Observation of Molecular Complexes in Oligo-Phenylenevinylene (OPV) Organogels by Neutron Diffraction

1. Introduction

Organogels are fascinating systems as they form three-dimensional networks from solutions containing a low percentage of relatively small-sized molecules designated as organogelators or low-molecular weight gelators (LMWG) [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16]. In a recent monograph these molecules have been regarded as “chimeras” due to the fact they bear simultaneously different kind of interactions, such as hydrogen bonding, π-π stacking, van der waals, and the like [5]. Due to this unusual chemical architecture their crystallization habit favours one specific growth face, which results in the making of very long fibrillar entities that are more or less connected so as to produce an infinite network.

The knowledge of their molecular structure is a topical issue [5,6,7,8]. One specific point raised quite recently concerns the formation of molecular compounds, that is the co-crystallization of the organogelator and the solvent [5,6]. Molecular compound is the generic terminology for systems that combine two or more molecules under a well-defined stoichiometry whatever the way they interact. This term is customary in the nomenclature used in temperature-concentration phase diagrams [17]. In practice, several names can designate these systems depending on the type of interactions: crystallo-solvates, molecular complexes, inclusion compounds, intercalates, guest-host compounds, and the like [17,18,19]. Systems form through hydrogen bonds are rather designated as molecular complexes, a situation that will be encountered in this paper.

Several authors have already suspected from circumstantial evidence that the solvent is not only a simple diluent but plays a role in the gelation process [20,21,22,23,24]. The existence of molecular compounds has only been recently demonstrated by means of neutron diffraction while investigating the gelation of triaryl trisamide molecules with a series of organic solvents [24,25]. It has been further shown that the occurrence of molecular compounds has a direct bearing on the gel properties [26,27]. Neutron diffraction by making use of the isotopic contrast brought about by using either hydrogenous or deuterated species can unambiguously demonstrate the occurrence of molecular compounds [25].

Based on circumstantial evidence, we had already conjectured in previous papers that the occurrence of molecular complexes in some OPVs through hydrogen bonds should be contemplated [21]. Here, we present neutron diffraction data that definitely establish the existence of these complexes with benzyl alcohol and an OPV molecule bearing OH terminal groups (see Scheme 1). Conversely, we further show that no molecular complex occurs when the terminal group is replaced by a non-hydrogen forming group [28].

2. Results and Discussion

The oligo vinylene phenylene molecules investigated herein presented in scheme 1 are typical “chimera” molecules[5,28]. They consist of a central core made of phenyl groups linked by ethylene groups, and six aliphatic armes containing 16 carbons each. The structure is either completed with two CH₃-OH terminal groups that give a hydrogen bonding character to this molecule (OPVOH scheme 1a), and that mimic part of the benzyl alcohol or a NCOOC₂H₅ group (OPVR scheme 1b) free of hydrogen bond character [28].

The present study of the molecular strcuture has been carried out by X-ray and neutron diffraction in systems prepared from hydrogenous benzyl alcohol (C₇H₉OH, called BzOH_H) and deuterated benzyl alcohol (C₇D₉OH, called BzOH_D).

2.1. Thermodynamics

In a previous paper we had reported on the temperature-concentration phase diagram for OPVOH/benzyl alcohol and OPVOH/benzyl methyl ether [21]. The chemical structure of benzyl methyl ether (BME) is very close to that of benzyl alcohol except that the OH function is replaced by OCH₃, which impairs the formation of hydrogen bonds.

The large difference in melting temperature between OPVOH/BzOH_H and OPVOH/BME together with the diffference in melting/formation enthalpies led us to suspect that benzyl alcohol may form a molecular complex with OPVOH through the formation of hydrogen bonds (see Figure 1a) That the extrapolated melting enthalpy to C_OPVOH= 1 gives a values much larger than that measured in the solid state for OPVOH/BzOH gels as opposed to what is seen for OPVOH/BME supports the hypothesis of a differing molecular structure[21]. This is further observed for the OPVR/BzOH systems in Figure 1b, where the extrapolated enthalpy is found virtually identical to that in the solid state.

Yet, no direct demonstration had able us to confirm this conjecture so far. Note that this effect cannot be related to the solvent quality, which does not affect the value of the melting enthalpy.

The T-C phase diagram of OPVOH and OPVR in hydrogenous and deuterated benzyl alcohol is shown in Figure 1. The use of deuterated benzyl alcohol instead of the hydrogenous version does not alter the thermodynamic properties of the organogels. The formation and melting temperatures are identical as are the associated enthalpies. Comparing the neutron diffraction patterns obtained with hydrogenous benzyl alcohol and deuterated benzyl alcohol is therefore a legitimate procedure, something already well-documented [29].

To extend further this study, we have studied another OPV molecule, namely OPVR, where the chemical structure of the terminal group theoretically prevents the molecule to establish hydrogen bonds with benzyl alcohol (Scheme 1b). As can be seen in the melting temperature of OPVR in benzyl alcohol is lower than that of OPVOH. Unlike OPVOH, extrapolation of the melting enthalpy to C= 1 does give a value nearly identical to that measured in the solid state (Figure 1). In this system, the formation of an OPVR/benzyl alcohol complex is therefore not expected. Here too, there is no difference in the melting properties of OPVR organogels whether hydrogenous or deuterated benzyl alcohol are used (Figure 1).

2.2. Morphology

The morphology of these organogels is also strongly dependent upon the terminal groups. OPVOH/benzyl alohol have un unusual morphology of these gels, which is of the hub-like type with thin ribbon-like fibrils that radiate from well-identified centers Figure 2 top left). Conversely, OPVR/benzyl alocohol organogels display the commonly observed morphology, namely randomly-dispersed fibrils (Figure 2 bottom). Interestingly, drying OPVOH/ benzyl alcohol gels for SEM observation leads to an apparent collapse of the gel structure (Figure 2 top right) unlike what is seen with OPVR gels. This effect may arise from the dismantling of the OPVOH/benzyl alcohol molecular complex in the course of the drying process.

2.3. Neutron Diffraction for Studying Molecular Compounds

The main technique for deriving the molecular structure is radiation diffraction, and more commonly X-ray diffraction. Neutron diffraction offers further possibilities, especially in the case of molecular complexes thanks to the availability of the isotopic labelling [30,31,32,33,34,35].

In the case of neutron diffraction, the intensity is written [31]:

$$I\left(q\right)={\displaystyle \frac{1}{N^2}}{\sum }_i{\sum }_j{a}_i{a}_j{exp}^{iq{r}_{ij}},$$

(1)

where a_i and a_j are the neutron scattering length of species i and j and r_ij the associated distance.

For systems comprising two types of molecules, 1 and 2, (1) is written:

$$I\left(q\right)={\displaystyle \frac{A_1^2}{N_1^2}}\sum _i\sum _j{exp}^{iq{r}_{ij}}+{\displaystyle \frac{A_2^2}{N_2^2}}\sum _k\sum _l{exp}^{iq{r}_{kl}}+{\displaystyle \frac{2A_1{A}_2}{N_1{N}_2}}\sum _m\sum _{n\ne m}{exp}^{iq{r}_{mn}}$$

(2)

where A₁ and A₂ are the scattering amplitudes of molecules 1 and 2, that are calculated by summing the scattering lengths of their constituting atoms ($A={\sum }_s{a}_s$) [30]. It is understood that for the third term of relation 2 only the terms where A_m≠ A_n are considered. As opposed to X-rays, these scattering amplitudes do not depend on q as neutrons interact only with the nucleus [30,32], whose size is negligible with respect to the van der Waals radius. Relation (2) is rewritten:

$$I\left(q\right)=A_1^2{S}_1\left(q\right)+A_2^2{S}_2\left(q\right)+{2A}_1{A}_2{S}_{12}\left(q\right)$$

(3)

For systems composed of a crystallizable component C and of a solvent S, two cases can occur: a solid solution where the solvent is totally expelled out of the crystal, and a molecular complex where the co-crystallization of both species occurs. The latter are characterized by a stoichiometric composition C_γ, corresponding to the number of solvent molecules S per crystallizable molecules C.

In the case of a solid solution, the cross-term in relation (2) vanishes so that:

$$I\left(q\right)=A_C^2{S}_C\left(q\right)+A_S^2{S}_S\left(q\right)$$

(4)

Under these conditions, the diffraction pattern of the crystals from molecules C, I(q) does not depend upon the value of the scattering amplitude of A_S (no change in peak intensity nor in peak number).

In the case of molecular complexes, the cross-term cannot be neglected. Also, there are two types of solvent molecules if the concentration is below the stoichiometric composition, those remaining liquid and those participating in the co-crystals. The intensity is written:

$$I\left(q\right)=A_C^2{S}_C\left(q\right)+A_S^2{S}_S^{co}\left(q\right)+A_S^2{S}_S^{liq}\left(q\right)+{2A}_C{A}_S{S}_{CS}\left(q\right)$$

(5)

where $S_S^{co}\left(q\right)$is the diffraction by the solvent molecules that are part of the molecular complex and thus organized, while $S_S^{liq}\left(q\right)$corresponds to the diffraction of solvent molecules in the liquid phase. As a result, the diffracted intensity from the molecular complex reads:

$$I_{comp}\left(q\right)=A_C^2{S}_A\left(q\right)+A_S^2{S}_S^{co}\left(q\right)+{2A}_C{A}_S{S}_{CS}\left(q\right)$$

(6)

Therefore, modifying the scattering amplitude of the solvent molecules results in altering the diffraction pattern of the system, such as appearance or disappearance of peaks as well as modification in relative intensities [25,33]. Note that the use of isotopes does not change the molecular structure [25,32,33].

Neutron diffraction therefore allows one to distinguish qualitatively and straightforwardly a solid solution from a molecular complex thanks to the availability of hydrogenous and deuterated solvents.

2.4. Diffraction Data

The diffraction data have all been recorded in a domain of transfer momentum q ($q={\displaystyle \frac{4\pi }{\lambda }}\mathit{sin}\left({\displaystyle \frac{\theta }{2}}\right))$ λ= wavelength, θ= scattering angle) ranging from 0.5 to 10 nm^-1. As the solvent displays a flat scattering in this q-range, only the empty cell has been subtracted from the raw spectra.

There is a significant difference between the diffraction patterns recorded by X-ray and by neutrons for the OPVOH/benzyl alcohol systems as seen in Figure 3a. While there is only one peak at q= 1.405 nm^-1 for the X-ray pattern, other peaks appear in the neutron patterns. A weak peak is seen for OPVOH/BzOH_H at q= 2.329 nm^-1, against three additional peaks for OPVOH/BzOH_D at q= 2.329 nm^-1, q= 2.79 nm^-1 and q= 4.672 nm^-1 (see Table 1). Miller indices are assigned to the peaks by taking the π-π stacking direction as the c axis (Table 1). The relative intensities of the neutron diffraction peaks for each systems , I(q_i) in %, is calculated by:

$$I\left(q_i\right)={\sigma }_i/{\sum }_1^4{\sigma }_i$$

(7)

where σ_i is the surface of peak i as determined by means of a fit with a Lorentzian function:

$$I\left(q\right)=S_B+{\displaystyle \frac{2A}{\pi }}\left[{\displaystyle \frac{∆q}{4{\left(q-q_o\right)}^2+{∆q}^2}}\right]$$

(8)

where S_B the flat background (solvent+incoherent), A a constant, q_o the peak maximum and ∆q the full width at half maximum.

The difference in relative intensities between peaks from OPVOH/BzOH_H and OPVOH/BzOH_D allows one to conclude unambiguously that OPVOH and benzyl alcohol form a molecular complex.

The peak at q= 2.329 nm^-1 for OPVOH/BzOH_D corresponds to a distance d= 2.7 nm close to the spacing between the terminal OH groups (2.6 nm). It therefore corresponds to the 100 plane which contains the BzOH molecules. The other peaks are simply second orders of the 010 and 100 planes, which shows the high degree of organization within the OPVOH fibrils. A possible model is drawn in Figure 3a showing the way the benzyl acohol molecule can interact with OPVOH.

Figure 4. a) tentative monoclinic crystalline lattice for OPVOH/benzyl alcohol gels with a tentative positioning of the benzyl alcohol molecules; a= 2.7 nm, b= 4.89 nm, c= 0.35 nm, α= 79± 2°, β= γ= 90°. b) tentative triclinic crystalline lattice for OPVR/benzyl alcohol gels based on a previous paper [28] a= 2.8 nm, b= 4.89 nm, c= 0.49 nm, α= 79±2°, β= 42±2°, γ= 142±4°.

Figure 4. a) tentative monoclinic crystalline lattice for OPVOH/benzyl alcohol gels with a tentative positioning of the benzyl alcohol molecules; a= 2.7 nm, b= 4.89 nm, c= 0.35 nm, α= 79± 2°, β= γ= 90°. b) tentative triclinic crystalline lattice for OPVR/benzyl alcohol gels based on a previous paper [28] a= 2.8 nm, b= 4.89 nm, c= 0.49 nm, α= 79±2°, β= 42±2°, γ= 142±4°.

Drying OPVOH/benzyl alcohol gel, which results in the removal of the solvent from the crystalline lattice, is likely to introduce some local disorganization. This process may account for the observed collapse of the gel morphology.

The outcomes differ conspicuously in the case of OPVR/benzyl alcohol systems. As seen in Figure 3b the intensities of the neutron diffraction peaks and their positions are virtually identical whether deuterated or hydrogenous benzyl alcohol is used. This demonstrates the absence of a molecular complex in the case of OPVR.

Interestingly, a third peak is observed by X-ray diffraction which is conspicuously absent in the neutron diffraction patterns. This peak corresponds to the 100 plane containing the C₂H₅0 group. The neutron scattering amplitude of this group is calcutated $A_{C2H50}$through:

$$A_{\mathrm{C}2\mathrm{H}50}=\sum _{Atome}{a}_i$$

(9)

where a_i is the scattering amplitude of each atome constituting the group.

Equation (9) gives a value close to $A_{C2H50}=$0 thanks to the negative value of hydrogen a_H= -0.375 [30]. This therefore accounts for the absence of this peak in the neutron difffraction patterns.

The results are gathered in Table 2 where the peaks are indexed by again taking the π-π stacking direction as the c-axis. A tentative crystalline lattice is displayed in Figure 3b.

3. Conclusions

Results presented here provide a conclusive answer to a previous conjecture that predicted the occurrence of a molecular complex of OPVOH molecules in the organogelation process [21]. Conversely, formation of this complex depends upon the terminal group as shown with OPVR molecules. This may have a direct impact on the differing morphologies of these gels (see figure 2), hub-like type with thin ribbon-like fibrils in OPVOH/benzyl alcohol systems against randomly-dispersed type in OPVR/benzyl alcohol gels, as well as the collape of the gel morphology of the latter (figure 2) [36]. It is worth emphasizing that the existence of molecular complexes has also been recently demonstrated in the case of triaryl tris amide organogels [24]. The occurrence of molecular complex is likely to occur in the gelation process of many other systems, and is something to consider when determining the molecular structure, and in accounting for the thermodynamic properties.

4. Materials and Methods

Materials: The synthesis and properties of oligo (p-phenylene vinylene) gelator (OPVOH) is described in reference 17. The basic chemical structure of the molecule used in this study is shown in figure 1. The solvent used for this investigation were benzyl alcohol (BzOH), and benzyl methyl ether (BME). The hydrogenous solvents of high purity grade, were purchased from Aldrich, and were used without further purification. The deuterated benzyl alcohol, C7D7OH, was purchased from Eurisotop (Saclay, France) and used as-received. The preparation conditions for the neutron and the X-ray experiments are the same (quenching temperature, and ageing).

Differential Scanning Calorimetry: Diamond DSC from Perkin Elmer has been used for determining the gel formation and gel melting. Heating and cooling rates ranging from 5°C/min to 15°C/min were used. Approximately 30 mg of previously prepared gels were transferred into stainless steel sample pan. These pans were then hermetically sealed by an O-ring to prevent solvent evaporation.

X-ray diffraction: X-ray diffraction experiments are performed by means of a diffractometer developed at the Institut Charles Sadron in the DiffériX platform. The instrument operates with A monochromatic beam of wavelength λ = 0.154 nm and a hybrid photon counting detector (HPC-Dectrics Pilatus®3 R 300K). The distance sample-detector is set so as to access scattering vectors q= 4πsin(θ/2)/λ (θ= diffraction angle) ranging from q= 0.5 to 10 nm^-1. Calibration of the detector is performed with a silver behenate sample. The gel samples are placed in home-made sealed cells of adjustable thickness.

Neutron diffraction: The neutron diffraction experiments have been carried out on D16, a camera located at Institut Laue-Langevin (Grenoble, France) [37]. D16 is equipped with a new curved detector 38 cm high which covers a 86° solid angle with a high angular resolution of 0.075°. This detector is based on the Trench-MultiWire Proportional Chamber (MWPC) detector technology developed at ILL: 6 modules are mounted side by side in an 3He-filled curved vessel [38]. Each module consists of 192 cathode blades positioned every 2 mm, and 192 anode wires spaced by 1.5 mm [39]. The neutron wavelength is selected by beam reflection onto a focussing pyrolytic graphite monochromator. The present experiments were obtained with a neutron wavelength of λ= 0.449 nm, giving access to the following transfer momenta range of 0.6 nm^-1≤ q ≤ 10 nm^-1.

Optical Microscopy: Pictures were taken with a NIKON Optiphot-2 equipped with CCD camera by means of Nomarsky phase contrast. The samples were prepared by re-melting between glass-slides those gels obtained beforehand in a test-tube.

Scanning electron microscopy: The electron microscopy images were obtained on a JEOL (JSM 6700F) FESEM instrument. Hot solutions of OPV/benzyl alcohol were drop casted on silicon wafer and subsequently quenched to 00C in order to form the gel network. The samples were then dried under vacuum at room temperature. The dried gels were further coated by indium sputtering.