ARTICLE pubs.acs.org/JPCC
Paper-Based Photoconductive Infrared Sensor Alejandro J. Gimenez,† J. M. Ya~nez-Limon,† and Jorge M. Seminario*,‡ †
Centro de Investigacion y Estudios Avanzados del Instituto Politecnico Nacional Unidad Queretaro, Libramiento Norponiente No. 2000, Fracc. Real de Juriquilla CP 76230 Queretaro Qro., Mexico ‡ Department of Chemical Engineering, Department of Electrical and Computer Engineering Materials Science and Engineering Program, Texas A&M University, College Station, Texas 77843, United States ABSTRACT:
We report sensitivity to infrared radiation from a simple paper-made device, which increases its conductivity when exposed to hot objects. We propose that the conductivity of this device is due to ionic currents involving electrolyte salts dissolved in thin film of liquid dispersed over the paper surface, thus the current increases because of heating caused by absorption of infrared light from hot sources. The fast response to stimulus exposure of this sensor suggests that the heating effect is related to a radiative interaction rather than to another kind of heat transfer such as convection or conduction.
1. INTRODUCTION Many products used commonly in several applications are made of cellulose; this is because cellulose is an abundant, low cost, and environment friendly material.1 Thus it is a good idea to employ it in several applications. It is been reported recently that paper and cellulose are used in sensing applications; our group reported the fabrication of UV photoconductive sensors mixing paper with ZnO crystals.2 Other groups are fabricating chemical sensors on paper for medical diagnosis, taking advantage of the hydrophilic properties of cellulose.3,4 In this work we analyze cellulose-based devices to sense electromagnetic radiation in the mid-infrared range. This is relevant because mid- and far-IR ranges are difficult to detect mainly because its wavelengths are not high enough energetically to generate electron transitions detectable in semiconductor devices and the wavelengths are too short to be processed by antennas and amplifiers as RF signals. However, sensing radiation in the mid- and far-infrared range is important because bodies close to room temperature emit electromagnetic radiation in this range as stated by the Planck’s law on blackbody radiation.5 One direct use of this kind of sensor would be to detect people, animals, or any type of body having a temperature above the background. Another potential application of devices sensitive to mid-IR is to sense specific molecules by recognition of their IR signatures. r 2011 American Chemical Society
The approach normally used to fabricate sensors in the midinfrared part of the spectrum is to use pyroelectric materials.6 Pyroelectricity in materials is related to the crystal structure and is present in crystals without a center of symmetry; some studies demonstrate that cellulose as other polymers feature piezoelectric7 and pyroelectric8 effects due to their asymmetry. In the case of our device, it is not likely that the cellulose pyroelectricity is the origin of its sensitivity because the pyroelectric effect has been observed only in specially aligned grown films, being sensitive in vacuum conditions; however our device uses paper as it is and works at room conditions. The experiments we report in this work suggest that the electrical conductivity of our IR sensor devices comes from an ionic current involving electrolyte salts dissolved in a thin liquid layer dispersed over the paper. It is well-know that electrolyte solution conductivities are strongly dependent on temperature,9 so it appears that the sensor paper substrate efficiently absorbs the IR radiation and turns it into heat that is translated to a higher conductivity of the device. Figure 1 shows a schematic representation that explains the conductivity and sensitivity observed in paper devices. Received: July 4, 2011 Revised: August 20, 2011 Published: August 25, 2011 18829
dx.doi.org/10.1021/jp206287f | J. Phys. Chem. C 2011, 115, 18829–18834
The Journal of Physical Chemistry C
Figure 1. Proposed mechanism explaining the conductivity and sensitivity observed in paper devices. As cellulose is very hygroscopic, it tends to form thin films of water on its surface. The presence of ions on this film makes it conductive. Cellulose is also a good absorber of mid-range IR radiation. This absorbance is transformed into heat that may trigger a temperature increase, which might be interpreted as a detectable electric current variation.
Figure 2. Sensor device showing the graphite drawn electrodes with a 5 mm gap. The squares at both ends of the electrodes perform as pads, which are connected to the electrical contacts for the electrical measurements.
2. METHODOLOGY We follow a theoreticalexperimental approach to propose a mechanism that explains the origin of IR sensitivity observed in devices we fabricate. The experiments are aimed to demonstrate the origin of the electrical current of the device and to evaluate the sensor performance; on the other hand, theoretical calculations help us to identify the cause of the absorption of IR radiation. Experiment. We fabricate our sensors using high-purity Whatman 4210 filter paper. Sensors electrodes are drawn using graphite conductive ink. Figure 2 shows a picture of a paper sensor device. To grant conductivity to the devices, we deposit 1 drop (ca. 35 mL) of an electrolyte solution of potassium bromide (KBr) to the paper surface. We use KBr because it is transparent to a wide spectrum of light radiation11 and in this manner, we isolate the origin of the optical absorption of the device. After the drop deposition, the paper sensor is dried on a hot plate at 120 °C for a few minutes in order to evaporate the water. To test the electrical properties of the device at different humidity and electrolyte concentration conditions, we measure the conductivity of devices when adding a drop of deionized (DI) water and a solution with conductive salts: 0.1 M KBr and 1 M KBr. We perform these measurements in a humidity controlled environment to observe the effects of the presence of water and electrolytes in the sensor. To improve conductivity and stability of our devices, we have tried mixed solutions using water and glycerol at compositions 0%, 10%, and 30% glycerol/water vol %. We use glycerol because it evaporates at a temperature much higher than water (290 °C);12 this way, we have an electrolyte solution less
ARTICLE
Figure 3. Experiment deployed to test IR sensitivity of paper sensor devices.
dependent on relative humidity, and besides, it will not evaporate when gently heated. To study the effect of relative humidity of our sensor devices, we measure their conductivity in a drying environment inside a closed container with silica gel as drying agent. This process is performed while measuring the relative humidity (hygrometer Extech model RHT10) of the closed container and is done with samples at various concentrations of electrolytes and glycerol. In order to test the temperature dependence of the devices, we measure the conductance in a temperature controlled environment at dry conditions. To evaluate the sensitivity of the devices, we deployed a mechanism, including a rotatory window that exposes the sensors to a hot resistor at 120 °C at a rate of 100 mHz (5 s on, 5 s off); this resistor is 5 cm away from the sensor, and the sensitivity measurements are done at dry conditions. Figure 3 features a schematic diagram showing the process to test the sensitivity of the sensors. Electrical conductance measurements are done using a Keithley 6517A electrometer. We use a Perkin-Elmer Spectrum GX FTIR to analyze the IR absorbance of the used paper, this is done by separating the cellulose fibers using hot water and a ultrasonic bath for 1 h in order to disperse some fibers in DI water; later, a few drops of DI water with cellulose are deposited over a silicon wafer and dried to remove water. Thus, we obtain a thin dispersion of cellulose fibers suitable for FTIR characterization. Glycerol is also characterized using the attenuated total reflectance (ATR) technique. Theory. To study the IR absorption and vibrational modes of the materials involved in the paper sensor, we perform ab initio calculations of a dimer of cellulose and a glycerol molecule using the program Gaussian 03.13 The first step to find the vibrational modes is to obtain an optimized geometry for the molecules using ab initio DFT.13 We use the B3PW91 hybrid functional which combines the nonlocal Becke-3 (B3) exchange functional,14 the generalized gradient approximation (GGA) of PerdewWang (PW91),15 and an exchange component calculated similarly to HartreeFock (HF) but using the KohnSham molecular orbitals instead. The basis set used in all these calculations is the 6-31G(d,p).16,17 Once the optimized geometry is calculated the vibrational frequencies and intensities are calculated from the second derivative of the energy with respect to the atom positions and the polarizability. These methods have been widely tested in several types of systems such as in graphene sensors,18 encoding and transmission of data,19,20 metal clusters,21,22 energetic materials,2326 optical polymers,27 and several others. 18830
dx.doi.org/10.1021/jp206287f |J. Phys. Chem. C 2011, 115, 18829–18834
The Journal of Physical Chemistry C
ARTICLE
Figure 4. Conductance vs relative humidity for devices with different content of KBr electrolyte.
Figure 6. Conductance vs relative humidity for devices with different additives.
Figure 5. Conductance vs relative humidity for devices with different content of glycerol.
Figure 7. Conductance vs temperature dependence of a paper sensor device.
3. RESULTS AND DISCUSSION The effect of temperature and humidity on the conductivity of the paper sensor devices is studied to explain the mechanism that allows paper devices to conduct electrical current. The sensor performance is tested using sensors fabricated implementing different amounts of additives. The origin of the IR sensitivity of these devices is studied experimentally and theoretically. Experiment. The conductance of devices at several concentrations of electrolyte and glycerol is measured while removing the relative humidity in a controlled closed environment. We find marked behaviors of the sensor devices with respect to humidity variations as well as with paper additives. If the conductivities of devices having different amounts of KBr are compared, there is a sharp increment in conductivity of the paper sensors when the electrolyte is added. When the humidity from the environment is removed, a strong decrease of the conductivity takes place. Figure 4 shows the conductance (G) versus relative humidity (RH). The device having more KBr shows higher conductance, but as the humidity is removed, the slope sharply decreases, suggesting that at zero humidity environments, the conductivity will be similar to the device without the conductive salt. This implies that in a dry surface there are no agents to form a conductive film of electrolyte solution. We also study the effect of glycerol in the sensor device; different concentrations of glycerol in the sensor devices result in a strong change of conductance, depending on the amount of glycerol; glycerol is also capable to create conductive salt solutions. In this case we are not adding conductive salts, but some electrolytes are present probably as remainders of the paper fabrication. Interestingly, Figure 5 shows that the effect of humidity is weaker compared to the devices having only ionic salts. Thus, the current at zero humidity will have different conductivities as can be inferred from the presence of the glycerol layer.
Figure 8. Arrhenius plot showing ln(G) vs 1000/T for a paper IR sensor device.
We analyze the case of including both KBr and glycerol in the same sensor devices; this combination provides an important improvement in the conductance. There are 5 orders of difference of the conductance measurements between a device without additives and a device having both KBr and glycerol (Figure 6); this is important because a device with a larger conductivity is easier to measure and less noisy which are conditions desirable in any kind of sensor. The mechanism we propose as the possible origin to IR sensitivity is that the current of the devices increment due to a rise in temperature caused by a radiative interaction. To be able to detect this kind of change implies a strong dependency of the conductivity with temperature. Thus, we perform a measurement, heating a sensor device in a dry environment. The sensor device heated has KBr and glycerol additives to improve its conductance. Figure 7 shows the change of conductance when the sample is heated from 21 to 36 °C; in this range, the conductivity changes from 1.7 to 4.8 μS (280% difference). This very strong dependency makes plausible that 18831
dx.doi.org/10.1021/jp206287f |J. Phys. Chem. C 2011, 115, 18829–18834
The Journal of Physical Chemistry C
ARTICLE
Figure 9. Signals measured and normalized dividing the current by the average current to compare the amplitude change when a 100 mHz IR light signal is applied: (a) no additives; (b) 0.1 M KBr; (c) 1 M KBr.
Figure 10. The solution from these samples included 10% of glycerol: (a) No KBr; (b) 0.1 M KBr; (c) 1 M KBr.
Figure 11. The solution from these samples included 30% of glycerol: (a) No KBr; (b) 0.1 M KBr; (c) 1 M KBr.
a very small change in temperature could be detected by electronic means. The exponential growth of the conductance when the temperature rises suggests that the conductance of the device follows an Arrhenius behavior; it is well-known that ionic currents in electrolytes follow the Arrhenius law28 given by G ¼ Go expðEa =kTÞ
ð1Þ
where Go is the pre-exponential factor; this factor is usually proportional to the amount of charge carriers in the device. Ea is the activation energy of the ionic conduction, k is the Boltzmann constant, and T is the absolute temperature. Plotting the graph ln(G) versus 1000/T (Figure 8), good correlation between the data and the equation can be observed. From the slope, an activation energy of 0.54 eV (48 kJ/mol) is found for this process.
Figure 12. FTIR spectra for cellulose and glycerol.
This value is similar to the values observed in measurements of solid and liquid ionic conductivities.2931 The measurements of sensitivity to IR of paper devices with different content of KBr and glycerol have been normalized by 18832
dx.doi.org/10.1021/jp206287f |J. Phys. Chem. C 2011, 115, 18829–18834
The Journal of Physical Chemistry C
ARTICLE
Figure 13. Molecular geometries optimized using B3PW91/6-31G(d): (a) optimized geometry of a glucose dimer polymer chain; (b) optimized geometry of a glycerol molecule. Atom color code: carbon (gray), oxygen (red), and hydrogen (white).
Figure 14. Spectra calculated using ab initio for cellulose and glycerol. These spectra are compared with the emission spectrum of a hot body at 120 °C. This comparison suggests that the peaks around 1000 cm1 are capable of absorbing energy from the IR radiation of a hot body.
dividing the measured current (I) by the average current (Ia) the goal of this normalization is to obtain the current change when the sensors are exposed to the same source of infrared radiation and to be able to compare the sensitivity from devices having different amounts of additives. Figures 911 show the curves for a small group of paper sensors; a sensor without additives yields a small noisy signal, and as the amount of additives is increased, the 100 mHz signal applied to the IR sensors becomes stronger and better defined. It can be observed that the signal amplitudes from samples 9c, 10a, 10b, 10c, 11a, 11b, and 11c are similar in the range 24%. The main difference from these signals is the impedance; device 9c has an impedance of 3 GΩ and device 11c 2.1 MΩ. There are about 3 orders of magnitude in between these devices; the important point to notice here is that a sensor with lower impedance is less noisy and technically easier to read by standard electronics. To understand the sensitivity, IR spectroscopies are made from materials that could be involved in the IR absorption of the sensor; we are mainly studying cellulose and glycerol because it is well-known that KBr is transparent in this range of the optical spectrum. Figure 12 shows the FTIR results from cellulose and glycerol on these spectra; some strong absorption peaks are around 1000 cm1 (CO stretching), 1400 cm1 (COH scissoring), and 3300 cm1 (CH stretching). Theory. Cellulose is formed by hundreds of thousands of glucose units; however, the goal in this work is to study the vibrational modes. Thus we calculate a dimer of two glucose cellulose oligomers. The optimized structure for a dimer of cellulose is shown in Figure 13a.
Some of the paper sensor devices tested in this work contain glycerol as an additive; for that reason we also have optimized a glycerol molecule in order to study its vibrational states. Figure 13b shows the optimized geometry of a glycerol molecule calculated using ab initio DFT methods. From the optimized geometries we calculate the vibrational modes of cellulose and glycerol. The intensity of the IR absorption is obtained as molar absorptivity ε (M1 cm1). This intensity is converted to mass absorptivity (g1 cm1) in order to make a better comparison between cellulose and glycerol. From the calculated spectra of these two materials, we notice in Figure 14 that glycerol and cellulose have similar absorption peaks, around 1000 and 3000 cm1. This is due to the presence of CO and CH bonds in both of the systems. To correlate the IR absorption to the IR radiation emitted by hot bodies, we compare the calculated absorption peaks with the blackbody emission for an object at 120 °C (393 K). From the blackbody emission spectrum, there is a large contribution of IR at about 1000 cm1; thus we infer that the sensitivity to IR noticed during the experiments using a hot resistor at 120 °C is being absorbed by the vibrational modes of cellulose and glycerol near to 1000 cm1. It is worth noticing that there is a good correlation between the spectrum obtained from experiment and the ones calculated by ab initio methods. In this case the advantage to have a tool to calculate the frequencies and intensities of different materials could help to evaluate the potential of new materials to develop sensors of specific wavelengths.
4. CONCLUSIONS In this work we develop a simple and low-cost sensor to detect IR radiation; the materials and processes involved in the fabrication of this kind of optical sensor are abundant and environment friendly. It is confirmed experimentally that the conductivity of the devices tested have a very important dependence on humidity and the addition of glycerol and electrolyte salts highly improves the conductivity of the devices. On the other hand, we also observe a large direct dependence of conductivity with temperature; this is a well-known behavior in electrolyte solutions because the mobility of the ions increments exponentially when temperature rises. The strong dependencies on humidity and temperature support our hypothesis that the origin of the conductivity of paper substrates is due to ionic transfer in a thin 18833
dx.doi.org/10.1021/jp206287f |J. Phys. Chem. C 2011, 115, 18829–18834
The Journal of Physical Chemistry C film of electrolyte liquid formed on the surface of the cellulose fibers. The origin of the sensitivity is explained as a radiative heating effect caused by the absorption of IR energy principally by the CO bonds present in cellulose and glycerol. These CO bonds have and important optical activity in frequencies around 1000 cm1, which corresponds to the main electromagnetic emission from bodies at temperatures slightly above room temperature. CH bonds do not contribute as strongly as CO because they absorb IR energies around 3000 cm1, which correspond to higher blackbody temperatures.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected].
’ ACKNOWLEDGMENT J.M.S. thanks the Defense Threat Reduction Agency through the Army Research Office for their support, Project Nos. W911NF-06-1-0231 and W91NF-07-1-0199. A.J.G. thanks Mr. Flores Farias for his assistance in the photoresponse measurements and acknowledges support from Conacyt. ’ REFERENCES (1) Klemm, D.; Heublein, B.; Fink, H.-P.; Bohn, A. ChemInform 2005, 36 (September 6). (2) Gimenez, A. J.; Ya~nez-Limon, J. M.; Seminario, J. M. J. Phys. Chem. C 2011, 115, 282. (3) Martinez, A. W.; Phillips, S. T.; Carrilho, E.; Thomas, S. W.; Sindi, H.; Whitesides, G. M. Anal. Chem. 2008, 80, 3699. (4) Nie, Z.; Nijhuis, C. A.; Gong, J.; Chen, X.; Kumachev, A.; Martinez, A. W.; Narovlyansky, M.; Whitesides, G. M. Lab Chip 2010, 10, 477. (5) Planck, M. The Theory of Heat Radiation, 2nd ed.; Blakiston’s Son & Co.: Philadelphia, PA, 1914. (6) Kuwano, Y.; Yokoo, T.; Shibata, K. Jpn. J. Appl. Phys. 1981, 20 (supplement 4), 221. (7) Kim, J.; Yun, S.; Ounaies, Z. Macromolecules 2006, 39, 4202. (8) Sharma, A. K.; Ramu, C. Br. Polym. J. 1990, 22, 315. (9) Robinson, R. A.; Stokes, R. H. Electrolyte solutions; Dover Publications: Mineola, NY, 2002. (10) www.whatman.com. (11) Johnson, K. W.; Bell, E. E. Phys. Rev. 1969, 187, 1044. (12) Lide, D. R.; Milne, G. W. A. Handbook of data on organic compounds; CRC Press: Boca Raton, FL, 1994. (13) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Scalmani, G.; Mennucci, B.; Barone, V.; Petersson, G. A.; Caricato, M.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Li, X.; Hratchian, H. P.; Peralta, J. E.; Izmaylov, A. F.; Kudin, K. N.; Heyd, J. J.; Brothers, E.; Staroverov, V. N.; Zheng, G.; Kobayashi, R.; Normand, J.; Sonnenberg, J. L.; Ogliaro, F.; Bearpark, M.; Parandekar, P. V.; Ferguson, G. A.; Mayhall, N. J.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Burant, J. C.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.;
ARTICLE
Chen, W.; Wong, M. W.; Pople J. A. Gaussian 03, revision B.04; Gaussian, Inc.: Wallingford, CT, 2003. (14) Becke, A. D. J. Chem. Phys. 1996, 104, 1040. (15) Perdew, J. P. In Electronic Structure of Solids; Ziesche, P., Eschrig, H., Eds.; Akademie Verlag: Berlin, 1991; p 11. (16) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (17) Hohenberg, P.; Kohn, W. Phys. Rev. B 1964, 136, 864. (18) Rangel, N. L.; Seminario, J. M. J. Chem. Phys. 2010, 132, 125102. (19) Yan, L.; Ma, Y.; Seminario, J. M. J. Nanosci. Nanotechnol. 2006, 6, 675. (20) Seminario, J. M.; Yan, L.; Ma, Y. J. Phys. Chem. A 2005, 109, 9712. (21) Seminario, J. M.; Tour, J. M. Int. J. Quantum Chem. 1997, 65, 749. (22) Seminario, J. M.; Zacarias, A. G.; Castro, M. Int. J. Quantum Chem. 1997, 61, 515. (23) Politzer, P.; Seminario, J. M. Chem. Phys. Lett. 1993, 207, 27. (24) Politzer, P.; Seminario, J.; Bolduc, P. Chem. Phys. Lett. 1989, 158, 463. (25) Seminario, J. M.; Concha, M. C.; Politzer, P. J. Comput. Chem. 1992, 13, 177. (26) Murray, J.; Redfern, P.; Seminario, J.; Politzer, P. J. Phys. Chem. 1990, 94, 2320. (27) Choi, D.-S.; Huang, S.; Huang, M.; Barnard, T. S.; Adams, R. D.; Seminario, J. M.; Tour, J. M. J. Org. Chem. 1998, 63, 2646. (28) De Berg, K. C. Sci. Educ. 2003, 12, 397. (29) Ngai, K. L.; Greaves, G. N.; Moynihan, C. T. Phys. Rev. Lett. 1998, 80, 1018. (30) Chen, X.; Xia, C.; Stimming, U. Electrochim. Acta 2007, 52, 7835. (31) Vieira, D. F.; Avellaneda, C. O.; Pawlicka, A. Electrochim. Acta 2007, 53, 1404.
18834
dx.doi.org/10.1021/jp206287f |J. Phys. Chem. C 2011, 115, 18829–18834
