Flexible Graphite-Based Humidity Sensor Using Green Technology

The low-cost graphite based pattern on cellulose paper was investigated in the present work. The graphite pattern used was fabricated by using normal inkjet printer on paper substrate that acted as working substrate as well as sensing material for humidity measurements. The quantitative electrical characterizations were measured by using different saturated salt-solutions producing relative humidity (RH) of 15%–92% at room conditions of 27 °C and 40%RH. The developed humidity sensor shows the sensitivity of 0.57 MΩ/%RH in the whole range of 15%–92%RH with a linearity co-efficient of R2 = 0.968, response (τ res) and recovery (τ rec) time of 294 s and 306 s respectively. The hydrophilic nature of the paper substrate is helpful for sensing, though the fabricated sensor is not so viable in terms of sensitivity, repeatability, and reuse but the method was simple, low-cost, bio-degradable, and use & throw which can be used for flexible and green electronics.

In the present scenario, the conventional silicon-based technology is not good enough for the manufacturing of flexible, lowcost, and environment-friendly integrated devices, so the components and devices on flexible substrates have gained increased attention [1][2][3][4][5][6][7][8][9][10][11] for the good mechanical, flexible, and non-planar electronic systems. Flexible substrates like cellulosic paper, textile, plastics, and rubber are more resilient, lesser in weight, low-cost, and give superior flexibility/bendability to absorb or experience stress/ strain than rigid substrates such as silicon and other conventional semiconductors. Micro Electro Mechanical Systems (MEMS) devices developed on substrates like paper, plastics, textile and rubber are flexible, bio-degradable, and foldable, which are essential for mechanical flexibility applications. Flexible substrates such as plastics and polyimide in all possible forms for printed electronics are selected in wide area of research because of the following advantages such as enhanced mechanical strength, high flexibility, good chemical resistance, and high-temperature stability. 12,13 Paper or paper-like substrates are gaining popularity as a potential replacement for plastic in next-generation flexible electronics. In our daily life, paper is by far the most widely used and costeffective material due to bio-degradable, low-cost and flexible in nature. Furthermore unlike traditional flexible substrates, papers combine flexibility and bendability with a variety of other characteristics. 14 MEMS device manufacturing based on the paper has the potential to be very easy and low-cost production techniques; prototyping can be done with simple tools, does not require cleanroom facilities, and potential for mass production (through printing and automatic paper cutting).
In recent years, we have observed an increased interest in the growth of paper based devices and technology for flexible and green electronics. The market or industry of paper based sensors and devices require a small investment in terms of price of the substrate, infrastructure, and expertise as compared to the conventional technology. The use of this kind of technology could easily be applied for real life applications like printed solar panels, printed batteries, sensors and more devices for medical, energy, military and other applications without access to clean-room laboratories.
Hoang-Phuong Phan et al. 15 showed a cantilever type magnetic actuator based on paper by using nanoparticles of ferromagnetic material for actuation and graphite for deflection measurement. The paper based magnetic actuator was fabricated in a clean free environment for the self sensing functions. Vivekananthan Balakrishnan et al. 16 showed a humidity sensor on paper by using graphite and nanoparticles of silver for monitoring of respiration. The group showed that the fabricated sensor has the sensitivity of 0.0564% with longer duration of time for respiration monitoring. Zhang et al. 17 fabricated a humidity sensor having improved sensitivity and response time of 25 s on modified cellulose paper by EPTAC i.e. glycidyl trimethyl ammonium chloride. The group showed that the modified paper was used as substrate as well as humidity sensing material which was used for different applications like breath sensing, non-contact switch and breath monitoring.
Graphite is a material knows for high sensitivity towards humidity. So, we utilized it as one of the sensing material for the present work. 16 Due to the exceptional material graphene, graphite, one of the carbon allotropes, has attracted the attention of researchers. 18 The pencil lead is a mixture of graphite and wax with clay in a small proportion, 19 and the pencil-trace created from it, is slowly making its way into electronic devices due to its distinctive performance, eco-friendliness, and ease of manufacture. [20][21][22] Researchers predict that our e-waste problem will get worse over time, because most sensor/electronics in the market today are designed for portability, not recyclability.
In the present work, graphite based pattern was fabricated on paper for analyzing the humidity effect using green technology which was portable as well as recyclable. The 10B grade pencil was used as sensing material because it consists of high quantity of graphite percentage than other variation. 23 The higher graphitic percentage was having higher conductivity. 24,25 The outline of the pattern was designed, printed, and fabricated by inkjet-printer on a cellulose paper by 10B grade pencil with the hand drawing method. The cellulose paper as well as graphite present in the 10B grade pencil was acting as sensing layer. The effect of humidity was characterized by using salt solution method. The different RH environment was created in a closed glass beakers by saturated salt solution of lithium chloride (LiCl), magnesium chloride (MgCl 2 ), sodium bromide (NaBr), sodium chloride (NaCl), and potassium sulphate (K 2 SO 4 ). The dynamic response of the sensor was characterized at room temperature of 27°C and 40% RH. This work is a step toward green, bio-degradable, and flexible sensors & electronic circuits; an alternative approach to environment -unfriendly traditional electronics. These facile, bio-degradable, and recyclable sensors can contribute to future developments in flexible and eco-friendly alternative electronics.

Experimental
Material and chemicals.-The normal cellulose A4 paper of thickness 88.53 μm and 10B grade pencil was procured from Century Star and Apsara, Hindustan Pencils Pvt. Ltd. The silver z E-mail: mansooriamir5@gmail.com conductive epoxy adhesive (8331-14 G) was purchased from MG Chemicals, USA. Acetone, Lithium Chloride (LiCl), Magnesium Chloride (MgCl 2 ), Sodium Bromide (NaBr), Sodium Chloride (NaCl), and Potassium Sulphate (K 2 SO 4 ) all are of AR grade was used in the experiments. The inkjet-printer used was HP LaserJet P1108.
Fabrication.-The outline of sensor were first designed by AutoCAD 2018 and printed on a normal A4 paper by using LaserJet printer (HP LaserJet P1108). The electrodes of the sensor are filled by hand drawing using 10B grade pencil (Apsara, Hindustan Pencils Pvt. Ltd.) on a normal A4 paper (Century Star). The copper wires were connected using silver conductive epoxy adhesive (8331, MG Chemicals, USA) for the electrical characterization of sensor.
Characterization and measurement.-The elastic property of the substrate was done by universal tensile test with digital image correlation (DIC) method. The surface roughness analysis of the substrate was examined by Atomic Force Microscope (Flex AFM, Nanosurf, Switzerland). Gwyddion 2.51 software was used for data analysis. The analysis of topographic and morphological properties was carried by scanning electron microscopy (FIB-SEM ZEISS Crossbeam 550). The hydrophilic behaviour of water with the cellulose paper was characterised by contact angle goniometer (Ossila, UK). The Raman spectra of the pencil trace was characterised by Raman Spectrometer (EnSpectr R532, Enhanced Spectrometry Inc., USA) having the wavelength of 532 nm. The optical images of the hand drawing sensor were measured by Optical microscope (Axioscope A1, Carl Zeiss Germany). The sensor was characterised in different RH environment using saturated salt solutions of LiCl (15%), MgCl 2 (35%), NaBr (55%), NaCl (73%), K 2 SO 4 (92%) respectively.
The saturated salt solutions were produced in closed glass beaker at room temperature of 27°C with an error of ±3% RH. The resistance of the sensor were measured by SMU (2450 Source Meter, Keithley USA) and the humidity of the different atmospheres was determined by humidity meter (HT 315, Lutron Japan). The HP LaserJet printer was shown in Fig. 1a

Results
Elastic property.-The elastic properties of cellulose paper was found using tensile test with digital image correlation (DIC) method by using 5 KN universal testing machine (UTM) as per ASTM E8 standards. The different samples (5 cm × 0.5 cm) of cellulose paper were tested as shown in Fig. 2. The digital image correlation (DIC) method is a non-contact method to calculate the elastic properties of  substrate. 26 The maximum load, ultimate tensile strength, elongation, poisson's ratio and young's modulus was calculated as 8.08 N, 6.47 MPa, 4.38%, 0.053 and 1039.29 MPa respectively. From the calculated results as shown in Fig. 3, it was observed that the cellulose paper will exhibit maximum load of 8 N, otherwise it will break.
Substrate characterization.-The atomic force measurement of 10 × 10 μm 2 scan area will determine the root-mean-square (RMS) roughness and peak-to-valley height. The AFM measurement was done by Nanosurf Flex AFM system. The AFM was having resonance frequency of 190 kHz and force constant of 48Nm −1 . The measurements are done at room temperature. The analysis of properties was simultaneously done by AFM using small area of scanning probe microscopy raster scan. 27,28 The Figs. 4a and 4b shows the topographic image and 3D image of cellulose paper respectively.
The conventional cellulose paper is extremely irregular having roughness and peak-to-valley height of around 206.43 nm and 1.765 μm respectively. The light and dark region represents the peak and pores of the sample respectively. The Figs. 5a and 5b shows the cross section view at 20 μm and 200 nm whereas Figs. 5c and 5d represents the top view 100 μm and 1 μm for the cellulose paper respectively. The top view of the paper confirms the cellulose fibres present in the paper used in the experiments. A droplet of water was dispensed using a glass syringe with a droplet volume of 3 μl as shown in Fig. 6. The contact angle, wetting energy, and work of adhesion were calculated as 104.81°, 9.08 mN m −1 , and 81.88 mN m −1 respectively.
The Fig. 7 shows the Raman spectrum of the graphite pencil trace on the cellulose paper. It can be observed from the Raman spectra that there are three prominent peaks at 1366, 1574 and 2682 cm −1 which corresponds to D, G and 2D bands of graphite respectively. The Fig. 8a shows the contact pad of the fabricated sensor whereas Figs. 8b-8d represents the dimension of the line and the distance between the two lines. As shown in Fig. 1c the line width is 0.5 mm and the spacing between two consecutive lines is 1.0 mm, but we found that the line width of fabricated sensor is approximately 0.72 mm and spacing is around 0.70 mm by optical images as shown in Figs. 8b-8d which shows the limitation of the hand drawing process. Before taking the SEM images, the samples was coated with Au:Pd in ratio of 80:20 with chamber pressure as 5 × 10 −3 mbar at room temperature. The SEM images of fabricated graphite pattern on paper (GPoP) sensor was clearly visible in the Fig. 9a whereas the Fig. 9b and inset of the Fig. 9b shows the flakes present in the graphite of 10B grade pencil at 10 μm and 200 nm respectively.

Discussion
Humidity characteristics.-The resistance measurements of graphite pattern on paper (GPoP) was characterised in a laboratory setup at room conditions of 27°C and 40% RH. The GPoP sensor  was characterised in the whole range of 15%-92% RH as shown in Fig. 10a. For the humidity sensing characteristics, the GPoP sensor was put in a closed glass vessel of a particular RH for 5 min and then it is put in another vessel of different RH. In this manner the adsorption and desorption curve have been taken with a particular RH. The dynamic response of the sensor is shown in Fig. 10c. The resistance of sensor was increased/decreased depends on n-type or ptype due to the adsorption/desorption of analyte. 29 It was observed that the resistance of the GPoP sensor was increased with the increasing RH due to the presence of hydrophilic functional groups present on graphite surface as it is explained by Alrammouz et al. 30 and Choo et al. 31 The water molecules connect to the hydrophilic functional groups on its surface by removing electrons when they   interact with these groups as a result of which they become less conductive in higher RH. As explained by, 32 the resistance of sensor is increasing with the increased RH because the sensing material will be a p-type material due to which absorption of water molecules will be acting as electron donor, so decrease of hole concentration due to water molecules and due to swelling of bare paper. Sensitivity is an essential parameter for determining the performance of humidity sensor for the practical field applications. The sensitivity of GPoP sensor was found to be 0.57 MΩ/%RH with a linearity of R 2 = 0.968 in a whole range of 15%-92% RH. Hysteresis is another important parameter to determine the reliability of the sensor, which is defined as the maximum difference of sensor characteristics in adsorption and desorption cycles. 31 Though, the GPoP sensor parameters were not good, but the method was simple, low-cost, and bio-degradable. The Fig. 10c represents the hysteresis curve of the GPoP sensor fabricated by hand drawing which shows the large difference in the adsorption and desorption process which is the limitation of the fabrication process.  Response time and recovery time are the two important parameters of the humidity sensor in practical field applications. In our work, response time is defined as the time taken by the sensor for changing the resistance from initial value to final value when the RH changed from 15% to 92%. Similarly, the recovery time is the time taken by the sensor for changing the resistance from final value to 90% of the initial value when the RH changes from 92% to 15%. Figure 10d shows the response and recovery curve for the 6 different cycles of the sensor. The response time was about 294 s whereas the recovery time was 306 s as shown in Fig. 10e. This is because the limitation of the hand drawing process as well as the normal cellulose paper resulting in slow response and recovery times. Accordingly, the specific application may be chosen so that these limitations do not restrict the usages. Table I shows the summary and comparative analysis of graphite based humidity sensor on cellulose paper. It was found that the sensing range and sensitivity was greater of our work as reported earlier. 31

Bending and Stability Study
The bending effect was experimentally studied by putting the GPoP sensor on glass cylinders of radius (R = 4 cm to 2 cm) at 27°C and 40%RH which was shown in Fig. 11. The Fig. 12 shows the percentage changes in sensor resistance with bending at 40%RH. The sensor resistance was increasing with the sensor's radius getting smaller as it bends. It was experimentally observed that the sensor resistance changed by 3.46% as the substrate's bending changed from R = 4 to 2 cm. The sensor resistance was changed largely i.e. >5% as it bends for radius <2 cm.
A set of 5 GPoP sensors (S61-S65) was patterned on a cellulose paper. Their resistance (Rs) was measured by SMU (2450 Source Meter, Keithley USA). The Fig. 13 shows the percentage variation of sensor resistance on cellulose paper at 27°C and 40%RH. It was observed that the minimum and maximum variations are 0.99% to 6.47% in resistive case. However, the overall trend seems to be well within 6%.

Human Blow Response
The graphite pattern on paper (GPoP) sensor was also responding to blow as shown in Fig. 14. The resistance of sensor was increasing from 190 KΩ to 210 KΩ as the humidity around the sensor changes from 30% RH to 75% RH by blowing the sensor in every 15 min interval. The sensor blow response and recovery time was around 2 s and 8 min respectively which can be used for non-contact switch.

Conclusions
In this study, we showed a low-cost and green method for developing a humidity sensor by drawing a graphite pattern on paper (GPoP) for numerous applications in the flexible and green electronics. The developed humidity sensor shows the sensitivity of 0.57 MΩ/%RH in the whole range of 15%-92%RH with a linearity coefficient of R 2 = 0.968, response (τ res ) and recovery (τ rec ) time of 294 s and 306 s respectively. The changes so introduced by using different compositions and processing parameters during surface modifications could possibly be used for tuning the sensor characteristics as per requirement. A more detailed study of the abovementioned aspects would certainly provide appropriate sensors for field applications. The present study of the sensors would certainly help in food packaging industry and medical healthcare for breath rate monitoring.