Exposure to toluene resulted in a partially reversible shift in the resonance depth and position of ... parts per billion level), yet there are inherent delays between.
ELSEVIER
Thin Solid Films 284285 (1996) 98-101
Toluene vapour sensing using copper and nickel phthalocyanine Langmuir-Blodgett films C. Granito ‘*’, J.N. Wilde ‘, MC Petty ‘, S. Houghton b, P.J. Iredale b aSchool of Engineering and Centre for Molecular Electronics, Universiry of Durham, South Road, Durham DHI 3LE, UK b Neotronics Limited, Parsonage Road, Takeley. Bishop’s Storrford, Herts. CM22 6PU, UK
Abstract Toluene vapour sensing has been successfully demonstrated using Langmuir-Blodgett films prepared from substituted phthalocyanines, containing copper or nickel as the central metal ions. High-quality layers have been built-up and the films characterised using surface plasmon resonance (SPR). Exposure to toluene resulted in a partially reversible shift in the resonance depth and position of the SPR curves. The optical sensor could detect concentrations of this organic vapour down to, at least, 50 parts per million. Keywords: Sensing;
Toluene; Phthalocyanine;Langmuir-Blodgen films
1. Introduction
There is an increased need for low-power, sensitive, selective and reliable gas sensors. This is partially a result of increased levels of pollution, especially from environmentally dangerous gases such as SO*, NO,, and health-threatening gases, such as toluene and benzene. The Agency for Toxic Substances and Disease Registry in the USA state that long-term low-level exposure (200 parts per million (ppm) for over 1 year) to toluene can cause damage to kidneys, memory loss and nausea. Higher levels of toluene (600 ppm for over 1 year) can cause permanent damage to the brain, and continuous exposure can result in death. At present, the most reliable way to determine concentrations of this organic vapour is in an analytical laboratory, where samples of air are sent and tested by chemical or gas chromatography techniques [ 11. These methods have the advantages of being very accurate and sensitive (down the parts per billion level), yet there are inherent delays between taking the sample and receiving the results. This problem is accentuated by draft legislation in the USA stating that employees cannot enter enclosed areas with more than 100 ppm of airborne toluene. Therefore, a portable, selective sensor has to be developed to meet this new standard. Inorganic compounds, such as metal oxides, are often used to detect gases or vapours, including toluene. Normally, such ‘ Present address: Departmentof MaterialsScience, University of Lecce, Via Amesaao 130.73100 Lecce, Italy. 0040-6090/96/$15,00 0 1996 Elsevier Science S.A. SSDIOO40-6090(95)08280-S
materials have to be operated at elevated temperatures to aid desorption of the gas. The associated power consumption of devices based on these compounds could limit portable gassensing applications. An alternative approach is to exploit thin films of organic materials, Many substances, such as phthalocyanines, are known to exhibit a high sensitivity to gases and good reversibility at room temperature [ 21. The Langmuir-Blodgett (LB) technique can be used to produce ultra-thin (molecular thickness) films of these compounds. Recently, toluene sensing has been reported using LB layers ( 120 nm thick) of copper 1,4-bis( 5hydroxypentoyloxy) 8,11,15,18,22,25-hexa-isopentyl phthalocyanine [ 31. Toluene gas concentrations were monitored using surface conductivity changes, resulting from interactions between the film and toluene. Optical sensing can offer advantages over electrical devices in certain measurement environments. Here, we report preliminary results on the development of an optical organic sensor, using copper and nickel phthalocyanine LB films. The reaction on the surface of the LB films due to the presence of toluene vapour was measured using the technique of surface plasmon resonance (SPR) [ 4,5]. The molecular formula for the compounds used is shown in Fig. 1. The first two compounds are copper phthalocyanines (CuPc) : copper ten&is-( 3,3-dimethyl- l-butoxycarbonyl) phthalocyanine (CuPcBC) and copper tetraphthalocyanine kis- (3,3-dimethyl- 1-neopentoxycarbonyl) (CuPcNC) . The other material contained nickel as the central metal: nickel tetrakis- (3,3-dimethyl- 1-butoxycarbonyl) phthalocyanine (NiPcBC) .
C. Granito et al. /Thin Solid Films 284-285 (19%) 98-101
R:
--c
-
~HFWCW3
WRW
-
C-VXH,,,
(CIRNC)
M : Cu or Ni Fig. 1. Structure of the tetrasubstituted copper and nickel phthalocyanines used in this work.
2. Experimental Glass slides were used as substrates, these were cleaned with a solution of Decon 90, and placed in an ultrasonic bath for 30 min. The cleaned glass slides were coated with a 50 nm thick layer of silver, by vacuum deposition. Phthalocyanine films were then deposited on top of the silver substrate using the LB technique. The synthesis and purification of the copper and nickel phthalocyanine molecules have ben discussed previously [ 61. The deposition process was undertaken using a constant-perimeter barrier LB trough located in a microelectronics clean room [ 71. The phthalocyanine compounds were transferred to the water surface using 25% mesitylene (by volume) in ethyl acetate, at a known concentration of 0.5 g 1-r. The solvents were ethyl acetate (99%) (Hopkins and Williams); mesitylene (99%) (BDH). 400 pl of the phthalocyanine solution was spread onto a subphase of pure water (pH of 5.8 f 0.2 at 20 f 2 “C) . The layer was left for 15 min so that the solvent could evaporate; the film was then compressed at 0.01 nm* molecule- ’ s- ’ and controlled at 30 mN m- ’ . The layers of phthalocyanine were deposited using an alternate-layer trough; one side had a clean water surface, the other a controlled layer of phthalocyanine. The substrate was lowered through the clean water surface and moved, underwater, through an interconnecting gate at 50 mm min - ‘. The substrate was withdrawn through the air-water interface at a spe4zd of 9 mm min- ‘, picking up a single layer of phthalocyanine, then left for 15 min to dry; the deposition ratio was approximately unity. This process was repeated to transfer further layers (i.e., Z-type deposition). The glass substrates were coupled to a glass prism which focused monochromatic light (p-polarised) from a He-Ne laser (h = 632.8 nm) onto the LB layer [ 41. The reflectivity from the surface of the film was collected via a photodiode, and its output was displayed on an X-Y chart recorder. Toluene vapour was generated using the diffusion cell shown in Fig. 2 [8]. This consisted of a toluene reservoir
Fig. 2. Schematic diagram of the diffusion cell used to generated toluene vapour at fixed concentrations [7].
immersed in a temperature-controlled water bath and an upper cell where nitrogen mixed with the toluene vapour. The design of the capillary tube was essential since this determined the operating temperature of the water bath and, thereby, the concentration of toluene. Using Eq. (3) in Ref. [ 81, we were able to compute the temperatures of the water bath to obtain concentrations of toluene in the range 50-200 ppm PI.
3. Results and discussion The three phthalocyanine single-layer LB films (CuPcBC, NiPcBC, CuPcNC) were fist character&d by SPR. The results shown in Fig. 3 are for CuPcNC. As the number of overlayers increases, the SPR curve broadens and the point
0.2
0 40
41 Internal
42 angle
43 44 [degrees]
45
Fig. 3. Surface plasmon resonance (h-632.8 nm) curves for: (a) 50 nm silver layer; (b) silver + 1 LB layer CuPcNC; (c) silver + 2 LB layers of CuPcNC.
100
C. Granito et al. /Thin Solid Films 284-285 (19%) 98-101
Table 1 Comparison between the refractive indices and layer thickness measurements for the substituted phthalocyanines, from SPR (A = 632.8 nm) and ellipsometry (A=405 nm) Phthalocyanine
Refractive index (SPR)
Layer thickness (SPR) (nm)
Refractive index (ellipsometry)
Layer thickness (ellipsometry) (nm)
CuPcBC NiPcBC CuPcNC
1.60 f 0.03 1.82*0.05 1.66 f 0.02
1.87*0.51 1.38 kO.52 2.42+0.13
1.62rtO.03 1.77rtO.02
2.37 f0.05 1.49 f 0.14
of minimum intensity shifts to higher angles [9]. The decrease in resonance depth can be attributed to absorption by the film, since the absorption peak for CuPcNC is very close to the wavelength of the incident laser light [ lo]. Similar data were recorded for NiPcBC and CuPcBC. The experimental results were modelled on a Unix workstation using an in-house program to give values for refractive index and film thickness [ 111. Table 1 summarises the results and compares these with ellipsometric data [ lo]. The film thickness data obtained from SPR and ellipsometry are the same within experimental error. The values support our previous suggestion that each LB layer transferred to the substrate surface consists of a single molecule thickness film in which the molecules are arranged in stacks with their large molecular faces perpendicular to the substrate surface [ lo]. A coated slide was mounted in the SPR system; the internal angle of the incident laser beam was chosen to be on the steepest gradient of the SPR curve, on the low-angle side of resonance. The sample was initially exposed to dry nitrogen. After the output had settled ( 10 min), toluene was added via the diffusion cell. The first film exposed to toluene vapour was a 50 nm layer of uncoated silver on a quartz glass substrate. This was to establish the effect of toluene on the underlying silver layer. Fig. 4 shows the change in the reflectivity curve on exposure to increasing concentrations of toluene. The slow and monotonic increase in the measured reflectivity corresponds to a shift in the SPR curve to higher angles. There was no observed recovery (in the time scale of the experiment) when the toluene was replaced by dry nitrogen. The phenomenon is possibly related to toluene condensing on the
0
20
40
60
80
100
120
140
160
[mini Fig. 4. 50 nm silver layer exposed to 50/100/200 ppm toluene vapour in nitrogen at a flow rate of 100 cm3 mm-‘.
275
0
20
40
60
80
loo
120
140
160
Time [min] Fig. 5. 1LB layer of CuPcNC on 50 nm silver exposed to 50/100/200 ppm toluene vapour in nitrogen at a flow rate of 100 cm3 min-‘.
silver surface. Similar effects have been documented for other organic vapours on gold [ 121. The above experiment was then repeated with a single LB film of CuPcNC. The results are presented in Fig. 5. In contrast to the data shown in Fig. 4 for the reference silver slide, exposure to toluene produced a rapid increase in the reflectivity. The reflectivity change seemed almost independent of toluene concentration over the range investigated (Xl-200 ppm) . For low concentrations, a partial recovery in reflectivity was observed when the toluene was turned off. However, the effect was irreversible at high toluene concentrations. We suggest that the irreversible part of the response is associated with the interaction of toluene and the silver film (the irreversible shift in the photodiode output for both Fig. 4 and 5 is approximately 30 mV for similar toluene exposures). The reaction between the organic vapour and the phthalocyanine film is probably reversible and could result from an increase in the film thickness or refractive index (or both). We have also studied the effect of increasing the LB film thickness. A 2-layer CuPcNC LB film exhibited a similar sensitivity, but both longer response and recovery times; again, the recovery was only partial. The CuPcBC LB film showed a similar response to toluene as measured for the CuPcNC LB film. In contrast, the sensitivity of the NiPcBC LB film was somewhat less. 4. Conclusions
Time
We have shown that toluene vapour can be detected with phthalocyanine Langmuir-Blodgett films, using the tech-
C. Graniro et al. /Thin Solid Films 284-285 (1996) 98-101
nique of surface plasmon resonance. Preliminary measurements reveal that this optical method can readily detect concentrations of the organic vapour down to 50 parts per million (the lowest concentration investigated). This is certainly sufficient for the development of new portable sensing instruments based on current legislation. One immediate problem, however, concerns the poor reversibility of the sensor. We suggest that this results from the interaction of toluene with the underlying silver substrate. We are currently exploring both the use of other metal coatings and methods to “seal” the silver surface. Experiments to determine the cross-sensitive of the sensor to other vapours are also under way.
Acknowledgements
The authors would like to thank the Engineering and Physical Sciences Research Council and Neotronics Limited for supporting this work.
101
References
[ 1I M. Whalen, J.N. Driscoll and CD. Wood, Amospheric Environmenr. 28 (1994) 567-570. [21 MC. Petty, Biosensors Bioelectron., 10 (1995) 129-134. [31 S. Mukhopadhyay, CA. Hogarth, SC. Thorpe and M.J. Cook, J. Mater. Sci.. 5 (1994) 321-323. [41 J.P. Lloyd, C. Pearson and M.C. Petty, Thin Solid Films, 160 (1988) 431-443. [51E. Kretschmann and H. Raether. Z. Naturforsch., 23a (1968) 21352136. [61 L. Pasimeni, M. Meneghetti, R. Rella, L. Valli, C. Granito and L. Tmisi, Thin Solid Films, in press. [71 G.G. Roberts, P.S. Vincett and W.A. Barlow, Phys. Technol., 12 (1981) 69. [81 J.M. Mckelvay and H.B. Hoelscher,Anal. Chem., 29 (1957) 123. [91 A. Otto, Phys. SrarusSolidi, 42 (1970) k37. [ 101C. Granite, L.M. Goldenberg, M.R. Bryce, A.P. Monkman, L. Troisi. L. Pasimeni and M.C. Petty, Lungmuir, submitted. [ 111J. Cresswell, Ph.D. Thesis, University of Durham, 1992. [121 P.S. Vkusic, G.W. Bradbeny and J.R. Sambles, Surf. Sci. Len., 277 (1992) L34-L38.
