Thermally induced J-band narrowing in merocyanine LB films

Thermally induced J-band narrowing in merocyanine LB films

Colloids and Surfaces A: Physicochem. Eng. Aspects 284–285 (2006) 509–513 Thermally induced J-band narrowing in merocyanine LB films Junpei Miyata, S...

212KB Sizes 0 Downloads 15 Views

Colloids and Surfaces A: Physicochem. Eng. Aspects 284–285 (2006) 509–513

Thermally induced J-band narrowing in merocyanine LB films Junpei Miyata, Shin-ichi Morita 1 , Yasuhiro F. Miura, Michio Sugi ∗ Graduate School of Engineering, Toin University of Yokohama, 1614 Kurogane-cho, Aoba-ku, Yokohama 225-8502, Japan Received 20 July 2005; received in revised form 17 October 2005; accepted 28 October 2005 Available online 2 December 2005

Abstract We have examined the effect of heat treatments (HT) at temperatures below 90 ◦ C by comparing UV–vis spectra of meorcyanine (MS)-cadmium arachidate (Cd-C20 ) mixed Langmuir–Blodgett (LB) films before and after the application of HT. It has been found that narrowing of the red-shifted J-band is thermally induced in the mixed LB films by HT at temperatures ranging from 60 ◦ C to 70 ◦ C. The J-band is centered at 590 nm in the as-deposited state, and it is associated with a prominent in-plane anisotropy. After HT, the band becomes sharper in spectral shape with its peak further red-shifted to 596 nm, and the in-pane anisotropy is drastically decreased at the same time. We have not found the original and the thermally induced bands coexisting in one and the same spectrum. It is suggested that the observed thermal annealing effect is due to a total reorganization of the MS chromophore arrangements rather than an additional growth of the original MS J-aggregates. The reorganization of MS chromophores seems to be closely related with the microbrownian motion of alkyl chains that is plausibly assumed to be activated in the present temperature range. The mild heat treatments as shown here is characterized as a purely physical process, and will present another method of modifying the properties of LB films with a lower risk of degradations compared with those chemical processes such as acid or basic treatments which are so far known to be effective for controlling the aggregation state. © 2005 Elsevier B.V. All rights reserved. Keywords: Merocyanine dye; Langmuir–Blodgett films; J-aggregate; Heat treatment; Annealing

1. Introduction A surface-active derivative of meorcyanine dye (MS shown in Fig. 1) is easily incorporated into monolayers when it is mixed with arachidic acid (C20 in Fig. 1) [1,2]. The monolayers formed on a Cd2+ -containing subphase are easily deposited onto substrates as Y-type Langmuir–Blodgett (LB) films by employing the conventional vertical dipping technique. A vast literature has been accumulated on the MS-C20 mixed LB films as model systems for studying optical, photoelectric and other physical properties [3–10], and also for trial manufacture of device prototypes [11–13]. The mixed LB films are known to be phase-separated into MS-rich and C20 -rich domains. The MS-rich domains are associated with a red-shifted band with its peak located around 590 nm, while the C20 -rich domains are transparent to visible light. The 590-nm band, which gives a blue color to the films, is iden∗

Corresponding author. Tel.: +81 45 974 5043; fax: +81 972 5972. E-mail address: [email protected] (M. Sugi). 1 Present address: Department of Chemistry, The University of Georgia, GA 30602-2556, USA. 0927-7757/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2005.10.056

tified as a J-band characterized by a remarkable red shift and narrowing of the peak with an intense fluorescence with zero or small Stokes shift. Here, the J-band is interpreted as originating from a low-dimensional array, often referred to as a J-aggregate, composed of dipole-coupled chromophores [14]. The aggregation state of MS in the films is found to be modified by various secondary treatments applied in both liquid and gas phases, resulting in drastic changes in the associated properties [6,15,16]. For thicker films with 10 or more monolayers, e.g., heat treatments (HT) as well as acid treatments (AT) dissociate the J-aggregates with the film color changing from blue to red, and basic treatments (BT) lead to restoration of the aggregates with the color returning to blue [6,16]. The changes in photoelectric properties has been also reported, showing that the lateral photoconductivity is enhanced more than a decade by successive application of AT and BT in gas phase [6,16]. As reported in previous papers, the J-band disappears in the spectrum for the thicker film case when HT at 90 ◦ C for 30 min is applied [6,16]. The film looks now red with a broad band centered around 530 nm, indicating that the J-aggregate is dissociated into monomeric state. The X-ray diffraction pattern after HT is, however, sharper in shape than that in the as-deposited

510

J. Miyata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 284–285 (2006) 509–513

Fig. 1. Chemical structures of film-forming materials: merocyanine dye (MS), arachidic acid (C20 ), and its deuterium derivative (C20 -d).

state before HT [17]. The results indicate that HT at 90 ◦ C has an annealing effect to enhance the ordering of Cd2+ ions at the expense of the dissociation of J-aggregates, since the diffraction pattern represents the ordering of Cd2+ -ions at each hydrophilic interface. Another remarkable feature reported for HT at 90 ◦ C is that the additivity law of total heating time does not hold: 1 × HT (30 min) and 2 × HT (15 min) share common spectral shape with each other [6]. In the spectra for 3 × HT (10 min) and 6 × HT (5 min), however, the J-band components are recognized as a shoulder and a peak, respectively, indicating the increased heat-tolerance of J-aggregates for which the intervening cooling down to room temperature should be responsible. There arises a question as to what happens at temperatures below 90 ◦ C with respect to the apparently enhanced heattolerance of J-aggregate. We have been examining the effect of “mild” HT at temperatures between room temperature and 90 ◦ C, where we should explicitly allow for the influence of microbrownian motion of alkyl chains [18,19]. In the range of 60–70 ◦ C, we have actually found that the original J-aggregate is reorganized by HT to form a novel phase of aggregate associated with a prominent band, which is sharper in shape and narrower in width than the original J-band in the as-deposited films. Here, it is noted that no report has yet come to our knowledge as to the heat treatment that exhibits an annealing effect to enhance the ordering of merocyanine chromophores caused by the heat treatments. The results will be reported in the present paper. 2. Experimental Surface-active meorcyanine MS, arachidic acid C20 and its deuterium derivative (C20 -d in Fig. 1) were purchased from Hayashibara Biochemical Laboratories Inc. Kankoh-Shikiso, Fluka AG and C/D/N Isotopes, respectively, and used without further purification. MS and C20 (or C20 -d) were dissolved in chloroform with the molar mixing ratio of [MS]:[C20 ] = 1:2. The concentration of MS was of the order of 10−4 M. The LB films were prepared using the standard vertical dipping method. The aqueous subphase and the deposition condition were the same as given in a previous paper [10]. The substrates, each one-fourth of an ordinary slide glass, were

soaked in an ethanol solution of pro analysi grade KOH for 4–6 h, and then rinsed with pure water. Immediately after rinsing, the substrates were precoated with five monolayers of Cd salt of pure C20 to make their surfaces hydrophobic. Ten monolayers of the MS-C20 mixture were then deposited onto the substrates. The films were of Y-type with a deposition ratio of approximately unity. For the heat treatment, each LB sample was sealed in an aluminum tube with a screw top at one end (ca. 20 mm in diameter and 150 mm long). The tube was then immersed in a water bath kept at a constant temperature. After a planned heating time, it was pulled out of the bath to cool down to room temperature. UV–vis and infrared absorption spectra A|| and A⊥ of the LB films were measured at room temperature using a Shimadzu UV-2100 and a JASCO FT/IR-300, respectively, where A|| and A⊥ refer to absorbances taken using linearly polarized light incidents with the electric vector parallel and perpendicular to the dipping direction in the deposition process, respectively. 3. Results 3.1. UV–vis spectra The results for HT at 90 ◦ C were consistent with those reported previously. Introduction of polarized incidents, however, adds novel information about the in-plane anisotropy to the previous results. Fig. 2(a) exemplifies the spectra of an LB film before and after HT at 90 ◦ C for 30 min. The as-deposited film is blue in color associated with a J-band located at 590 nm with a dichroic ratio R = A|| /A⊥ appreciably larger than unity (R = 1.84 for this example). After HT, the 590-nm band disappears, indicating that the J-aggregates in the MS-rich domains are dissociated. At the same time, the in-plane anisotropy drastically decreases down to R ≈ 1, and the A|| and A⊥ spectra are now approximately identical with each other. A broad band located around 530 nm in each spectrum is plausibly assigned to MS monomers [6]. The film turns from blue to red in color as a consequence. It should be noted that neither the crossing point of both A|| curves nor that of A⊥ ones corresponds any more to the isosbestic point that is correctly given as the crossing point of the absorbance curves for the non-polarized light. An HT at 80 ◦ C for 30 min also resulted in the dissociation of the J-aggregate and the appearance of the isotropic 530-nm band as is the case with HT at 90 ◦ C. A different effect on the MS aggregation state has been observed for the mild HT below 80 ◦ C. As shown in Fig. 2(b), a narrowing of red-shifted band with an increased peak height is observed after HT at 70 ◦ C for 15 min, indicating the enhanced order in the chromophore arrangement. The in-plane anisotropy disappears at the same time, with the A|| and A⊥ peaks slightly red shifted to 596 nm. Besides the red-shifted band, each spectrum after HT involves another component centered around 530 nm, which is plausibly assigned to monomers as already mentioned. Fig. 2(c) exemplifies the changes in A|| and A⊥ spectra caused by HT at 60 ◦ C for 30 min. A similar narrowing to the 70 ◦ C-HT

J. Miyata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 284–285 (2006) 509–513

Fig. 2. Absorption spectra A|| (—) and A⊥ (· · ·) of [MS]:[C20 ] = 1:2 10-layered LB films before and after a heat treatment (HT) at 90 ◦ C for 30 min (a), HT at 70 ◦ C for 15 min (b), and HT at 60 ◦ C for 30 min. A|| and A⊥ in the as-deposited state for each case are also shown by the thin solid and the thin dotted lines, respectively.

case is observed. The peak height is, however, appreciably lower than that for HT at 70 ◦ C for 15 min, and the in-plane anisotropy is still recognized although it decreases drastically compared to the as-deposited state (from R = 1.70 down to 1.14 at peaks in this case). The fraction of monomeric component may be larger in this case than that in the 70 ◦ C-HT case as suggested by the different spectral shapes between both cases at shorter wavelengths. The changes in A|| and A⊥ spectra were traced employing one and the same sample by successive application of HT cycles, each consisting of HT at a given temperature for the same length of heating time followed by cooling down to room temperature. We have found that the additivity law of total heating time does not hold for the present annealing effect. The spectral shape after the first HT cycle remained unchanged for further application of HT cycles, indicating that the growth of 596-nm band is hindered after the first cooling. We have also examined the effect of the length of heating time when one single HT was applied to the film. Fig. 3 shows the spectra for HT at 60 ◦ C for various heating times taken on samples prepared in one and the same batch. The spectra for the 15-min case present a contrast to those for other three cases with

511

Fig. 3. Absorption spectra A|| (—) and A⊥ (· · ·) of [MS]:[C20 ] = 1:2 10-layered LB films after heat treatments (HT) at 60 ◦ C for 0 min, 15 min, 30 min, 45 min and 60 min.

longer heating times. After HT at 60 ◦ C for 15 min, the band is shifted towards longer wavelengths, and its peak is now centered at 594 nm. The bandwidth is at the same time subjected to an appreciable narrowing with a decreased in-plane anisotropy. The spectra for 30, 45 and 60 min share common features in magnitude and shape. The 596-nm band is therefore fully developed for these three cases, indicating that the annealing process is completed within a range 15 min < tH < 30 min, where the 15min case can be interpreted as representing the transient stage of annealing. Similar experiments were carried out for evaluating tH for 70 ◦ C, and we have obtained a range 6 min < tH < 12 min. It is noted that broadening or splitting around the peak should be seen in the spectrum if the original 590-nm band and the heatinduced 596-nm band (and the transient 594-nm band as well) coexist in one and the same spectrum. In this respect, no evidence of the coexistence has been recognized among more than four hundred spectra we have so far measured. 3.2. Infrared spectra: a preliminary measurement The change in the alkyl chain conformation associated with HT can be separately examined by introducing the deuterium derivative C20 -d [17,20,21]. Fig. 4 shows the polarized infrared (IR) absorption spectra before and after HT at 70 ◦ C for 30 min in the range of 3000–2800 cm−1 , where all the C20 components

512

J. Miyata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 284–285 (2006) 509–513

4. Discussion We have described the thermally induced J-band narrowing in MS-C20 mixed LB films observed at temperatures below 90 ◦ C. As described in Section 3.1, the additivity law of heating time does not hold for the present annealing effect, where the spectral shape after the first HT cycle remains unchanged for further application of the cycles. This indicates that the cooling down to room temperature hinders further growth of the heat-induced 596-nm band. The in-plane anisotropy is drastically decreased by HT, and no evidence has been so far found for coexistence of the 590-nm and the 596-nm bands. The 596-nm band is therefore suggested to be due to a complete reorganization to form another crystalline phase of MS chromophores rather than a result of additional growth of the original J-aggregate. We consider that the present reorganization process should involve at least three stages as follows:

Fig. 4. Infrared absorption spectra A|| and A⊥ of an [MS]:[C20 -d] = 1:2 10layered LB film deposited on a substrate precoated with C20 -d monolayers before and after heat treatments (HT) at 70 ◦ C for 30 min.

were replaced by C20 -d. No systematic difference in the UV–vis spectra has been observed between the MS-C20 and the MSC20 -d samples. We clearly recognize the antisymmetric and the symmetric modes, νas (2920 cm−1 ) and νs (2850 cm−1 ) of CH2 stretching vibration in the C18 H37 chain attached to an MS choromophore. The occurrence of both bands is reminiscent of the spectra of the Cd-C20 and the C20 free acid LB films in both relative magnitude and spectral shape [17]. The third prominent band observed for the Cd-C20 and the free acid films (2954 cm−1 , typically) is assigned to the νas stretching mode of CH3 [17]. In the present spectra, however, the corresponding band is hardly recognized, hidden by relatively high noise levels. The νas (CH2 ) band for the as-deposited film exhibits a slight in-plane anisotropy R ≈ 0.8, while the νs (CH2 ) band with 0.9 < R < 1.0 is almost isotropic. The anisotropy tends to disappear after HT (1.0 < R < 1.1 for both bands) in accordance with the UV–vis spectra we have seen in Figs. 2 and 3. The anisotropy observed only for the νas (CH2 ) suggests that there exists a specific conformation preferred by the C18 H37 chain attached to each MS chromophore, since the transition dipole moments of νas (CH2 ) and νs (CH2 ) are orthogonal to each other. The rotation around the long axis of an extended alkyl chain should lead to the same R-value for both modes, and the statistical occurrence of gauche conformation should lead to R = 1. The apparent absence of νas (CH3 ) band in the present spectra may also suggest the preferred conformation of the MS alkyl chains. It should be mentioned that the present reorganization process is found to be sensitive to the substrates used, and that we have not yet been successful in securing sufficient reproducibility when we use such substrates as CaF2 plates that are more appropriate for IR measurement.

(1) The original J-aggregates, each involving several hundred MS molecules, are thermally activated into a transient state in which the aggregates may dissociate to release MS monomers; (2) The remaining fraction of J-aggregates in the transient state is then subjected to a phase transition to change into another crystalline state associated with the 596-nm band; (3) Each aggregate with the 596-nm band grows in size by merging monomeric and oligomeric clusters in the neighborhood to form the fully developed 596-nm aggregate. Stage (1) may be completed within tH = 15 min for 60 ◦ C and 6 min for 70 ◦ C. Stage (2) as a phase transition may have much faster reaction rates than those of Stage (1), while the rates for Stage (3) should correspond to the tH -values: 15 min < tH < 30 min and 6 min < tH < 12 min for 60 and 70 ◦ C. As is well known, the assemblies of alkyl chains may retain the intramolecular mobility above the glass transition temperature that can be far below the room temperature as typically seen in the rubber-like materials [4]. Such mobility, referred to as the microbrownian motion, will trigger the present reorganization process. Although the thermal behavior of the present MS-C20 mixed LB films is open for a thorough investigation, there is a considerable literature on the thermal behavior of LB films of pure Cd salts of fatty acids. In the present mixed system, the monolayers are phase-separated into the MS-rich and C20 -rich domains [4], the latter of which should reflect the properties of pure CdC20 LB films. In this respect, it has been revealed by X-ray diffraction measurement that the Cd–Cd spacing in the Cd-C20 LB films tends to decrease with temperature from ca. −40 ◦ C on, suggesting that the alkyl chains are subjected to swaying motion due to the increasing fraction of gauche conformations with temperature [18]. It has been also revealed that the degree of order of Cd2+ lattice remains unchanged up to ca. 80 ◦ C [15,18]. These remarks on the Cd-C20 system have been later supported by an infrared spectroscopic study including the same Cd-C20

J. Miyata et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 284–285 (2006) 509–513

LB system by Naselli et al. [19]. These remarks suggest that the swaying motion of alkyl chains is relevant to the present reorganization process observed in the range of 60–70 ◦ C. The role of Cd2+ ions should be emphasized as a stabilizing component to hold the LB lamellar structure by anchoring one end of each alkyl chain to a hydrophilic interface. The alkyl chains are allowed to sway like reeds in the wind below the macroscopically observed melting point that should correspond to the dissociation of Cd2+ -ion lattice. The previously estimated value for Cd-C20 LB film, 100–120 ◦ C [22], is comparable to 110 ◦ C for Cd-C20 bulk crystal obtained by Naselli et al. employing differential scanning calorimetry (DSC) [19]. These values are actually higher than the melting points of the related bulk crystals with smaller binding energies, e.g., 75.5 ◦ C for C20 free acid with hydrogen bonds, and 36–37 ◦ C for icosane (CH3 (CH2 )18 CH3 ) known as a van der Waals solid [23]. Here, the binding energies are typically of the order of 102 , 10 and 1 kJ/mol for ionic bonds, hydrogen bonds and van der Waals forces, respectively [24]. As to the IR spectra, the anisotropy seen only for νas (CH2 ) and the apparent absence of νas (CH3 ) (cf. 3.2) suggest, however, that there is a specific conformation preferred by the C18 H37 sidechain of an MS in the original J-aggregate, although we may say nothing decisive in the present stage due to the substratesensitive nature of the present annealing process. 5. Concluding remarks We have reported the annealing effect caused by the mild heat treatments below 90 ◦ C detected by means of UV–vis spectroscopy. The MS chromophores are reorganized by HT at 60 ◦ C and 70 ◦ C, resulting in a novel phase of MS-aggregate with its spectra sharper in shape than those of the original J-band. We have proposed a tentative model consisting of three stages, activation, phase transition and growth, to explain the present annealing process. It is suggested that the reorganization process of MS chromophores is, in each stage, closely related with the swaying motion of either or both of the alkyl species of MS and C20 in the LB system. The contribution from each species will be separately evaluated by means of IR spectroscopy by introducing the deuterium derivative C20 -d. The preliminary measurement of IR spectra on the MS-C20 -d samples has indicated that there is a preferred conformation of the alkyl chains attached to MS chromophores in the as-deposited state. The substrate-sensitive nature of the present process should be solved as a prerequisite for observing the IR spectra of CD2 stretching modes with a sufficient accuracy. The study along this line is now in progress and the results will be reported elsewhere. The annealing process described here is a physical one, in which the lower risk of degradation of films is expected in com-

513

parison with those chemical processes such as acid or basic treatments. Further, the exchange of molecules and ions is apt to occur between a sample and environment in the chemical processes. In these respects, the present process based on the mild heat treatment will present another method of effective modification of the properties of LB films. Acknowledgements The authors express their thanks to Mr. S. Mouri for his technical assistance in the present experiment. They are also indebted to Dr. Y. Hirano for his valuable comments in the early stages. References [1] M. Sugi, S. Iizima, Thin Solid Films 68 (1980) 199. [2] M. Sugi, T. Fukui, S. Iizima, K. Iriyama, Mol. Cryst. Liq. Cryst. 62 (1980) 165. [3] M. Sugi, M. Saito, T. Fukui, S. Iizima, Thin Solid Films 88 (1982) L15. [4] M. Sugi, M. Saito, T. Fukui, S. Iizima, Thin Solid Films 99 (1983) 17. [5] S. Kuroda, M. Sugi, S. Iizima, Thin Solid Films 99 (1983) 21. [6] M. Sugi, M. Saito, T. Fukui, S. Iizima, Thin Solid Films 129 (1985) 15. [7] S. Kuroda, K. Ikegami, M. Sugi, S. Iizima, Solid State Commun. 58 (1986) 493. [8] H. Nakahara, K. Fukuda, D. M¨obius, H. Kuhn, J. Phys. Chem. 90 (1986) 6144. [9] S. Nishikawa, Y. Tokura, T. Koda, K. Iriyama, Jpn. J. Appl. Phys. 25 (1986) L701. [10] Y. Hirano, T.M. Okada, Y.F. Miura, M. Sugi, T. Ishii, J. Appl. Phys. 88 (2000) 5194. [11] M. Saito, M. Sugi, T. Fukui, S. Iizima, Thin Solid Films 100 (1983) 117. [12] K. Sakai, M. Saito, M. Sugi, S. Iizima, Jpn. J. Appl. Phys. 24 (1985) 865. [13] M. Sugi, K. Sakai, M. Saito, Y. Kawabata, S. Iizima, Thin Solid Films 132 (1985) 69. [14] See, for example T. Kobayashi (Ed.), J-aggregates, World Scientific, Singapore, 1996. [15] T. Fukui, M. Saito, M. Sugi, S. Iizima, Thin Solid Films 109 (1983) 247. [16] M. Saito, M. Sugi, K. Ikegami, M. Yoneyama, S. Iizima, Jpn. J. Appl. Phys. 25 (1986) L478. [17] M. Saito, Y. Tabe, K. Saito, K. Ikegami, S. Kuroda, M. Sugi, Jpn. J. Appl. Phys. 29 (1990) L1892. [18] T. Fukui, M. Sugi, S. Iizima, Phys. Rev. B22 (1980) 4898. [19] C. Naselli, J.F. Rabolt, J.D. Swallen, J. Chem. Phys. 82 (1985) 2136. [20] Y. Hirano, S. Morita, Y.F. Miura, M. Sugi, Thin Solid Films 438–439 (2003) 225. [21] S. Morita, Y.F. Miura, M. Sugi, Y. Hirano, J. Appl. Phys. 94 (2003) 4368. [22] T. Fukui, A. Matsuda, M. Sugi, S. Iizima, Bul. Electrotech. Lab. 41 (1977) 423. [23] See, for example Encyclopedic Dictionary of Chemistry, Kagakudojin, Tokyo, 1989. [24] See, for example P.W. Atkins, The Elements of Physical Chemistry, 2nd ed., Oxford University Press, Oxford, 2001.