient RH and the low RH level produced by fluxing dry air inside the measurement chamber. The response of the sensors to RH changes was evaluated by DC measurements performed by means of an Agilent 34970 multimeter. Expositions to calibrated amounts of ammonia were performed by fluxing a combination of dry air and dry ammonia into the measurement chamber. Ammonia fluxes were generated by a permeation tube provided by fine permeation tubes, calibrated to provide a flux of 14 ppm ammonia when kept at 50 ◦ C. 3. Results and discussion
up to about 70 ◦ C, and R vs. T plots that remain stable during repeated measurements. At higher temperature the R vs. T plot departures from the linear law, suggesting that irreversible processes, likely due to substrate deformation take place. The temperature coefficient of resistance (TCR) ˛ was obtained by fitting the experimental data (Fig. 4) according to the first order polynomial R(T) = R(T0 )[1 + ˛(T − T0 )], with T0 = 25 ◦ C, and resulted of about 2 × 10−3 K−1 in all the examined samples, which is about two times less than the reported value of 3.9 × 10−3 K−1 . This latter discrepancy arises from the dependence of the TCR value of copper films on the film thickness. Narayandas et al. [15] show that the measured TCR decreases by more than one half with respect to those of bulk copper, as the copper film thickness decreases. As the temperature exceeds 70 ◦ C, the copper films undergo irreversible increases of resistance, and the relationship between resistance and temperature becomes no longer linear. To investigate on the effects of thermal treatment, XPS measurements have been performed on films obtained from the same batch and subject to subsequent thermal treatment in air. Thermal treatment does affect the photoelectron spectra, as it can be evidenced by
3.1. Sensor platform development
3.1.1. Thermal stability and ageing effects The results of resistance measurements performed on a variety of copper films deposited on transparency sheets (see layouts of Fig. 1), while subjected to an increase of temperature, have in common a resistance that linearly increases with temperature
83 82 Resistance (Ω)
The sensor platforms are fabricated by using transparency sheets, coated on both sides with patterned copper films by means of thermal vacuum evaporation. First, the thermal stability of the evaporated copper films is investigated by means of resistance measurements vs. temperature. The structural modifications suffered by the samples when heated are investigated by means of SEM–EDX and X-ray Photoelectron Spectroscopy. SEM measurements are used also to investigate on the slow degradation of the electrical conductivity of copper films aged in air under moisture for months. By monitoring the electrical resistance of samples subjected to bending cycles at constant temperature, an evaluation of the mechanical stability of the developed flexible sensor platform is also performed.
81 80 79 78
40
45
50
55
60
65
70
Temperature (°C) Fig. 4. Resistance of a copper film deposited on a transparency sheet vs. temperature. Three curves, almost superimposed, are present in the plot relative to repeated measurements on the same sample. A TCR of about 2 × 10−3 K−1 can be calculated from the plots.
G. Scandurra et al. / Sensors and Actuators B 157 (2011) 473–481 6
1.0
1.0x10
Cu 2p
C 1s 2.0x10
5
1.0x10
6.0x10 Counts
0.8
1.5x10
Normalised Counts
Counts
5
8.0x10
477
275
280
285
290
295
Binding Energy (eV) 5
4.0x10
O 1s 5
2.0x10
500
600
700
800
900
1000
Binding Energy (eV)
0.6 0.4 0.2 0.0 930
932
934
936
938
Binding Energy (eV) Fig. 5. XPS spectra of as deposited copper film (upper black line), and of a copper film subject to 24 h heating in air at 70 ◦ C, followed by 2 h treatment at 100 ◦ C (lower red line). The inset compares the carbon 1 s photoelectron peak of the two samples. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
comparing the results of wide range XPS measurements performed on as deposited copper, and on a copper film after 24 h heating in air at 70 ◦ C, followed by 2 h treatment at 100 ◦ C (Fig. 5). The surface composition derived from the analysis of the XPS spectra reveals the presence of carbon (about 40% in the untreated film), in addition to copper and oxygen. The percentage of detected carbon originating from the plastic substrate increases by about 20% after 24 h heating in air at 70 ◦ C. Two hours annealing at 100 ◦ C of the sample previously treated at 70 ◦ C, makes the percentage of surface carbon increase by approximately 90% with respect to that of the as deposited film. At the same time, according to the XPS spectrum it seems that the Oxygen surface concentration decreases after annealing at 100 ◦ C. Aimed at explaining this latter finding, SEM-EDX analysis were performed on the annealed samples, but any attempt to perform the measurements did fail because of the extreme sensitivity of the film to the electron beam, even at low voltage. Such a kind of problem arose only in samples annealed at 100 ◦ C. In addition, visual inspection of copper films heated in air at 100 ◦ C shows that the samples optical transmittance remarkably increases as a consequence of heating. It can be suggested that thermal processing at 100 ◦ C in air has serious consequences on the structure and elemental composition of the coated films. To obtain further insight into the effects of thermal treatment, Fig. 6 compares the Cu 2p3/2 photoelectron peak recorded on as deposited and on thermally treated samples. The spectra of Fig. 6 show that untreated copper has a weak broad band positioned in the higher binding energy side of the main peak. This latter band is found to slightly grow after 24 h thermal treatment at 70 ◦ C. After 2 h annealing at 100 ◦ C of the sample previously heated for 24 h at 70 ◦ C, the intensity of the band in the high energy side of the main peak has a significant increase, and the photoelectron spectrum shows a slight shift towards higher binding energy. The line shape modification of the Cu 2p3/2 photoelectron peak suggests that thermal treatment at temperature higher than 70 ◦ C has an effect on the oxidation state of copper. These findings can explain the irreversible behaviour of electrical resistance vs. temperature evidenced by the TCR measurements above 70 ◦ C. Moreover, the surface carbon enrichment evidenced by XPS and SEM observations carried out suggests that the degradation of thermal treated samples can be also related to a reduced thickness of the Cu film. Speculatively, it can be hypothesized that the surface of the Cu was allowed to oxidize upon exposure to air, producing a thin Cu2 O surface layer. The presence of humidity causes the formation of Cu(OH)x phases, permitting the solvation of Cu, i.e. the formation
Fig. 6. Cu 2p3/2 photoelectron peak of as deposited copper film (solid line), of a copper film after 24 h heating in air at 70 ◦ C (dotted line), and of a copper film after 24 h heating in air at 70 ◦ C, followed by 2 h at 100 ◦ C.
of Cu+ (aq) ions in solution. This interpretation is however to be proved and work is in progress in order to better understand this phenomenon. As far as the stability at room temperature is concerned, resistance measurements performed on samples exposed to air moisture for weeks, indicate that the electrical resistance of the untreated copper films deposited on transparency sheets has a trend to slowly increase with ageing. A detailed SEM analysis carried out on as deposited copper films and on films aged in air for months suggests that this behaviour is related to relevant film surface modification (Fig. 7a and b). Indeed, while as prepared Cu film shows an homogeneous surface structure without any structural imperfection even at high magnification, the samples aged in air for months have irregular surface morphology, arising from massive formation of microcracks. Due to the small thickness of the Cu films, microcracks expose the underlying layer of the substrate support, as confirmed by presence of carbon in the EDX spectra (not shown). The progressive film degradation can be therefore attributed to a corrosion process occurring in humid atmosphere, in according with previous reports [16]. It is found that ageing effects can be prevented by spraying commercially available water resistant polyacrylic layers on the top of the deposited copper films. 3.1.2. Thermal characteristics Thermal maps on the top side of the double layer system with the heather in operation are obtained as described in the previous paragraph. A typical temperature map over an area coinciding with the surface overlying the heater is reported in Fig. 8. The image is the result of the scan on a 32 × 32 points grid. With a power of about 250 mW, the top surface above the heather can reach a temperature of about 10 ◦ C above ambient temperature. Within an area of about 3 cm2 near the centre a good uniformity is obtained as the temperature change is within 0.5 ◦ C. It is conceivable, however, that a proper design of the heather layout can help in obtaining a better uniformity of the entire area above the heather. 3.1.3. Mechanical stability The mechanical stability of the copper films thermally evaporated on transparency sheets has been tested by evaluating the evolution of the electrical resistance of the films while subjected to cyclic bending. Fig. 9 shows the electrical resistance of a typical copper film recorded at 30 ◦ C, as a function of the number of bending cycles. According to Fig. 9 sensitive changes of the measured electrical resistance are observable after more than 104 bending
478
G. Scandurra et al. / Sensors and Actuators B 157 (2011) 473–481
R/R0 (%)
0,2
0,1
0,0
0
1x10
5
2x10
5
Bending Cycles Fig. 9. Percentage change of the resistance vs. bending cycles (R0 = 385 ).
stability. Tests performed by adhesive tapes show that the copper films have good adhesion to their flexible substrates, as on removal of the tape they are not detached from the substrates. 3.2. Sensor development and sensing tests
Fig. 7. SEM analysis of the Cu film surface: (a) as prepared Cu film and (b) Cu film aged in air for months.
cycles. Detailed analysis of the experimental results shows that the way electrical resistance evolves in response to substrate bending, critically depends on the copper films thickness. However, on the basis of measurements performed on a variety of samples, it can be inferred that the films resistance has satisfactory mechanical
ΔT (C ) O
y (mm)
20
9,0 8,5 8,0 7,5 7,0 6,5 6,0 5,5 5,0
10
0
10
20
x (mm) Fig. 8. Temperature map (T with respect to ambient temperature) over the surface of the plastic substrate with the heather of Fig. 1a (at the bottom) supplied with 250 mW. An area of about 3 cm2 is obtained at the centre of the substrate with a temperature uniformity better than 0.5 ◦ C.
Once the of transparency sheets substrates having heaters on the back and calibrated resistance on the top has been prepared and passivated by applying the polyacrylic protective layers on the copper patterns, the sensing platforms are completed by applying humidity and ammonia sensing materials on the top of the substrates. The composite used as a humidity sensing material is a stable conducting ink containing MWCNTs well above percolation threshold [17], dispersed with the aid of aqueous suspensions of conducting PEDOT:PSS. Bare MWCNTs [18], and MWCNTs dispersed in insulating polymer hosts, have been successfully used to develop chemiresistive relative humidity sensors [18,19]. In the case of bare MWCNTs, it has been found that while environmentally absorbed water does not affect the behaviour of individual nanotubes, reduction of the electrical conductivity of the carbon nanotubes network arises from the presence of water absorbed at the nanotubes interfaces [17]. The performance of RH sensors based on MWCNTs dispersions, and the way such sensors respond to RH changes, are strictly related to the carbon nanotubes concentration. In particular, according to a model developed in [19], at concentrations well below the percolation threshold, the sensing mechanism is dominated by interaction between environmental water and the polymer host. When MWCNTs concentration approaches the percolation threshold, in addition to electrical resistance arising from the polymer host, carbon nanotubes also do contribute to the electrical resistivity of the composites, both individually and as a network. At concentrations higher than the percolation threshold the response of the composites towards RH is dominated by the way carbon nanotubes interact with water, and by the way the carbon nanotubes network is affected by the swelling of the polymer host [20]. As far as it concerns PEDOT:PSS, the medium used by us as a host for MWCNTs, it is a hole conducting polymer consisting of grains of oxidized PEDOT, dispersed in the nearly insulating PSS matrix which provides the SO3 − groups charge, required in order to achieve the electrical neutrality [21]. Charge transport is due to the motion of positively charged defects through percolative paths joining the PEDOT+ grains. PEDOT:PSS exists in a variety of more or less conducting formulations, depending on the ratio between PEDOT and PSS. The way the charge transport properties of PEDOT:PSS changes with RH is the result of a compromise between a number of effects [22,23], including morphology changes due
G. Scandurra et al. / Sensors and Actuators B 157 (2011) 473–481
(a)
Resistance ( )
3450
3400
3350 25
30
35
40
45
50
Temperature (°C)
Resistance ( )
3800
(b)
3700
3600
3500 0
5
10
15
20
Time (minutes) Fig. 10. (a) Temperature dependence of the resistance of a thin MWCNT/PEDOT:PSS film, measured at 15% RH; and (b) resistance changes of a thin MWCNT/PEDOT:PSS film kept at 25 ◦ C, from the baseline values (at 15% RH), in response to abrupt RH changes obtained by fluxing humidity saturated air pulses inside the measurement chamber for a few seconds.
Films of PANI:DBSA are prepared starting from suspensions of infusible emeraldine base (EB), obtained by using DBSA both as a dopant and as a surfactant to aid dispersion. Such films have been shown to have electrical resistivity that decreases with increasing RH [14]. In addition, measurements performed at constant temperature, and at 40% RH, have shown that the electrical resistance of PANI:DBSA films exposed to ammonia increase linearly with the NH3 concentration in the range between a few ppm and a few tens ppm. Fig. 12 shows the results of electrical measurements performed at 32 ◦ C, at 17% RH, by exposing a typical PANI:DBSA film prepared as it is described in [14] to 120 s lasting pulses of 5 ppm ammonia. It can be noticed that the resistance change in response to the NH3 pulses has a repetitive shape, with average recovery time of the order of 30 min. The discrepancy between this latter values, and the 15 min recovery time reported in [14], likely originates
3800
Resistance ( Ω)
to polymer swelling, electrical conductivity changes due to polarization effects arising from absorbed water, perturbation of the acid/base equilibrium, increased ionic conductivity arising from the dissolution into absorbed water of PSS and charged species originating from the environment. As a matter of fact, the response towards RH is influenced also by the PSS to PEDOT ratio, and by the films thickness. The electrical resistance measured at constant temperature on freshly deposited PEDOT:PSS films prepared by us from a commercially available highly conducting formulation, increases as the relative humidity level decreases. Unfortunately the PEDOT:PSS films developed by us cannot be used as humidity sensitive layers because of their lack of stability, evidenced by a progressive degradation of charge transport properties with ageing in air. On the contrary, the electrical conductivity of conducting inks obtained by dispersing MWCNT in PEDOT:PSS, at concentrations above the carbon nanotubes percolation threshold are found to be stable over months. Current–voltage measurements carried out at room temperature, in the presence of moisture, indicate that it follows the Ohm law, without any hysteresis, i.e. the electrical behaviour of the sample is mainly resistive, and electronic conduction predominates over the ionic one that occurs due to the presence of physisorbed water. So, though detailed information concerning the charge transport properties are obtainable by performing impedance spectroscopy measurements, being the electrical behaviour mainly resistive, quantitative indications about the way the sample respond to RH changes can be derived in this case by using DC measurements only. The MWCNT:PEDOT composite used to detect RH changes, unlike PEDOT, is found to be highly stable with ageing in the presence of moisture and, according to electrical measurements performed on a large number of thicker and thinner samples, shows reversible resistance changes in response to RH changes. The resistance changes are always found to be linearly related to the RH level, in the explored experimental range. According to the irreversible degradation of the electrical transport properties of PEDOT films with ageing in the presence of moisture, the stable and reversible response to RH changes observed by us is ascribable to the changes of percolative conduction paths through carbon nanotubes network induced by the polymer host swelling. Fig. 10a shows the temperature dependence of the electrical resistance of a thin film consisting of MWCNTs dispersed in PEDOT:PSS, measured by maintaining 15% relative humidity inside the measurement chamber. The plot shows that at constant RH, the film resistance changes by about 3% as the temperature changes between 20 ◦ C and 48 ◦ C. Fig. 10b shows how the sample’s resistance changes in response to abrupt changes of RH. The baseline represents the stationary value of the resistance, as recorded in stationary condition at 15% RH, at 25 ◦ C. When few seconds lasting pulses of humidity saturated air are fluxed inside the measurement chamber, the resistance is found to rapidly increase reaching a maximum. As the humid air flux is stopped, the sample’s resistance decreases, and the baseline value is recovered in a few minutes. The plot of Fig. 10b evidences the good repeatability of the MWCNT/PEDOT:PSS electrical resistance changes in response to rapid RH variations. Fig. 11 shows the electrical resistance of a MWCNT/PEDOT:PSS film measured in stationary conditions at 25 ◦ C, as a function of RH. The RH levels are established inside the measurement chamber by fluxing dry air mixed with humidity saturated air at different ratios, and are measured using a reference sensor. The data are fitted to a first order polynomial, and comparison between the best fit curve and experimental data suggests that the resistance is linearly related to the RH in the whole investigated range. This finding is in agreement with those reported in literature for MWCNTs dispersions above percolation threshold [20].
479
3700
3600
3500
3400 15
20
25
30
35
Relative Humidity (%) Fig. 11. Resistance values of a thin MWCNT/PEDOT:PSS film measured at T = 25 ◦ C, in stationary conditions, at different RH, plotted as a function of the corresponding RH level measured through a reference sensor.
480
G. Scandurra et al. / Sensors and Actuators B 157 (2011) 473–481
4. Conclusions
20
ΔR/R (%)
16 12 8 4 0 0
20
40
60 80 Time (min)
100
120
140
Fig. 12. Relative changes of resistance of a PANI:DBSA film, exposed to 120 s lasting 5 ppm NH3 pulses. Measurements are carried out at 32 ◦ C, at 17% RH.
from the different relative humidity condition: indeed, it has been observed that the higher the environmental RH, the faster unperturbed conditions are recovered. In addition, slower recovery times compared to those previously reported [14], may arise from traces of DBSA, that may be not completely removed from polymer films, especially in case of thicker samples. The results of a sensing test performed at constant temperature on a prototype transparency sheet equipped with a copper spiral to provide heat on the back, and with a copper resistance to be used as temperature sensor on the top, are shown in Fig. 13. MWCNT/PEDOT:PSS, and PANI:DBSA films coated on the top area of the sheet, are used as humidity and ammonia sensitive materials. The measurements are performed supplying power to the heater in order to set a substrate temperature, evaluated by the copper TCR, of 32 ◦ C, and by recording the time evolution of the electrical resistance of the sensing films, while the humidity level inside the measurement chamber is made to change between 12% and 50% by fluxing 500 cm3 /min dry air. The electrical resistances of MWCNT/PEDOT:PSS and PANI:DBSA have opposite behaviour as the relative humidity level changes. At about 34 min from the starting time, under dry air flux, a 120 lasting pulse of 7 ppm ammonia is fluxed into the chamber. As it can be noticed in Fig. 13, the 7 ppm ammonia pulse has sensitive effect on the resistance of PANI:DBSA only.
Fig. 13. Time evolution of the resistances of MWCNT/PEDOT:PSS (red line) and PANI:DBS (black line) deposited on the same transparency sheet, exposed to dry air flux (500 cm3 /min), and to 7 ppm dry ammonia (500 cm3 /min). Measurements are carried out at 32 ◦ C, between 50% and 12% RH. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)
This work explores the possibility to develop flexible sensing platforms, suitable for low temperature gas sensing applications, using inexpensive materials and time and cost saving processing. It is found that simple copier grade transparency sheets, vacuum coated with copper, can be readily patterned using local chemical etching, avoiding time consuming photolithography steps and minimizing hazardous chemicals wastes. The use of the copper patterned transparency sheets, coated with a polyacrylic protective layers in order to prevent copper degradation with ageing, as a platform for low temperature ammonia and relative humidity sensing applications, has been demonstrated. The results presented and discussed in this paper are the first approach to the use of common copier grade transparency sheets as substrates for low temperature sensing applications. The shapes, the geometry and the dimensions of the electrodes, and the size and geometry of serpentine heather are far to be optimized. As far as it concerns overall dimensions, the lower limit that can be reached using the approach described in the paper is conditioned by the spatial resolution of local chemical etching. Such kind of problem can be readily overcome by using conventional photolithography.
References [1] M.A.M. Leenen, V. Arning, H. Thiem, J. Steiger, R. Anselmann, Printable electronics: flexibility for the future, Phys. Status Solidi A 206 (2009) 588–597. [2] J. van den Brand, R. Kusters, M. Barink, A. Dietzel, Flexible embedded circuitry: a novel process for high density, cost effective electronics, Microelectron. Eng. 87 (2010) 1861–1867. [3] V. Casey, B. McAree, M.C. Moloney, P. Grace, Wearable sub-bandage pressure measurement system, in: Sensors Applications Symposium (SAS), 2010, IEEE, 23–25 February, 2010, pp. 41–45. [4] Y.J. Yang, M.Y. Cheng, W.Y. Chang, L.C. Tsao, S.A. Yang, W.P. Shih, F.Y. Chang, S.H. Chang, K.C. Fan, An integrated flexible temperature and tactile sensing array using PI-copper films, Sens. Actuators A 143 (2008) 143–153. [5] D. Briand, S. Colin, J. Courbat, S. Raible, J. Kappler, N.F. de Rooij, Integration of MOX gas sensors on polyimide hotplates, Sens. Actuators B 130 (2008) 430–435. [6] E. Abad, S. Zampolli, S. Marco, A. Scorzoni, B. Mazzolai, A. Juarros, D. Gómez, I. Elmi, G.C. Cardinali, J.M. Gómez, F. Palacio, M. Cicioni, A. Mondini, T. Becker, I. Sayhan, Flexible tag microlab development: gas sensors integration in RFID flexible tags for food logistic, Sens. Actuators B 127 (2007) 2–7. [7] A. Oprea, J. Courbat, N. Bârsan, D. Briand, N.F. de Rooij, U. Weimar, Temperature, humidity and gas sensors integrated on plastic foil for low power applications, Sens. Actuators B 140 (2009) 227–232. [8] L. Xu, B. Dong, Y. Wang, X. Bai, Q. Liu, H. Song, Electrospinning preparation and room temperature gas sensing properties of porous In2 O3 nanotubes and nanowires, Sens. Actuators B 147 (2010) 531–538. [9] J.W. Jeong, Y.D. Lee, Y.M. Kim, Y.W. Park, J.H. Choi, T.H. Park, C.D. Soo, S.M. Won, I.K. Han, B.K. Ju, The response characteristics of a gas sensor based on poly-3hexylithiophene thin-film transistors, Sens. Actuators B 146 (2010) 40–45. [10] N. Van Hieu, N.Q. Dung, P.D. Tam, T. Trung, N.D. Chien, Thin film polypyrrole/SWCNTs nanocomposites-based NH3 sensor operated at room temperature, Sens. Actuators B 140 (2009) 500–507. [11] S. Ju, J.M. Lee, Y. Jung, E. Lee, W. Lee, S.J. Kim, Highly sensitive hydrogen gas sensors using single-walled carbon nanotubes grafted with Pd nanoparticles, Sens. Actuators B 146 (2010) 122–128. [12] L. Wang, J. Luo, J. Yin, H. Zhang, J. Wu, X. Shi, E. Crew, Z. Xu, Q. Rendeng, S. Lu, M. Poliks, B. Sammakia, C.J. Zhong, Flexible chemiresistor sensors: thin film assemblies of nanoparticles on a polyethylene terephthalate substrate, J. Mater. Chem. 20 (2010) 907–915. [13] G. Scandurra, C. Ciofi, C. Pace, F. Speroni, F. Alagi, True constant temperature MTF test system for the characterization of electromigration of thick Cu interconnection lines, Microelectron. Reliab. 42 (2002) 1347–1351. [14] G. Rizzo, A. Arena, N. Donato, M. Latino, G. Saitta, A. Bonavita, G. Neri, Flexible, all-organic ammonia sensor based on dodecylbenzene sulfonic acid-doped polyaniline films, Thin Solid Films 518 (2010) 7124–7127. [15] K. Narayandas, M. Radhakrishnan, C. Balasubramanian, Defect density and electrical properties of vacuum evaporated copper films from annealing studies of electrical resistance, Electrocomp. Sci. Technol. 9 (1982) 171–178. [16] N. Nagai, T. Hironaka, T. Imai, T. Harada, M. Nishimura, R. Mimori, H. Ishida, Study of interaction between polyimide and Cu under a high humidity condition, Appl. Surf. Sci. 171 (2001) 101–105. [17] J. Li, J. Liu, C. Gao, J. Zhang, H. Sun, Influence of MWCNTs doping on the structure and properties of PEDOT:PSS films, Int. J. Photoenergy (2009), doi:10.1155/2009/650509.
G. Scandurra et al. / Sensors and Actuators B 157 (2011) 473–481 [18] J.T.H. Tsai, C.C. Lu, J.G. Li, Fabrication of humidity sensors by multi-walled carbon nanotubes, J. Exp. Nanosci. 5 (2010) 302–309. [19] P.-G. Su, C.-S. Wang, In situ synthesized composite thin films of MWCNTs/PMMA doped with KOH as a resistive humidity sensor, Sens. Actuators B 124 (2007) 303–308. [20] K. Yoo, L.T. Lim, N.K. Min, M.J. Lee, C.J. Lee, C.W. Park, Novel resistive-type humidity sensor based on multiwall carbon nanotube/polyimide composite films, Sens. Actuators B 145 (2010) 120–125. [21] S. Kirchmeyer, K. Reuter, Scientific importance, properties and growing applications of poly(3,4-ethylenedioxythiophene), J. Mater. Chem. 15 (2005) 2077–2088. [22] P. Pacher, A. Lex, S. Eder, G. Trimmel, C. Slugovc, E.J.W. List, E. Zojer, A novel concept for humidity compensated sub-ppm ammonia detection, Sens. Actuators B 145 (2010) 181–184. [23] M. Kus, S. Okur, Electrical characterization of PEDOT:PSS beyond humidity saturation, Sens. Actuators B 143 (2009) 177–181.
Biographies Graziella Scandurra received the Degree in Electronic Engineering and the Ph.D. Degree in information technology from the University of Messina in 2001 and 2005, respectively. She worked with the Dipartimento di Ingegneria dell’Informazione, Pisa, Italy, in the design of dedicated instrumentation. She is currently a Researcher at the Dipartimento di Fisica della Materia e Ingegneria Elettronica, University of Messina. Her current research interests include the design of dedicated and low noise instrumentation and the development and characterization of sensors. Antonella Arena is Associate Professor of Experimental Physics, at the Department of Matter Physics and Electronics Engineering of the University of Messina. Her main research interests are focused on preparation and electrical characterization of conjugated polymers, nanostructured materials, composites, and conducting inks, to be applied in the development of photosensitive devices, sensors, and flexible electronics applications.
481
Carmine Ciofi received the Degree in Electronic Engineering from the University of Pisa, Pisa, Italy in 1989, and the Ph.D. Degree from the Scuola Superiore di Studi Universitari e Perfezionamento S. Anna, Pisa, Italy, in 1993. He joined the Dipartimento di Ingegneria dell’Informazione: Elettronica Informatica e Telecomunicazioni, University of Pisa, where he remained until 1998. He is currently Associate Professor of electronics at the University of Messina, Messina, Italy. His main research interests include the characterization and the reliability of electron devices and the design and realization of dedicated electronic instrumentation. Antonino Gambadoro received the Degree in Electronic Engineering in 2008. He is currently a Ph.D. student at the Dipartimento di Fisica della Materia e Ingegneria Elettronica, University of Messina. His main research interests include the design of control systems for the characterization of electron devices and sensors. Francesco Barreca received the Degree and the Ph.D. Degree in Physics from the University of Messina in 1997 and 2002, respectively. He worked with the Dipartimento di Fisica della Materia e Ingegneria Elettronica, University of Messina, and with the Istituto per i Processi Chimico-Fisici of C.N.R. section of Messina. He is currently a Researcher at the “Advanced and Nano Materials Research srl”, academic spin-off of the University of Messina. His current interests are: Laser Ablation in Liquids of nanostructured materials and study of their structural and optical properties; analysis of the surface chemistry of materials by means of X-ray photoelectron spectroscopy. Gaetano Saitta is Full Professor of Physics, at the Department of Matter Physics and Electronics Engineering of the University of Messina. His research activities include statistical data handling, characterization of optical and electronic properties of semiconductors through spectrophotometry and electron spectroscopies, and development and characterization of photosensitive devices and solid state sensors. Giovanni Neri is Full Professor of Chemistry. He has been Director of the Department of Industrial Chemistry and Materials Engineering of the University of Messina. His research activity, covers many aspects of the synthesis, characterization and chemical–physics of materials with particular emphasis to catalytic and sensing properties. In the latter research area his work has been focused on the preparation of metal oxide thick and thin films, organic–inorganic hybrid nanocomposites, novel coordination complexes and their application in gas sensors.
