sensors Article

Humidity Sensing Properties of Paper Substrates and Their Passivation with ZnO Nanoparticles for Sensor Applications Georgios Niarchos 1, *, Georges Dubourg 1 , Georgios Afroudakis 2 , Markos Georgopoulos 2 , Vasiliki Tsouti 2 , Eleni Makarona 2 , Vesna Crnojevic-Bengin 1 and Christos Tsamis 2 1 2

*

Nano and Microelectronics Group, BioSense Institute, 1 Zorana Djindijca, 21000 Novi Sad, Serbia; [email protected] (G.D.); [email protected] (V.C.-B.) Institute of Nanoscience and Nanotechnology, NCSR “Demokritos”, Patriarhou Gregoriou & Neapoleos, 15310 Aghia Paraskevi, Greece; [email protected] (G.A.); [email protected] (M.G.); [email protected] (V.T.); [email protected] (E.M.); [email protected] (C.T.) Correspondence: [email protected]; Tel.: +381-21-485-2137

Academic Editors: Christos Riziotis, Evangelos Hristoforou and Dimitrios Vlachos Received: 5 December 2016; Accepted: 4 February 2017; Published: 4 March 2017

Abstract: In this paper, we investigated the effect of humidity on paper substrates and propose a simple and low-cost method for their passivation using ZnO nanoparticles. To this end, we built paper-based microdevices based on an interdigitated electrode (IDE) configuration by means of a mask-less laser patterning method on simple commercial printing papers. Initial resistive measurements indicate that a paper substrate with a porous surface can be used as a cost-effective, sensitive and disposable humidity sensor in the 20% to 70% relative humidity (RH) range. Successive spin-coated layers of ZnO nanoparticles then, control the effect of humidity. Using this approach, the sensors become passive to relative humidity changes, paving the way to the development of ZnO-based gas sensors on paper substrates insensitive to humidity. Keywords: humidity passivation; paper substrate; ZnO nanoparticles; disposable sensors

1. Introduction Cellulose fiber-based paper has been used for long time in missives, flyers, books or newspapers, because it is both chemically and mechanically stable under atmospheric conditions and absorbs ink readily. In the present-day, paper can be also used as substrate for the fabrication of electronic devices such as antennas, sensors, photovoltaics or energy storage devices [1–4]. Compared to other substrates like silicon or glass, paper as a substrate offers an important number of advantages: the raw materials required for its manufacturing process are very abundant organic materials, it is inexpensive, lightweight, recyclable, disposable and environmentally benign [5]. In addition, recent advancements in printed electronics using electrically functional inks and traditional printing technologies such as flexo, screen, offset and gravure, allow the fabrication of paper-based electronic devices in a simple, large-scale and low-cost manner [6–8]. Paper is a good candidate for gas-sensing applications thanks to its inherent porous structure and rough surface, which provides a large surface area compared to flat glass or a nonporous plastic substrate [9]. Several kinds of gas sensors have already been presented in the literature like ethanol, hydrogen sulfide or nitrogen dioxide sensors [10–12] and are based on metal oxides [13], nanoparticles [14] or polymers [15,16]. Paper is also the perfect candidate for low-cost and disposable humidity sensors due to its capacity to absorb water vapor. However, only a few concrete examples of humidity sensors using Sensors 2017, 17, 516; doi:10.3390/s17030516

www.mdpi.com/journal/sensors

Sensors 2017, 17, 516

2 of 11

paper substrate as the sensing element can be found in the literature. For instance, a fully printed chip-less radio frequency identification sensor for humidity monitoring applications was reported in [17]. Another example presents a capacitive-type humidity sensor using silver interdigitated electrodes printed on paper substrate. This work investigates the effect of water molecule absorption on recycled paper capacitance and conductance [18]. The ability of a paper to absorb moisture largely depends on its manufacturing process, since this process affects the morphology of cellulose fibers on its surface. Therefore, we need to further investigate paper manufacturing methods. Although paper can be a promising candidate for humidity sensing, its use for other gas sensors can be restricted by the same effect, affecting the gas sensor’s selectivity and influencing their performances. Many studies have focused on devising methods for improving the surface hydrophobicity of paper, such as laser ablation to modify the surface morphology and/or change the surface energy [19] and plasma-induced polymerization to create hydrophobic polymer chains on the paper’s surface to make it water repellent [20]. Other interesting examples include the possibility to improve surface hydrophobicity of paper by using a coating of organic or inorganic nanoparticles [21,22]. Moreover, an interesting technique is proposed in [23], which describes a coating process aimed at reducing the effect of humidity on paper-based electronics. Even though this process is effective, it requires a specific type of paper, takes longer, and incurs additional costs. In this context, we propose a simple, effective and low-cost approach to study and reduce the effect of humidity on the electrical properties of simple printing papers with different morphology. As a first step, we fabricated paper-based microdevices by laser ablating metal layers deposited on the surface of the paper substrates and characterized them in detail under known relative humidity levels at room temperature. This enabled us to explore the role of surface morphology in the ability of a particular paper to act as a humidity sensor. As a second step, we examined a time- and cost-efficient methodology—inspired by Nature’s approach to exploit nanoparticle formations for functionality enhancement [24]—as a potential route for passivating paper substrates with respect to humidity. The development of the proposed methodology was partly based on a previous work that employed sol-gel solutions with suspended ZnO nanoparticles in an attempt to control the wetting properties of wood surfaces [25]. ZnO nanostructures [13,26] (nanoparticles, nanowires, nanoflakes, etc.) are exceptionally interesting due to their great potential in many practical applications such as energy harvesting [27], sensors [28] and solar cells [29].In our case, we thoroughly investigated how ZnO nanoparticles affect sensor operation, in an attempt to pave the way for disposable and selective sensors on paper substrates with no impact from humidity. 2. Materials and Methods 2.1. Fabrication of Paper-Based Devices Even though the most popular paper used in sensor technology is the Whatman no.1 chromatography paper (Sigma-Aldrich Chemie Gmbh, Taufkirchen, Germany), this work focuses on two other paper substrates: a plain printing paper of a 80 gm−2 basis weight and a glossy, photographic quality paper of a 200 gm−2 basis weight. Interdigitated electrodes (IDEs) were directly patterned on them using an inexpensive and fast fabrication process, which does not require the high-cost semiconductor manufacturing equipment normally used for silicon microfabrication. The process sequence is schematically illustrated in Figure 1. The papers were cut into 3 × 2 cm2 rectangular pieces (Figure 1a) that were loaded onto a sputtering deposition system and a thin layer of Au/Cr (100 nm/10 nm) was directly deposited on them (Figure 1b) without any pre-treatment, showing good adhesion properties. The resulting layer was directly patterned by laser ablation using a short pulse laser (Nd:YAG-1064 nm, Rofin-Sinar Laser Gmbh, Hamburg, Germany), in order to define the IDEs design (Figure 1c).

Sensors 2017, 17, 516 Sensors 2017, 17, 516

3 of 11 3 of 11

Figure 1. 1. Microdevice Microdevice fabrication fabricationprocess. process.(a) (a)Paper Papersubstrate, substrate,(b) (b)Deposition Depositionof ofAu Aulayer, layer, (c) (c) Laser Laser Figure micromachining ofofthe theinterdigitated interdigitatedelectrodes, electrodes, Spin-coating ZnO-nanoparticle layers, micromachining (d)(d) Spin-coating the the ZnO-nanoparticle layers, (e) (e) Photograph of the finished fabricated devices paper devices coinofof11cent), cent),(f) (f)Scanning Scanning Photograph of the finished fabricated devices onon paper (2 (2 devices onona acoin Electron Microscope Microscope (SEM) (SEM) image image of of the the resulting resulting nanotextured nanotextured ZnO ZnO film. film. Electron

The ablationprocess processwas wasoptimized optimized modification of the current values at a constant The laser laser ablation byby modification of the current values at a constant raster raster speed and selection of appropriate frequencies. The maximum available frequency 65 kHz speed and selection of appropriate frequencies. The maximum available frequency of 65ofkHz and and low raster of 80 were mm/schosen were to chosen obtain well-defined andablation selective a lowa raster speedspeed of 80 mm/s obtaintowell-defined electrodeselectrodes and selective of ablation of the metal layer without destroying the paper as shown in Figure 1e. the metal layer without destroying the paper as shown in Figure 1e. 2.2. 2.2. Preparation Preparation of of ZnO ZnO Nanoparticles Nanoparticles Synthesis Synthesis of of one-dimensional one-dimensional ZnO-based ZnO-based nanostructures nanostructures has has been been achieved achieved with with various various techniques as electrodeposition electrodeposition[30], [30],hydrothermal hydrothermal method [31], sol-gel method pyrolysis techniques such such as method [31], sol-gel method [32],[32], pyrolysis [33] [33] the commercial method sol-gel method, which the most widely and and the commercial method [34]. [34]. The The sol-gel method, which is theismost widely used,used, offersoffers good good homogeneity, low processing temperatures and large-area is low-cost and allows homogeneity, low processing temperatures and large-area coating, coating, is low-cost and allows control of control of thecomposition. solution composition. the solution In ourwork, work, sol-gel solutions were byprepared dissolving zinc(Zn(CH acetate dihydrate In our sol-gel solutions were prepared dissolving by zinc acetate dihydrate 3 COO) 2 .2H2 O, (Zn(CH 3 COO) 2 .2H 2 O, Merck Millipore, Darmstadt, Germany) into analytical grade ethanol (C 2H6O, Merck Millipore, Darmstadt, Germany) into analytical grade ethanol (C2 H6 O, Carlo Erba Reagents Carlo deconcentration Ruil, France) of at a40concentration of 40 mM [25]. The resulting S.A.S.,Erba Val Reagents de Ruil, S.A.S., France)Val at a mM [25]. The resulting suspended ZnO suspended ZnO nanoparticles were deposited on the paper substrates using successive nanoparticles were deposited on the paper substrates using successive spin-coatings (Figure 1d), spin-coatings number of one which wasfor ranged to ten for theeach 40 mM solution. the number of(Figure which1d), wasthe ranged from to ten the 40from mMone solution. After coating step, After each coating step, theat samples were 100°C in fororder 10 min a hotplate in order for the samples were annealed 100 ◦ C for 10annealed min on a at hotplate for on ZnO nanoparticles to bind ZnO nanoparticles to bind together and form a film and to remove any traces of the solvent left together and form a film and to remove any traces of the solvent left inside the paper. This temperature inside the paper. Thisto temperature was selected in the order to avoidand deformation bothFigure the electrodes was selected in order avoid deformation of both electrodes the paper of itself. 1f shows and the paper itself. 1f shows a photograph the fabricated the paper along a photograph of the Figure fabricated microdevice on the of paper along withmicrodevice a magnifiedon Scanning Electron with a magnified Scanning Electron Microscope (SEM) image of the nanotextured ZnO film on the Microscope (SEM) image of the nanotextured ZnO film on the electrodes. electrodes. Prior to the deposition of the nanoparticles, their size was measured using a Zetasizer Nano ZS Prior to the deposition of the nanoparticles, theirsize sizedistribution was measured using a Zetasizer Nano ZS (Malvern Instruments Ltd., Worcestershire, UK) their shown in Figure 2a. The sol-gel (Malvern Instruments Ltd., condensation Worcestershire, their size distribution shown in Figure 2a. The process involves hydrolysis, andUK) polymerization of monomers, growth of particles and sol-gel process After involves hydrolysis, condensation and polymerization monomers, growth of agglomeration. nucleation and growth, the average particle size mayofchange by aging, leading particles and agglomeration. After nucleation and growth, the average particle size may change by to unsystematic agglomeration, formation of large monodisperse nanoparticles, compact or porous aging, leading micrometer-size to unsystematicparticles, agglomeration, of large monodisperse nanoparticles, polycrystalline spheres formation or faceted particles. compact or porous polycrystalline micrometer-size particles, spheres faceted particles. In order to examine whether the successive coating with the ZnO or sol-gel results in a uniform film, In order to examine whether the successive coating with the ZnO sol-gel results in a the uniform a patterned silicon wafer with IDE electrodes was deposited with ZnO nanoparticles under same film, a patterned silicon wafer with IDE electrodes was deposited with ZnO nanoparticles under the same conditions (40 mM ZnO sol-gel with thermal annealing between successive coatings). The SEM

Sensors 2017, 17, 516

4 of 11

Sensors 2017, 17, 516

4 of 11

conditions (40 mM ZnO sol-gel with thermal annealing between successive coatings). The SEM image of the film film (Figure 2b) shows that athat porous film forms, even even on a smooth and solid surface image ofresulting the resulting (Figure 2b) shows a porous film forms, on a smooth and solid like Si. This that for thefor selected spin coating conditions film uniformity and structural surface like Si. indicates This indicates that the selected spin coating conditions film uniformity and morphology on the papers may also depend on surface roughness. structural morphology on the papers may also depend on surface roughness. Size districution of ZnO nanoparticles on ethanol solution 35 30

Number (%)

25 20 15 10 5 0 0

2

4

6

8

10

Particle diameter (nm)

(a)

(b)

Figure 2. (a) SizeSize distribution of suspended ZnO nanoparticles in sol-gel solution andand (b)(b) SEM image Figure 2. (a) distribution of suspended ZnO nanoparticles in sol-gel solution SEM image of aofpatterned with IDEs Si wafer coated with 15 layers of ZnO nanoparticles from a 40 mM sol-gel. a patterned with IDEs Si wafer coated with 15 layers of ZnO nanoparticles from a 40 mM sol-gel.

2.3. Measurement Setup 2.3. Measurement Setup A custom-designed experimental setup was developed in order to perform the measurements A custom-designed experimental setup was developed in order to perform the measurements in in a controlled environment. The setup consists of a sealed Teflon chamber, inside which the sample a controlled environment. The setup consists of a sealed Teflon chamber, inside which the sample is is carefully placed, along with the sensing tip of a portable hydrometer, in order to measure the carefully placed, along with the sensing tip of a portable hydrometer, in order to measure the changes changes in the RH in real-time. Mass-flow controllers and flow meters from Brooks Instruments in the RH in real-time. Mass-flow controllers and flow meters from Brooks Instruments (Hatfield, PA, (Hatfield, PA, USA) were used to determine the actual flow rates and concentration of gases. USA) were used to determine the actual flow rates and concentration of gases. Nitrogen was used as Nitrogen was used as a carrier gas for the water vapors. The gas was driven inside a sealed glass a carrier gas for the water vapors. The gas was driven inside a sealed glass bottle containing water bottle containing water (bubbler), which caused the formation of bubbles and the resulting vapors (bubbler), which caused the formation of bubbles and the resulting vapors were then guided to the were then guided to the Teflon chamber where the device was located. Responses to relative Teflon chamber where the device was located. Responses to relative humidity changes were recorded humidity changes were recorded by measuring the drop in resistance across the IDEs, using an by measuring the drop in resistance across the IDEs, using an amperometer (Keithley, Tektronix, Inc., amperometer (Keithley, Tektronix, Inc., Beaverton, OR, USA) driven by a custom-designed Beaverton, OR, USA) driven by a custom-designed Labview-based interface. All measurements were Labview-based interface. All measurements were performed in room temperature (T = 25 °C). performed in room temperature (T = 25 ◦ C). 3. Results and Discussion 3. Results and Discussion 3.1.3.1. Structural Characteristics of Uncoated Devices Structural Characteristics of Uncoated Devices The effect of of humidity onon paper largely depends onon its its surface morphology and composition, The effect humidity paper largely depends surface morphology and composition, which in in turn depend onon thethe manufacturing process. A structural characterization of of both types of of which turn depend manufacturing process. A structural characterization both types paper is thus required in order to understand paper behavior under controlled humidity levels. paper is thus required in order to understand paper behavior under controlled humidity levels. Energy-dispersive X-ray spectroscopy analyze the thecomposition compositionofofthe Energy-dispersive X-ray spectroscopy(EDX) (EDX)was wasused usedin in order order to to analyze thevarious various papers. Figure 3 is the EDX spectrum of both the plain and glossy paper where the of papers. Figure 3 is the EDX spectrum of both the plain and glossy paper where the presence presence calcium andcan oxygen can be identified. The existence of these is elements is due to calciumofand oxygen clearly beclearly identified. The existence of these elements due to the calcium thecarbonate calcium carbonate (CaCO3) introduced into during the pulp during paper manufacturing order to fill of (CaCO3 ) introduced into the pulp paper manufacturing in order toinfill the pores thethe pores of the cellulose matrixpaper and quality, improveespecially paper quality, especially scattering, ink cellulose matrix and improve light scattering, ink light absorbance and surface absorbance andSimilar surface smoothness. obtained forinthe smoothness. results were also Similar obtainedresults for the were glossyalso paper. However, this glossy case, inpaper. addition However, in this in addition calcium and oxygen, silicon, as well as aluminum peaks were to calcium and case, oxygen, silicon, astowell as aluminum peaks were identified. identified. Although the exact manufacturing conditions of the paper substrates used is not known, manufacturing conditions of the paper is not known,of thethe theAlthough presencethe ofexact Si and Al is not unusual since it can besubstrates attributedused to the specifics presence of Si and Al is not unusual since it can be attributed to the specifics of the manufacturing manufacturing processes for improving paper quality and characteristics. Dry-strength additives, processes for improving paper quality and characteristics. Dry-strength additives, or dry-strengthening agents, are chemicals that improve paper strength under normal conditions. These improve the paper's compression strength, bursting strength, tensile breaking strength, and

Sensors 2017, 17, 516

5 of 11

or dry-strengthening agents, are chemicals that improve paper strength under normal conditions. Sensors 2017, 17, 516 5 of 11 Sensors 2017, 17, 516    paper’s compression strength, bursting strength, tensile breaking strength,5 of 11  These improve the and delamination resistance. resistance. Typical Typical chemicals chemicals used used include include cationic cationic starch starch and and polyacrylamide polyacrylamide (PAM) (PAM) delamination delamination resistance. Typical chemicals used include cationic starch and polyacrylamide (PAM)  derivatives. These These substances substances work work by by binding binding fibers, fibers, often often with with the the aid aid of of aluminum aluminum ions ions in in paper paper derivatives. derivatives. These substances work by binding fibers, often with the aid of aluminum ions in paper  sheet. In during paper fabrication, a fact thatthat could explain the EDX signal. SEM sheet. In addition, addition,SiSiisisused used during paper fabrication, a fact could explain the EDX signal. sheet. In addition, Si is used during paper fabrication, a fact that could explain the EDX signal. SEM  characterization of both paper substrates with patterned IDE determine SEM characterization of both paper substrates with patterned IDEwas wasperformed performedin inorder order to to determine characterization of both paper substrates with patterned IDE was performed in order to determine  the differences differencesin inmorphology, morphology,as asshown shownininFigure Figure4.4. the the differences in morphology, as shown in Figure 4.  700 700

Plain paper 80 gm-2

600

PGlossy lain p ap er 80 200 g m -2 paper gm-2 G lo ssy p ap er 200 g m -2

600

EDX counts EDX counts

500 500

400 400

300 300

O

200

O

200

Ca

Al Si

100 C a

Al

100

Ca

Si

Ca

0 0

0

1

2

0

1

2Energy (keV) 3

3

4

5

4

5

E n erg y (keV )

 

−−2 2 plain Figure3.3.EDX EDXspectra spectraofofthe the8080gm gm plain paper paper and and the the 200 200 gm gm−−22glossy Figure glossypaper. paper. −2 Figure 3. EDX spectra of the 80 gm  plain paper and the 200 gm−2 glossy paper.   

(a)

(b)

 

  Figure  4. SEM images of the surface of the (a) 80 gm−2 plain paper and (b) 200 gm−2 glossy paper,    Figure 4. SEM images of (c) the surface of the (A) 80 gm−2 plain paper and(d) (B) 200 gm−2 glossy paper, (c) and (d) are higher magnifications of each film respectively.  (C) and (D) are higher magnifications of each film respectively. Figure 4. SEM images of the surface of the (a) 80 gm−2 plain paper and (b) 200 gm−2 glossy paper, The  cellulose  fibers magnifications that  form  the  paper  can  be  clearly  identified  in  the  case  of    (c) and (d) are higher of each filmstructure  respectively.

devices on the 80 gm−2 plain paper indicating a rough and porous surface morphology (Figure 4a),  −2 glossy paper exhibits a smoother surface and appears more compact (Figure 4b).  while the 200 gm The cellulose fibers that form the paper structure can be clearly identified in the case of   indicating a rough and porous surface morphology (Figure 4a), devices on the 80 gm−2 plain paper while the 200 gm−2 glossy paper exhibits a smoother surface and appears more compact (Figure 4b).

Sensors 2017, 17, 516

6 of 11

The cellulose fibers that form the paper structure can be clearly identified in the case of devices on the 80 gm−2 plain paper indicating a rough and porous surface morphology (Figure 4A), while the −2 glossy Sensors 17, 516 paper exhibits a smoother surface and appears more compact (Figure 4B). 6 of 11 200 gm2017,

3.2. Humidity Devices 3.2. Humidity Characteristics Characteristics of of Uncoated Uncoated Devices Depending surface structure, structure, paper the inherent inherent ability ability to to absorb absorb water water molecules molecules Depending on on its its surface paper has has the resulting to a change in its electrical properties. Thus, we must consider the effect of humidity on resulting to a change in its electrical properties. Thus, we must consider the effect of humidity on plain plainglossy and glossy papers. and papers. Figure 5a presents the change change in in resistance resistance of ofthe theuncoated uncoated80 80gm gm−−22 paper-based Figure 5a presents the paper-based devices devices under under controlled 70%, where in resistance resistance as the controlled RH RH levels levels ranging ranging from from 20% 20% up up to to 70%, where we we observe observe aa drop drop in as the humidity level increases. increases.Figure Figure5b 5bshows showsthe theresulting resulting sensing response, calculated as the ratio of humidity level sensing response, calculated as the ratio of the −2 paper-based − 2 the electrical resistance (R o /R), as a function of humidity level for 80 and 200 gm electrical resistance (Ro /R), as a function of humidity level for 80 and 200 gm paper-based devices; devices; plain paper aexhibits higher response in to the50% 20%RH to rage 50% RH than the glossy one. the plainthe paper exhibits higher aresponse in the 20% thanrage the glossy one. Humidity response of the plain paper at controlled RH

Sensing responses for both paper substrates

1E10 14

Plain paper 80 gm-2

Plain paper Glossy paper

10

20% Ro\R

Resistance (Ohm)

12

1E9 40%

8 6

60% 4

70% 2 0

1E8 0

20

40

60

Time (min)

(a)

80

100

20

30

40

50

60

70

RH (%)

(b)

Figure 5. (a) (a)Humidity Humidity response at controlled RH levels ranging fromto 20% (b)response sensing Figure 5. response at controlled RH levels ranging from 20% 70%, to (b)70%, sensing response for both asof function for both papers as papers function the RH.of the RH.

Indeed, a porous and rough paper surface is more prone to water molecule absorption than a Indeed, a porous and rough paper surface is more prone to water molecule absorption than a smooth one, since its surface area is larger. In fact, at moderate RH (40%), diffusion of water vapor smooth one, since its surface area is larger. In fact, at moderate RH (40%), diffusion of water vapor further inside the volume of paper occurs through the network of air spaces among the cellulose further inside the volume of paper occurs through the network of air spaces among the cellulose fibers. As water progresses through the stack, an amount is sorbed onto the fibers’ surface, which fibers. As water progresses through the stack, an amount is sorbed onto the fibers’ surface, which results in water to lose its original concentration. The cellulose fibers absorb and desorb water into results in water to lose its original concentration. The cellulose fibers absorb and desorb water into the capillary air space next to them, with little to no transmission of water molecules between them, the capillary air space next to them, with little to no transmission of water molecules between them, therefore any further movement of wateroccurs almost entirely through air spaces. This absorption therefore any further movement of wateroccurs almost entirely through air spaces. This absorption causes the paper resistance, as measured by the IDE electrodes, to decrease. At higher percentages of causes the paper resistance, as measured by the IDE electrodes, to decrease. At higher percentages of RH, water dissociation occurs on the moist cellulose fiber, creating H+ and OH−. Thus, the current RH, water dissociation occurs on the moist cellulose fiber, creating H+ and OH− . Thus, the current can can flow due to ionic conduction decreasing resistance along the IDE electrodes and resulting to a flow due to ionic conduction decreasing resistance along the IDE electrodes and resulting to a high high sensing response for both papers. The dependence of electrical resistance on water diffusion sensing response for both papers. The dependence of electrical resistance on water diffusion through through the porous structure of paper is also indicated by the difference in the response and the porous structure of paper is also indicated by the difference in the response and recovery times recovery times obtained for the two paper substrates. obtained for the two paper substrates. Figure 6a shows the response times of the uncoated paper-based devices, where response time Figure 6a shows the response times of the uncoated paper-based devices, where response time is defined as the time required for the device to reach 90% of its steady state value. We observe that is defined as the time required for the device to reach 90% of its steady state value. We observe that the response time of the plain paper is considerably lower than that of the glossy paper. This can in the response time of the plain paper is considerably lower than that of the glossy paper. This can in part be attributed to differences in structure and porosity between the two papers due to different part be attributed to differences in structure and porosity between the two papers due to different manufacturing processes. manufacturing processes. Generally, a rough and porous surface is more capable of water absorption than a smooth one, which results to a faster response time. The recovery time follows the same trend as the response time with a faster dehumidification for the plain paper up to 60% RH as shown in Figure 6b. The slower recovery time at higher RH could be attributed to the presence of higher partial pressure of water vapor close to the surface of the paper substrate when it recovers through the desorption of water molecules.

Sensors 2017, 17, 516 Sensors 2017, 17, 516

7 of 11 7 of 11

Response times for the uncoated devices

Recovery times for the uncoated devices

40

Plain paper

10

Glossy paper 200 gm-2

Plain paper Glossy paper 8

Time (min)

Time (min)

30

20

6

4

10 2

0 20

30

40

50

RH (%)

(a)

60

70

20

30

40

50

60

70

RH (%)

(b)

Figure 6. 6. (a) (a) Response Response time time and and (b) (b) recovery recovery time time of of the the uncoated uncoated devices devices for for both both plain plain22 (triangles) (triangles) Figure and 200 200 glossy glossy paper paper (circles). (circles). and

The humidity sensing performance of both papers in terms of response/recovery time and Generally, a rough and porous surface is more capable of water absorption than a smooth one, sensitivity are summarized in Table 1 and are compared with previous humidity sensors made on which results to a faster response time. The recovery time follows the same trend as the response time paper substrate. We notice that porous paper even without additional functionalization exhibits a with a faster dehumidification for the plain paper up to 60% RH as shown in Figure 6b. The slower high sensitivity in a range from 0 to 70% level of humidity. Contrary to previous work found in the recovery time at higher RH could be attributed to the presence of higher partial pressure of water vapor literature using paper substrate, in our case no additional material is used to improve sensing close to the surface of the paper substrate when it recovers through the desorption of water molecules. performances. The humidity sensing performance of both papers in terms of response/recovery time and sensitivity summarized in Table 1 and of are compared with previous humidity sensors made on Tableare 1. Humidity sensing performance different sensor devices fabricated on paper substrate. paper substrate. We notice that porous paper even without additional functionalization exhibits a high Sensing Material RH (%) Response/Recovery Reference Year 0 to 70% sensitivity in a range from level of humidity. Minimum Contrary to previous work found in the time literature [35] Polypyrrole 40 418 s/418 s using paper substrate,2011 in our case no additional material is used to improve sensing performances. [36] Carbonnanotube 20 6 s/120 s 2012 Table 1. Humidity sensing performance of different sensor devices fabricated on paper [37] Aluminumoxide 25 / substrate. 2014 [38] PVA 50 / 2015 Reference Year Sensing Material Minimum RH (%) Response/Recovery Time [18] Paper 30 4 min/6 min 2015 [35] 2011 Polypyrrole 40 418 s/418 s This[36] work Porouspaper 20 1 min/2–10 2016 2012 Carbonnanotube 20 6 s/120 smin [37] 2014 Aluminumoxide 25 / min This work Smooth paper 40 10 min/8 2016 [38] 2015 PVA 50 / [18] 2015 Paper 30 4 min/6 min 3.3. Effect ZnO Nanoparticles on the Sensitivity of the Paper-Based Devices This of work 2016 Porouspaper 20 1 min/2–10 min This work 2016 Smooth paper 40 10 min/8 min

The effect of humidity on paper can diminish its potential for use in electronic devices, therefore, this issue needs to be addressed. In order to eliminate the humidity effect, we propose an 3.3. Effectsolution of ZnO Nanoparticles thedeposition Sensitivity of Paper-Based DevicesLayers of ZnO nanoparticles original consisting inonthe ofthe ZnO nanoparticles. wereThe deposited a standard low-cost its spin-coating method to allow devices, the creation of a effect of using humidity on paperand can diminish potential for use in electronic therefore, uniform layer on the entire surface of the paper. this issue needs to be addressed. In order to eliminate the humidity effect, we propose an original For consisting both kinds resistance as theLayers number of successive spinwere coated layers solution in of thepaper, deposition of ZnOdecreases nanoparticles. of ZnO nanoparticles deposited increases. For theand plain paper,spin-coating a 30% reduction from 1.96 MΩ to 1.55 MΩ, using a standard low-cost methodintoresistance, allow the creation of a uniform layer onwas the observed after 10 deposited layers from a 40 mM sol-gel and the glossy paper exhibits a similar entire surface of the paper. behavior, except for the first deposition. deposition a singleofZnO layer, resistance increased For both kinds of paper, resistanceUpon decreases as the of number successive spin coated layers compared to the uncoated paper. Subsequent layer deposition resulted in ever lower resistance, as increases. For the plain paper, a 30% reduction in resistance, from 1.96 MΩ to 1.55 MΩ, was observed can be seen in Figure 7. This different behavior could be attributed to different surface roughness after 10 deposited layers from a 40 mM sol-gel and the glossy paper exhibits a similar behavior, except andthe differences in the manufacturing process the two papers. for first deposition. Upon deposition of abetween single ZnO layer, resistance increased compared to Figure 8 shows the ratio of the electrical resistance (R IDE the uncoated paper. Subsequent layer deposition resulted in evero/R), lowermeasured resistance,between as can bethe seen in −2 and 200 gm−2 paper-based devices as a function of the RH, after each electrodes, of the 80 gm Figure 7. This different behavior could be attributed to different surface roughness and differences in spin-coating of ZnO nanoparticles. Wetwo observe that the device’s response to humidity decreases the manufacturing process between the papers. with the number of coatings. After five deposited layers, the effect of humidity on the paper is

Sensors 2017, 17, 516

8 of 11

Sensors 2017,817, 516 Figure shows

8 of 11 the ratio of the electrical resistance (Ro /R), measured between the IDE electrodes, − 2 200 gm paper-based devices as a function of the RH, after each spin-coating 8 of of 11 considerably decreased. These results reveal that the ZnO layer, at least under the experimental ZnO nanoparticles. We observe that the device’s response to humidity decreases with the number of conditions used thisThese work, acts as a effect passivation humidity prohibiting molecules considerably decreased. results reveal that thelayer ZnO to layer, at least thewater experimental coatings. After fiveindeposited layers, the of humidity on the paper is under considerably decreased. from interacting with the cellulose fibers. All measurements were conducted in fixed time conditions used in this actslayer, as a at passivation to humidity conditions prohibitingused water These results reveal thatwork, the ZnO least underlayer the experimental in molecules thisintervals work, (30 min) for each %RH, where the device’s passivation remain unchanged for the duration from interacting with the cellulose fibers. All measurements were conducted in fixed time intervals acts as a passivation layer to humidity prohibiting water molecules from interacting with the cellulose of each measurement. (30 min) for each %RH, were whereconducted the device’s passivation remain(30unchanged for %RH, the duration of fibers. All measurements in fixed time intervals min) for each where the each measurement. device’s passivation remain unchanged for the duration of each measurement. −2 of the 2017, 80 gm Sensors 17, 516and

Resistance changes vs number of coatings

15

Resistance (MΩ) Resistance (MΩ)

15

Resistance changes vs number of coatings 80 gm-2 grm-2 80 200 gm-2 10 200 grm-2

10

5 5

0 0

5

0

10

Number of ZnO layers

0

5

10

Number of ZnO layers

Figure 7. Resistance changes between the electrodes as a function of the number of the ZnO layers coated both the 80 gm−2 (triangles) andelectrodes the 200 gm papers. Figure Resistance changes between the the electrodes as−2 aa(circles) function of the the number number of of the the ZnO Figure 7. 7. in Resistance changes between as function of ZnO layers layers − 2 (triangles) and the 200 gm −2 (circles) papers. coated in both the 80 gm −2 −2 coated in both the 80 gm (triangles) and the 200 gm (circles) papers. Effect of ZnO layers to electrical resistance for glossy paper devices

Effect of ZnO layers to electrical resistance for plain paper devices

-2

14ZnO layers to electrical resistance Plain paper 80 gm Effect of for plain paper devices 14 12

8

6

6 4

6

4

4 2

4

2

2

0

Ro/R

Ro/R

Ro/R

8

8 6

2 0

20

30

40

50

60

30

40

50

RH (a) (%)

10

70

20

30

60

70

40

50

60

70

RH (%)

0

RH (%)

0 20

Blank Glossy paper 200 gm-2 1 layer 5 layers Blank 10 layers 1 layer 5 layers 10 layers

12 10 10

10 8

Ro/R

Glossy paper 200 gm

14 12

Plain paper 80 gm-2

Blank 1 layer 5 layers Blank 10 layers 1 layer 5 layers 10 layers

12 10

Effect of14 ZnO layers to electrical resistance for glossy paper -2devices

10

20

30

40

50

60

70

RH(b) (%)

(b) coated with 1 (hollow Figure 8. Electrical (a) resistance changes for devices on the (a) 80 gm−2 paper Figure 8. Electrical resistance changes for devices on the (a) 80 gm−2 paper coated with 1 (hollow circles), Electrical 5 (hollowresistance squares) changes and 10 (hollow reverse triangles) layers of ZnO nanoparticles from a −2 paper Figure forreverse devices on the (a) 80 of gm coated with circles), 8. 5 (hollow squares) and 10 (hollow triangles) layers ZnO nanoparticles from1 a(hollow sol-gel −2 sol-gel5of solution 40 mM and as a of function ofon RH and (b) gm Measurements paper. Measurements infrom each −2200 circles), (hollow squares) 10RH (hollow reverse triangles) layers of ZnO nanoparticles aRH solution 40 mMof as a function and (b) 200on gm paper. in each RH were were conducted in fixed time intervals of 30min. −2 sol-gel solution of 40 mM as a function of RH and on (b) 200 gm paper. Measurements in each RH conducted in fixed time intervals of 30min. were conducted in fixed time intervals of 30min.

In the case of glossy paper (Figure 8b) the decrease is more profound than that observed in the In the glossy paper (Figure 8b) the decrease is more profound than that observed in the −2 case of 80 In gm and results in the complete passivation after 10 deposited Thisincan thepaper case of glossy paper (Figure 8b) the decrease of is paper more profound than thatlayers. observed thebe − 2 80 gm paper and results in the complete passivation of paper after 10 deposited layers. This can be the results different fabrication process of the two different papers, as itbe can 80attributed gm−2 papertoand in morphology the completeand passivation of paper after 10 deposited layers. This can attributed to the different morphology and fabrication process of the two different papers, as it can be be seen from SEM images (Figure 4). Indeed, the ZnO nanoparticles are absorbed by the porous attributed to the different morphology and fabrication process of the two different papers, as it can seen from SEM images (Figure 4). Indeed, the ZnO nanoparticles are absorbed by the porous paper making theimages formation of a 4). uniform and layer at theare paper surface difficult, bepaper seen from SEM (Figure Indeed, thesolid ZnOZnO nanoparticles absorbed bymore the porous making the formation of a uniform and solid ZnO layer at the paper surface more difficult, while a while a smooth surface facilitates the formation of a uniform solid thin film resulting in better paper making the formation of a uniform and solid ZnO layer at the paper surface more difficult, smooth surface facilitates the formation of a uniform solid thin film resulting in better passivation passivation to humidity. Even though an important factor for paper-based electronics is the effect while a smooth surface facilitates the formation of a uniform solid thin film resulting in better of to humidity. Even though an important factor for paper-based electronics is the effect of long-term long-termtohumidity, not considered here, sincefactor the aim this work electronics is to develop low-cost and passivation humidity.it Even though an important for of paper-based is the effect of disposable sensors (e.g., for agricultural applications). The long-term stability effect has been studied long-term humidity, it not considered here, since the aim of this work is to develop low-cost and in an indirect way. Allfor fabricated sensors were thoroughly tested instability a period of 6 has months their disposable sensors (e.g., agricultural applications). The long-term effect beensince studied in an indirect way. All fabricated sensors were thoroughly tested in a period of 6 months since their

Sensors 2017, 17, 516

9 of 11

humidity, it not considered here, since the aim of this work is to develop low-cost and disposable sensors (e.g., for agricultural applications). The long-term stability effect has been studied in an indirect way. All fabricated sensors were thoroughly tested in a period of 6 months since their fabrication time and in several cases measurements were repeated after a couple of months, with results showing only little variation. In addition, new fabricated sensors exhibited similar behavior. The two key factors that control sensitivity towards humidity are porosity and surface area. Due to the fibrous structure of the paper, one layer of ZnO nanoparticles is not enough to create a solid thin film on the surface of the paper, so the measured sensing response is similar to that of the uncoated one. As the number of coating layers increases, more ZnO is deposited both on the surface and inside the body of the paper, causing fewer cellulose fibers to come in contact with water molecules, as most of them are blocked by the ZnO nanoparticles. The cellulose fibers gradually become completely covered, the air gaps among them are filled up and this leads to the passivation of the paper against humidity, as evidenced by measuring smaller changes in resistance. 4. Conclusions We have performed a structural characterization of plain and glossy paper made bydifferent manufacturing processes in order to determine the effect of paper morphology on its humidity sensing properties. We also investigated the electrical resistance of both paper types under various humidity levels. This study showed that plain paper with porous and rough surface exhibits higher sensitivity and faster response time compared to a smooth glossy paper, proving the possibility to improve the humidity sensing performance of paper by controlling its surface porosity and roughness without additional sensing materials. Furthermore, we demonstrated a promising method for the passivation of paper substrate towards humidity. This method consists in controlling the effect of humidity by successive spin-coated layers of ZnO nanoparticles on top of the interdigitated electrodes. Tested for both types of paper substrate, plain and glossy paper, the successive coating of the electrodes with ZnO nanoparticles induces a gradient passivation towards relative humidity changes, making the paper impervious to moisture. With this approach, total passivation of glossy paper has been achieved for relative humidity ranging from 20% to 70%. Therefore, the fabricated paper-based devices have the potential to be used for low-cost and disposable humidity sensors, and for designing selective paper-based chemical sensors insensitive to humidity. Future works will aim to use ZnO nanostructured layers for the detection of volatile organic components (VOC) like ethanol. Acknowledgments: The authors would like to thank Vladimir V. Srdi´c and Jovana Stanojev for measurements of the nanoparticles’ size. This work has been funded by FP7-REGPOT INNOSENSE GA No. 316191 (Reinforcement of BioSense Center—ICT for Sustainability and Eco-Innovation). Author Contributions: G.N. and G.D. conceived the idea and designed the experiments; G.A., M.G. and V.T. performed the experiments; E.M. contributed in material preparation (ZnO sol-gel); G.N., G.D. and C.T. analyzed the data; V.C.-B. and C.T. coordinated the research; G.N. wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4.

Huang, G.W.; Feng, Q.P.; Xiao, H.M.; Li, N.; Fu, S.Y. Rapid Laser Printing of Paper-based Multilayer Circuits. ACS Nano 2016, 10, 8895–8903. [CrossRef] [PubMed] Hübler, A.; Trnovec, B.; Zillger, T.; Ali, M.; Wetzold, N.; Mingebach, M.; Wagenpfahl, A.; Deibel, C.; Dyakono, V. Printed Paper Photovoltaic Cells. Adv. Mater. 2011, 1, 1018–1022. [CrossRef] Liana, D.D.; Raguse, B.; Gooding, J.J.; Chow, E. Recent advances in paper-based sensors. Sensors 2012, 12, 11505–11526. [CrossRef] [PubMed] Nguyen, T.H.; Fraiwan, A.; Choi, S. Paper-based batteries: A review. Biosens. Bioelectron. 2014, 54, 640–649. [CrossRef] [PubMed]

Sensors 2017, 17, 516

5. 6. 7. 8. 9.

10. 11. 12. 13. 14. 15.

16.

17. 18. 19. 20. 21.

22.

23. 24. 25. 26.

27.

10 of 11

Zheng, G.; Cui, Y.; Karabulut, E.; Wågberg, L.; Zhu, H.; Hu, L. Nanostructured paper for flexible energy and electronic devices. MRS Bull. 2013, 38, 320–325. [CrossRef] Sarfraz, J.; Ihalainen, P.; Määttänen, A.; Peltonen, J.; Lindén, M. Printed hydrogen sulfide gas sensor on paper substrate based on polyaniline composite. Thin Solid Films 2013, 534, 621–628. [CrossRef] Hardman, D.J.; Slater, J.H.; Reid, A.G.; Lang, W.K.; Jackson, J.R. Biochemical and Immunochemical Assay Device. U.S. Patent 6,573,108, 3 June 2003. Hossain, S.M.; Brennan, J.D. β-Galactosidase-based colorimetric paper sensor for determination of heavy metals. Anal. Chem. 2011, 83, 8772–8778. [CrossRef] [PubMed] Yang, J.; Vak, D.; Clark, N.; Subbiah, J.; Wong, W.W.H.; Jones, D.J.; Watkins, S.E.; Wilson, G. Organic photovoltaic modules fabricated by an industrial gravure printing proofer. Sol. Energy Mater. Sol. C 2013, 109, 47–55. [CrossRef] Arena, A.; Donato, N.; Saitta, G.; Bonavita, A.; Rizzo, G.; Neri, G. Flexible ethanol sensors on glossy paper substrates operating at room temperature. Sens. Actuators B Chem. 2010, 145, 488–494. [CrossRef] Sarfraz, J.; Määttänen, A.; Ihalainen, P.; Keppeler, M.; Lindén, M.; Peltonen, J. Printed copper acetate based H2 S sensor on paper substrate. Sens. Actuators B Chem. 2012, 173, 868–873. [CrossRef] Liu, H.; Li, M.; Voznyy, O.; Hu, L.; Fu, Q.; Zhou, D. Physically flexible, rapid-response gas sensor based on colloidal quantum dot solids. Adv. Mater. 2014, 26, 2718–2724. [CrossRef] [PubMed] Comini, E.; Sberveglieri, G. Metal oxide nanowires as chemical sensors. Mater. Today 2010, 13, 36–44. [CrossRef] Segev-Bar, M.; Haick, H. Flexible Sensors Based on Nanoparticles. ACS Nano 2013, 7, 8366–8378. [CrossRef] [PubMed] Bearzotti, A.; Macagnano, A.; Pantalei, S.; Zampetti, E.; Venditti, I.; Fratoddi, I.; Russo, M.V. Alcohol vapor sensory properties of nanostructured conjugated polymers. J. Phys. Condes. Matter 2008, 20, 474207–474212. [CrossRef] Fratoddi, I.; Bearzotti, A.; Venditti, I.; Cametti, C.; Russo, M.V. Role of nanostructured polymers on the improvement of electrical response-based relative humidity sensors. Sens. Actuators B Chem. 2016, 225, 96–108. [CrossRef] Mraovi´c, M.; Muck, T.; Pivar, M.; Trontelj, J.; Pleteršek, A. Humidity Sensors Printed on Recycled Paper and Cardboard. Sensors 2014, 14, 13628–13643. [CrossRef] [PubMed] Feng, Y.; Xie, L.; Chen, Q.; Zheng, L.R. Low-cost printed chipless RFID humidity sensor tag for intelligent packaging. IEEE Sens. J. 2015, 15, 3201–3208. [CrossRef] Chitnis, G.; Ziaie, B. Waterproof Active Paper via Laser Surface Micropatterning of Magnetic Nanoparticles. ACS Appl. Mater. Interfaces 2012, 4, 4435–4439. [CrossRef] [PubMed] Song, Z.; Tang, J.; Li, J.; Xiao, H. Plasma-induced polymerization for enhancing paper hydrophobicity. Carnohydr. Polym. 2013, 92, 928–933. [CrossRef] [PubMed] Ogihara, H.; Xie, J.; Okagaki, J.; Saji, T. Simple method for preparing superhydrophobic paper: spray-deposited hydrophobic silica nanoparticle coatings exhibit high water-repellency and transparency. Langmuir 2012, 28, 4605–4608. [CrossRef] [PubMed] Stanssens, D.; Van den Abbeele, H.; Vonck, L.; Schoukens, G.; Deconinck, M.; Samyn, P. Creating water-repellent and super-hydrophobic cellulose substrates by deposition of organic nanoparticle. Mater. Lett. 2011, 65, 1781–1784. [CrossRef] Bollstrom, R.; Pettersson, F.; Dolietis, P.; Preston, J.; Osterbacka, R.; Toivakka, M. Impact of humidity on functionality of on-paper printed electronics. Nanotechnology 2014, 25, 094003. [CrossRef] [PubMed] Carmona-Ribeiro, A.M. Biomimetic nanoparticles: Preparation, characterization and biomedical applications. Int. J. Nanomed. 2010, 5, 249–259. [CrossRef] Koutzagioti, C.; Ntalos, G.; Tsamis, C.; Makarona, E. Low-cost multi-functional bioinspired wood coatings based on ZnO nanostructures. In Proceedings of the E-MRS 2015 Spring Meeting, Lille, France, 11–15 May 2015. Makarona, E.; Skoulikidou, M.C.; Kyrasta, T.; Smyrnakis, A.; Zeniou, A.; Gogolides, E.; Tsamis, C. Controllable fabrication of bioinspired three-dimensional ZnO/Si nanoarchitectures. Mater. Lett. 2015, 142, 211–216. [CrossRef] Niarchos, G. Design, Fabrication and Characterization of a Vibrational-Driven Piezoelectric Microgenerator. Ph.D. Thesis, National and Kapodistrian University of Athens, Athens, Greece, 2013.

Sensors 2017, 17, 516

28.

29. 30. 31.

32. 33. 34. 35. 36. 37. 38.

11 of 11

Gurav, K.V.; Gang, M.G.; Shin, S.W.; Patil, U.M.; Desmukh, P.R.; Agawane, G.L.; Suryawashi, M.P.; Pawar, S.M.; Patil, P.S.; Lokhande, C.D.; et al. Gas sensing properties of hydrothermally grown ZnOnanorods with different aspect ratios. Sens. Actuators B Chem. 2014, 190, 439–445. [CrossRef] Venditti, I.; Barbero, N.; Russo, M.V.; Di Carlo, A.; Decker, F.; Fratoddi, I.; Barolo, C.; Dini, D. Electrodeposited ZnO with squaraine sensitizers as photoactive anode of DSCs. Mater. Res. Express 2014, 1, 015040. [CrossRef] Leprince-Wang, Y.; Yacoubi-Ouslim, A.; Wang, G.Y. Structure study of electrodeposited ZnO nanowires. Microelectron. J. 2005, 36, 625–628. [CrossRef] Guo, M.; Diao, P.; Wang, X.; Cai, S. The effect of hydrothermal growth temperature on preparation and photoelectrochemical performance of ZnO nanorod array films. J. Solid State Chem. 2005, 178, 3210–3215. [CrossRef] Oskam, G. Metal oxide nanoparticles: Synthesis, characterization and application. J. Sol-Gel Sci. Technol. 2006, 37, 161–164. [CrossRef] Hosono, E.; Fujihara, S.; Honma, I.; Zhou, H. The fabrication of an upright-standing zinc oxide nanosheet for use in dye-sensitized solar cells. Adv. Mater. 2005, 17, 2091–2094. [CrossRef] Saito, M.; Fujihara, S. Large photocurrent generation in dye-sensitized ZnO solar cells. Energy Environ. Sci. 2008, 1, 280–283. [CrossRef] Mahadeva, S.K.; Yun, S.; Kim, J. Flexible humidity and temperature sensor based on cellulose– polypyrrolenanocomposite. Sens. Actuators A Phys. 2011, 165, 194–199. [CrossRef] Han, J.W.; Kim, B.; Li, J.; Meyyappan, M. Carbon Nanotube Based Humidity Sensor on Cellulose Paper. J. Phys. Chem. C 2012, 116, 22094–22097. [CrossRef] Balde, M.; Vena, A.; Sorli, B. Fabrication of porous anodic aluminium oxide layers on paper for humidity sensors. Sens. Actuators B Chem. 2015, 220, 829–839. [CrossRef] Amin, E.M.; Bhuiyan, M.S.; Karmakar, N.C. Development of a low cost printable chipless RFID humidity sensor. IEEE Sens. J. 2014, 14, 140–149. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).