Available online at www.sciencedirect.com
Sensors and Actuators B 130 (2008) 343–350
A study of partially irreversible characteristics in
a TTF–TCNQ gas sensing system
夽
Jung-Yu Liao
a, Kuo-Chuan Ho
a,b,∗aDepartment of Chemical Engineering, National Taiwan University, Taipei 10617, Taiwan bInstitute of Polymer Science and Engineering, National Taiwan University, Taipei 10617, Taiwan
Available online 10 August 2007
Abstract
The irreversible gas sensing behavior is investigated for TTF–TCNQ complex in the presence of NO2or O2. Reactions with the gas molecules
are the main reason to cause the irreversibility, which has been reported and supported by analytical data. The totally irreversible behavior was noticed when a TTF–TCNQ thin film was brought in contact with NO2gas. As for O2gas, however, only partially irreversible phenomenon was
observed. For a TTF–TCNQ thin film that interacts with NO2or O2gas, the degree of irreversibility is determined by the competition ratio of the
desorption rate to that of the reaction rate. A theory based on the competition concept was proposed and a general expression for three possible behaviors (totally reversible, totally irreversible, and partially reversible cases) in a gas sensing system was obtained. The O2sensing data match
the theoretical prediction when the reaction and the adsorption are limited at the surface. © 2007 Published by Elsevier B.V.
Keywords: Competition theory; Desorption; Irreversibility; Nitrogen dioxide sensor; TTF–TCNQ complex
1. Introduction
Tetrathiafulvalene (TTF)–tetracyanoquinodimethane (TC-NQ) complex is a highly conductive[1]organic material, and is found to be sensitive to several gas species, such as NO2,
O2, CO2, etc.[2] Its individual components, TTF and TCNQ
molecules and their individual derivatives, were reported to be ambient gas sensitive as well. There have been several discus-sions about how these materials are affected by the gases that surround them[3–8].
It is conceived that the structure of TTF[3]has been changed in IR spectrum (805–835 and 1270–1420 cm−1) during NOx
exposure. The TTF derivative thin film[4]have been reported as well to become radical cation when dipping in iodine vapor. Hen-rion et al.[5]also reported that redox reaction, color change, as well as resistivity change can be observed when TCNQ-derived LB films were exposed to iodine gas. Perez et al.[6] discov-ered the change of charge transfer band of TCNQ derivative films under NO2exposure. Hertler et al.[7]and Suchanski and
夽 Manuscript No. TP20#356 presented at the 11th International Meeting on
Chemical Sensors, Brescia, Italy, 16–19 July 2006.
∗Corresponding author at: Department of Chemical Engineering, National
Taiwan University, Taipei 10617, Taiwan. Fax: +886 2 2362 3040. E-mail address:kcho@ntu.edu.tw(K.-C. Ho).
Duyne[8]showed that TCNQ and TCNQ2−anion decomposed in the presence of NO2and O2gases.
Compare to the existing evidence for TTF, TCNQ and their individual derivatives, there are only limited information about the effect of ambient gases on the sensing properties of a TTF–TCNQ complex. For example, Jouve et al. [2] reported that a TTF–TCNQ complex containing a polymer film shows irreversibility in NO2detection and partial reversibility in O2
and H2O detections. Based on the possible reactive nature of
TTF–TCNQ complex, our previous work[9]had focused only on the irreversible sensing and modeling for NO2 detection.
One of the objectives of this study is to find out the reaction mechanism for a TTF–TCNQ complex when exposing to NO2
or O2 gas. Another objective is to establish a general model
(completely reversible, partially reversible for O2, and totally
irreversible for NO2) for those sensing materials that might be
reactive.
2. Experimental
TTF–TCNQ complex is obtained by dissolving both TTF (Fig. 1(a)) and TCNQ (Fig. 1(b)) powders (Aldrich, TTF:97%; TCNQ:98%) into acetonitrile to form a 1:1 molar ratio radical cation–radical anion pair complex, as a precipitate
0925-4005/$ – see front matter © 2007 Published by Elsevier B.V.
Fig. 1. The molecular structures of: (a) Tetrathiafulvalene, TTF; (b) Tetra-cyanoquinodimethane, TCNQ; (c) TTF–TCNQ complex.
(Fig. 1(c)). The preparation steps follow the procedure described in literature [1]. The complex powder is further purified by recrystallization in acetonitrile. Sensing thin film is thermally evaporated onto various substrates, including KBr (for FTIR), glass (for UV–vis, SEM and XRD), Pt coated quartz (for QCM), and Al2O3with printed interdigitated gold electrode (for
con-ductance) under 3× 10−5Torr vacuum. Film thicknesses are controlled at about 0.15–0.2m for both conductance and QCM measurements and 0.50–1.00m for other analytical experi-ments (XRD, UV–vis, FTIR and SEM). The deposition rate is
ca. 5–10 ˚A/s.
The sensing film that face against the gas inlet at a flow rate of 200 ml/min is placed in a sensing chamber described elsewhere
[9]. The sensing was carried out under room temperature. NO2
and O2(from liquid air) gases are diluted with pure N2that is
inert in this system. Unless stated otherwise, the standard sens-ing (exposure) process for analytical experiments was carried out for at least 12 h with 500 ppm NO2gas stream at 200 ml/min.
The conductance was recorded by EG&G model 273A poten-tiostat/galvanostat. The QCM measurements were accomplished by Seiko EG&G QCA917, respectively. The UV–vis and FTIR spectra were carried out by Shimadzu UV-1601PC and Bio-rad FTS-40. The XRD and SEM data were obtained by Philips PW1710 (λ = 1.540 ˚A, scan rate = 3 deg/min) and Hitachi S-800, respectively. The detail experimental setup can be found in our previous work[9].
3. Results and discussions
3.1. Reaction involved sensing mechanism
In our previous work [9], a totally irreversible reaction is assumed in the model deduction. We are interested in finding the mechanism involving the conductive property and how the conductivity is affected. Several methods have been carried out to identify the possible reactions and conductive mechanisms.
3.1.1. Irreversible sensing response
Fig. 2(a) is a typical irreversible conductance response for NO2detection using a TTF–TCNQ complex. First of all, thin
trend which is consistent with the conductivity transient in
Fig. 2(a).
After confirming the immobilization of NO2, as judged
by the increasing mass and decreasing conductance of a TTF–TCNQ complex, two reasonable possibilities may be drawn regarding the incoming NO2gas molecules, which are
either “absorbed” within the film or “reacted” with the complex to form additional products and being immobilized inside the film.
Fig. 2. Typical irreversible responses of NO2detection in (a) conductivity and
Fig. 3. XRD intensity for fresh and NO2-exposed TTF–TCNQ thin films on
glass substrate.
3.1.2. Crystal structure
The standard X-ray diffraction (XRD) pattern with three characteristic peaks[10]representing different packing planes is shown in Fig. 3. The crystalline intensity is dramatically decreased when the exposure process is finished. This suggested that the penetration of NO2molecules accompanied by reaction
ruins the standard packing of this complex, hence lowering the intensity in the XRD pattern.
It is generally accepted that the distances between two par-allel TTF and two parpar-allel TCNQ molecules in the complex are about 3.47 and 3.17 ˚A[11], respectively. According to the kinetic theory of gas, the molecular diameter of a spherical molecule can be estimated by the viscosity of the gas[12]
d2=2.67 × 10
−20× (MT)1/2
η (1)
where, d is the diameter of spherical molecule, M the molar mass,
T the temperature in Kelvin, and η is the viscosity of the gas.
Assuming that all molecules are spherical molecules and that the viscosities for N2, NO2and O2at 25◦C are 1.78× 10−5,
1.32× 10−5and 2.04× 10−5(pa s), respectively[13]. Thus, the diameters calculated from Eq.(1)for N2, NO2and O2are 3.71,
4.87 and 3.58 ˚A, respectively, which are larger than the inter-planar distance of TTF–TCNQ complex. It is highly unlikely that TTF–TCNQ complex itself will “absorb” the NO2molecule
and allow the gas to diffuse into the interplanar space which is relatively narrower. Since the immobilization was observed in either conductance or QCM measurement (as seen in Fig. 2), it indicates that the immobilization of gas molecules will bring the destruction of the standard packing of the complex, as evi-denced in Fig. 3. If a real stick type (O2 and N2) or curvy
molecular (NO2) shape is taken into account, the smoothly slip
of molecules into the interplane becomes more difficult than a spherical one, and thus unlikely to happen.
3.1.3. Spectra analysis
Fig. 4(a) is the UV–vis spectra for a fresh and a NO2-exposed
film samples dissolved in acetonitrile. It can be found that three absorption peaks, which represent TTF•+(318 nm, may has been shifted from 340 nm) and TCNQ•−(743 and 838 nm, according
Fig. 4. (a) Visible spectra for dissolved fresh and NO2-exposed TTF–TCNQ
thin films; (b) IR spectrum for fresh and NO2-exposed TTF–TCNQ thin films
on KBr substrate.
to literatures[8,14–17], as summarized inTable 1) decrease after NO2exposure. The major peak at 394 nm, which is due to the
oxidation of TCNQ•− to form TCNQ0(both species absorb at 390–400 nm, referring toTable 1), still maintains its intensity after NO2exposure. Minor absorption at 533 nm represents the
formation of TTF2+.
It has been reported that TCNQ0molecule reacts with NO2to
form terephthaloyl cyanide[7]and TCNQ2−ion reacts with O 2
to form DCTC−(␣,␣-dicyano-p-toluoyl-cyanide) ion[8]. Both reactions are basically similar and are mainly involved in a side chain transformation (two sides for NO2and one side for O2).
The difference is mainly determined by the oxidizing capability resulting from the active radical’s existence in NO2 molecule.
Table 1
The summarized UV–vis absorption peak ranges for TTF and TCNQ related species
Species Peak ranges (nm) References
TCNQ•− 400–455; 660; 728–760; 812–850 [14–16]
TCNQ0 390–412; 430–432 [14–16]
TTF•+ 340; 393–405; 435; 575; 653 [16,17]
TTF2+ 533 [16]
(Table 1).
The FTIR spectra for the fresh and the NO2-exposed
TTF–TCNQ thin films are shown inFig. 4(b). The major absorp-tion peaks (475, 860, 1354 and 1543 cm−1, as summarized in
Table 2)[18]found after NO2exposure represent the formation
of TCNQ0. Based on both UV and FTIR analyses, it is inferred that the major reaction involving TCNQ•−is
NO2+ TCNQ •− ↔ TCNQ0+ NO− 2, G 0 2 (2)
It is also reported that TTF0 is oxidized when exposing to NO2 with IR absorption region detected at 805–835 and
1270–1420 cm−1[3]. It is possible that the peak at 1377 cm−1 is due to the oxidation of TTF•+, TTF2+or their mixture[19]. Because the NO2-exposed film is yellowish in appearance, not
deep purple (TTF•+)[17], it is suggested that TTF•+is further oxidized to yellowish di-cation (TTF2+) in the actual reaction
NO2+ TTF •+ ↔ TTF2++ NO− 2, G 0 3 (3)
Since the bi-molecular exchanging from NO2to N2O4 are
known to exist, the formation of NO2− (absorbance peaks at
1325, 1270, and 829 cm−1[20]) is kinetically possible and ther-modynamically feasible as described by the following reduction equation:
N2O4(= 2NO2)+ 2e−↔ 2NO−2,
E40= 0.88 V (versus NHE) (4)
In addition, the potentials for the redox couples TTF2+/TTF+ and TCNQ−/TCNQ0are[21]
TCNQ0+ e−↔ TCNQ−,
E50= 0.13 V (versus SCE) or 0.37 V (versus NHE) (5)
TTF2++ e−↔ TTF+,
E60= 0.66 V (versus SCE) or 0.90 V (versus NHE) (6)
Therefore, the G02for Eq.(2)is
G02= −1 × F × (E04− E05)= −1 × 96485 × 0.51
= −49.2 kJ/mol
molecules and the previous reactions repeat again. The remained porous product is called “tarnishing film[9].”
To summarize our previous discussion, a net reaction is pro-posed
2NO2+ TTF − TCNQ k
−→TTF2++ TCNQ0+ 2NO−
2 (7)
This is to compare with the equation proposed in our previous work[9]
aNO2(g)+ TTF − TCNQ(s) k
−→[(NO2)a− (TTF − TCNQ)](s)
(8) which can be denoted as
aB(g)+ A(s) k
−→ABa(s) (9)
where “a” is a stoichiometric factor between the reactants, which takes two NO2molecules to react with both TTF•+and
TCNQ•−, B represents NO2, and A represents TTF–TCNQ
com-plex, ABamay be viewed as the comination of the mixed three
products listed in Eq.(7).
3.1.4. Hypothetical mechanism
As far as we know, NO2does react with TTF–TCNQ
com-plex, and it is reasonable to assume that the reaction is initiated by the strong interaction involving the unpaired electron (rad-ical) located in the molecular orbital of NO2 and the radical
cation–radical anion pair of TTF–TCNQ complex (Fig. 1(c)). It is possible that the unpaired electron of NO2 was attracted
initially by the radical hopping through TTF–TCNQ complex packing. After the attracted NO2reacted with the complex, the
conductive route is interrupted, as evidenced by both the contin-uous decrease in the conductance (Fig. 2(a)) and the increase in the mass (decrease in frequency,Fig. 2(b)) during NO2exposure.
Based on the well-known structure of the TTF–TCNQ complex
[11], which possesses 1-D (y-direction) an-isotropic conduction
[22] and our previous discussion, we propose a hypothetical mechanism, as depicted inFig. 5, to visualize the interaction. In this mechanism, the highly conductive property of the com-plex is contributed by the “active” hopping radical provided by the TTF–TCNQ complex. The un-paired, highly reactive NO2
is attracted and reacted with the complex. The porous products are left behind (TTF2+, TCNQ0, NO2−, and small amount of
Fig. 5. Hypothetical conducting and reacting mechanism for the NO2–TTF–TCNQ system.
DCTC−) and the conduction route is destroyed, while the un-reacted NO2 keeps penetrating through the remains and react
with the fresh complex.
3.1.5. Surface information from SEM
In the previous discussions, we imagine that the tarnishing film formed is a porous layer and allows larger molecules such as NO2to penetrate through. According to the SEM graphs shown
inFig. 6(a), which is the SEM image of a fresh TTF–TCNQ film coated onto a glass substrate, the surface of each chip of fresh TTF–TCNQ crystalline is quite smooth comparing to the NO2-exposed film inFig. 6(b). The increase in surface roughness
after NO2exposure supports the formation of a porous tarnishing
film.
3.2. General modeling of TTF–TCNQ-based sensors
The TTF–TCNQ complex not only is sensitive to NO2but
also sensitive to O2, as shown inFig. 7. When exposing to O2, the
transient conductance shows partially irreversible sensing char-acteristics for the TTF–TCNQ complex. This sensing behavior is different from that of NO2, as seen from Fig. 2(a). This
means that the interaction between O2and TTF–TCNQ
com-plex involves simultaneous adsorption/desorption and reaction processes.
In order to describe the partially irreversible behavior, the totally “irreversible reaction” model (proposed in our previous work[9]) and the totally “reversible adsorption/desorption[23]” model are adopted to form a general model.
Fig. 7. Experimental data and modeling prediction for a partially irreversible response of O2 sensing using a TTF–TCNQ complex. Segment α represents
the actual reacted amount; β represents the actual adsorbed amount; γ is the predicted reacted amount predicted; and δ is adsorbed amount predicted.
The reversible adsorption/desorption theory is based on Langmuir isotherm, which has been reported to explain the sens-ing characteristic of NO2 and a lead phthalocyaine thin film
(PbPc)[23]. The general Langmuir expression is written as
aB(g)+ A(thin film surface)←→ka
kd A − Ba(adsorbed) (10)
where “a” is the stoichiometric factor, B represents the incoming gas, A the sensing film which offers the surface area, and A− Ba
represents the adsorbed species.
Given the fact that O2 is also reactive with the film, the
adsorbed species (A− Ba) would react further with O2 (since ABain Eq.(7)represents the final products), or
aB(g)+ A(thin film surface)←→ka kd
A
−Ba(adsorbed) k
−→ABa(reacted) (11)
Several assumptions are made to validate the model before pursuing further analysis:
(i) Both adsorption/desorption and reaction are assumed to occur at the thin film surface only, which implies that the reacted amount and the rate of reaction are negligible. Once the reaction is taking place inside the film, this assumption is not valid.
(ii) In this work, “A” represents TTF–TCNQ, which is assumed to offer two (TTF•+and TCNQ•−) “identical sites (denoted as A)” for the adsorption.
(iii) The reactive gas, B (e.g., O2 or any other oxidizing gas),
behaves similarly to NO2as seen in Eq.(7)with a
stoichio-metric factor of 2. Therefore, each site (A) adsorbs only one gas molecule. Therefore, the general model, Eq.(11), can be simplified as
B(g)+ A k←→a kd
A− B(adsorbed)−→Ak B(reacted) (11)
(semiconductor-type) kinetics concept (the adsorption of electron-drawing gases will localize the conducting radi-cals)[23], and
G = P × CA (12)
where P is a proportional constant.
(v) In order to quantify the partial irreversibility, the competi-tion factor, fC, is defined as
fC= desorption rate reaction rate = kdCA−Ba kCA−Ba =kd k (13)
when fC 1 (or kdis two orders higher than k, says fC> 100,
the reaction term can be neglected), Eq.(11)is simplified to a purely reversible adsorption/desorption case, such as Eq.
(10); and when fC 1 (or kdis two orders smaller than k,
says fC< 0.01, the desorption term can be neglected), Eq.
(11)is then simplified to a pure reaction case, as depicted in Eq.(9), and ka= kin this case.
When 0.01 < fC< 100, the desorption and reaction processes
are in competition with each other. Both desorption and reaction cannot be ignored and should be taken into account. The rate expressions for Eq.(11)are
dCA−B dt = kaCB0(CA0− CA−B− CAB) −kdCA−B− kCA−B (14) dCA−B dt = k C A−B (15)
where CB0is the concentration of incoming gas B, which is a
constant; CA0the initial surface site concentration of fresh site A; CA−Bis the surface concentration of adsorbed sites; CAB
is the surface concentration of reacted sites; CA is the fresh
site concentration (=CA0− CA−B− CAB); ka, kd, and kare the
adsorption, desorption, and reaction rate constants, respectively. The schematic expression for a competitive desorption/reaction model is shown inFig. 8.
In order to solve Eqs.(14) and (15), two simplified forms are written as
C1= −EC1− FC2+ I (16)
Fig. 8. The schematic representation for a competitive reaction/desorption model.
where C1= CA−B, C2= CAB, E = (kaCB0+ kd+ k), F = kaCB0, I = kaCB0CA0, H = k, CA0and CB0are constants. By assuming
the initial conditions as C1(t = 0) = C10and C2(t = 0) = C20, Eqs.
(16) and (17)are solved by the Laplace transformation, and the solutions can be classified into two cases:
(a) Adsorption process (CB0> 0)
C1= C10exp −E 2t cosh ⎛ ⎝ E2 4 − FHt ⎞ ⎠ +(I − FC 20− (EC10/2)) (E2/4) − FH exp −E 2t × sinh ⎛ ⎝ E2 4 − FHt ⎞ ⎠ (18) C2= I F+ C20−I F exp −E 2t cosh ⎛ ⎝ E2 4 −FHt ⎞ ⎠ +(C10H + (C20E/2) − (IE/2F )) (E2/4) − FH exp −E 2t × sinh ⎛ ⎝ E2 4 − FHt ⎞ ⎠ (19) (b) Desorption process (CB0= 0)
Eqs. (18) and (19) can be further simplified when the desorption process starts by letting I = 0 = F, I/F = CA0, and E = kd+ k(because of CB0= 0). One obtains
C1= C10 exp[−(kd+ k)t] (20) C2= C20+ k
C10 kd+ k
[1− exp[−(kd+ k)t]] (21) The totally reversible case is obtained by letting k(or H) = 0,
C2= 0, C20= 0, E = kaCB0+ kd. Eqs.(18) and (20)can be reduced
to C1= I E+ C10− I E exp(−Et) (22) C1= C10 exp[−kdt] (23)
3.3. Results and data fitting
Both the experimental conductance transient and the fit-ted conductance transient, along with the fitfit-ted parameters, are shown in Fig. 7. Where CA0 is set to be 100 (real value
is unknown); P = 0.0312, which is the proportional constant disscussed in Eq. (12). The obtained fitting parameters are:
ka= 1.0× 10−5, kd= 7.5× 10−3, k= 2.0× 10−4, the
competi-tion factor, fc, is 37.5, the range at which the competition of
reaction and desorption does exist. It should be noticed that each adsorption/desorption step (every 10 min each step inFig. 7) is fitted separately using the same set of ka, kd, and k. That is, the
value of C1(CA−B) and C2(CAB) at the end of one section are
the “initial” values (C10 or C20) for the next step. The values
of C1and C2for the first step (2% O2adsorption) are set to be
zero.
The model prediction fits the experimental data very well when operating under a lower concentration of O2. The
devia-tion begins to enlarge when the O2concentration is higher than
8%. Taking the final run (21% O2) as an example, the actually
reacted amount is represented by segement α inFig. 7, however, the predicted value from the model is indicated by segement γ. On the other hand, the actually adsorbed amount is indicated by segement β, while the prediction from the model is segement δ. It can be found that the experimental amounts for both adsorp-ton and reaction are deviated from the model prediction at a higher O2 concentration, especially noticeable for the
adsorp-tion process. It is possible that the numbers of adsorbing site are insufficient for further adsorption when exposed to a higher con-centration of O2. Smaller deviation is observed for the reaction
process involves, and the actual amount of reaction is smaller than that of the predicted one (α < γ), which can be explained by the modified tarnishing film model proposed in our previous work[9]. Since the reacted films are porous and the diffusion (or penetration) is dominant, as shown inFig. 8, assumption (i) is no longer applied, thus explaining the deviation.
4. Conclusions
In this work, the irreversible behavior observed in NO2
sensing with a vacuum evaporated TTF–TCNQ thin film is char-acterized and modelled. Both conductance and mass transients suggest the irreversibility has something to do with NO2
immo-bilization. With the estimated diameters of gas molecules based on the kinetic theory of gas, the NO2most likely reacts with the
complex before diffusing through the film is proposed. Accord-ing to UV–vis and IR spectra data, as well as information from literatures, a net equation is proposed. In addition, XRD inten-sity, decreasing after NO2exposure, helps to explain the ongoing
reaction that destroyed the crystal packing of the complex. Among all information available, we proposed a hypothetical
This work was sponsored by the National Research Council of the Republic of China (Taiwan) under grant numbers NSC 92-2214-E-002-037 and NSC 93-2214-E-002-020.
References
[1] J. Ferraris, D.O. Cowan, V. Walatka Jr., J.H. Perlstein, Electron transfer in a new highly conducting donor–acceptor complex, J. Am. Chem. Soc. 95 (1973) 948–949.
[2] C. Jouve, D. Jullien, B. Remaki, Conductive polyethylene as sensitive layer for gas detection, Sens. Actuators B 28 (1995) 75–80.
[3] R.J. Ewen, C.L. Honeybourne, European Patent, 0,490,572A1 (1992). [4] S. Clemendot, M. Vandevyver, J.-P. Bourgoin, G. Derost, A.
Ruaudel-Teixier, A. Barraud, Evaluation of conducting LB films based on ethylenedithiotetrathiafulvalene (EDT-TTF) derivatives for phosphine sensing, Sens. Actuators B 6 (1992) 197–201.
[5] L. Henrion, G. Derost, A. Ruaudel-Teixier, A. Barraud, Investigation of n-docosyl pyridinium, TCNQ Langmuir–Blodgett films as gas sensors, Sens. Actuators 14 (1988) 251–257.
[6] H. Perez, W. Budach, S. Palacin, G. Derost, A. Barraud, Nitrogen dioxide detection using low-conducting Langmuir–Blodgett films, Sens. Actuators B 26-27 (1995) 140–143.
[7] W.R. Hertler, H.D. Hartzler, D.S. Acker, R.E. Benson, Substituted quinodimethans. III. Displacement reactions of 7,7,8,8-tetracyanoquinodimethan, J. Am. Chem. Soc. 84 (1962) 3387– 3393.
[8] M.R. Suchanski, R.P.V. Duyne, Resonance Raman spectroelectrochem-istry. IV. The oxygen decay chemistry of the tetracyanoquinodimethane dianion, J. Am. Chem. Soc. 98 (1976) 250–252.
[9] K.C. Ho, J.Y. Liao, NO2 gas sensing based on vacuum-deposited
TTF–TCNQ thin films, Sens. Actuators B 93 (2003) 370–378.
[10] T.H. Chen, B.H. Schechtman, Preparation and properties of polycrystalline films of (TTF)(TCNQ), Thin Solid Films 30 (1975) 173–188.
[11] T.J. Kistenmacher, T.E. Phillips, D.O. Cowan, The crystal structure of the 1:1 radical cation-radical anion salt of 2, 2-bis-1,3-dithiole (TTF) and 7,7,8,8-tetracyanoquinodimethane (TCNQ), Acta Cryst. B 30 (1974) 763–768.
[12] D.R. Lide, CRC Handbook of Chemistry and Physics, 81st ed., CRC Press, 2001, 6–46.
Weinheim, 1988.
[19] S. Matsuzaki, N. Koga, I. Moriyama, IR spectra of solid state reaction products of TTF and iodine, Bull. Chem. Soc. Jpn. 56 (1983) 2090–2092. [20] T. Hanawa, S. Kuwabata, H. Yoneyama, Gas sensitivity of polypyrrole films
to NO2, J. Chem. Soc., Faraday Trans. 1 (84) (1988) 1587–1592.
[21] C.D. Jaeger, A.J. Bard, Electrochemical behavior of donor-tetracyanoquinodimethane electrodes in aqueous media, J. Am. Chem. Soc. 102 (1980) 5435–5442.
[22] A. Figueras, J. Caro, J. Fraxedas, V. Laukhin, TTF–TCNQ thin films grown by organic CVD: texture and transport properties, Synthetic Met. 102 (1999) 1611–1612.
[23] C.J. Liu, J.C. Hsieh, Y.H. Ju, Response characteristics of lead phthalo-cyanine gas sensor: effect of operating temperature and postdeposition annealing, J. Vac. Sci. Technol. A 14 (3) (1996) 753–756.
Biographies
Jung-Yu Liao received his BS and PhD degrees in chemical engineering from
National Taiwan University, Taipei, Taiwan, in 1999 and 2004, respectively. His research interest surrounds thin film process in general and gas sensors in particular. Currently, he is a researcher working on organic light emitting diodes (OLED) at the Material and Chemical Research Laboratories, Industrial Technology Research Institute (ITRI), Chutung, Hsinchu, Taiwan.
Kuo-Chuan Ho received BS and MS degrees in chemical engineering from
National Cheng Kung University, Tainan, Taiwan, in 1978 and 1980, respec-tively. In 1986, he received the PhD degree in chemical engineering at the University of Rochester. The same year he joined PPG Industries, Inc., first as a senior research engineer and then, from 1990 until 1993, as a research project engineer. He has worked on the electrochemical properties of vari-ous electrode materials. He has applied surface science/interfacial engineering and electrochemistry/electrochemical engineering principles for improving the performances of electrochemical devices, including electrochemical sensors, electrochromic devices, and dye-sensitized solar cells, with emphasis on improv-ing their electrochemical, at-rest, and thermal stabilities. Followimprov-ing a 6-year industrial career at PPG Industries, Inc., he joined his alma mater at National Cheng Kung University in 1993 as an Associate Professor in the Chemical Engineering Department. In 1994, he moved to the Department of Chemical Engineering at National Taiwan University. Currently, he is a Professor jointly appointed by the Department of Chemical Engineering and Institute of Polymer Science and Engineering at National Taiwan University.
