Hydrogen-Bonding versus van der Waals Interactions in Self

Nov 10, 2010 - Department of Chemistry, Princeton University, Princeton, New Jersey ... made available by participants in CrossRef's Cited-by Linking ...
0 downloads 0 Views 6MB Size
pubs.acs.org/Langmuir © 2010 American Chemical Society

Hydrogen-Bonding versus van der Waals Interactions in Self-Assembled Monolayers of Substituted Isophthalic Acids Pearl N. Dickerson, Amber M. Hibberd, Nuri Oncel,† and Steven L. Bernasek* Department of Chemistry, Princeton University, Princeton, New Jersey 08544, United States. †Current Address: Department of Physics and Astrophysics, University of North Dakota, Grand Forks, North Dakota 58202, United States. Received September 1, 2010. Revised Manuscript Received October 28, 2010 Self-assembled monolayers of a series of isophthalic acids (5-octadecyloxyisophthalic acid, 5-decyloxyisophthalic acid, 5-hexyloxyisophthalic acid, and 5-pentyloxyisophthalic acid) formed on highly ordered pyrolytic graphite (HOPG) at the solid-liquid interface were studied using scanning tunneling microscopy (STM). Although these molecules have the same dicarboxyl headgroup, their hydrocarbon tails are of different lengths. Hydrogen-bonding between headgroups and van der Waals interactions between the hydrocarbon tails control the final morphology of the monolayer. The STM images show that both van der Waals interactions (vdWs) and hydrogen-bonding (H-B) compete to control the structure, but the final structure of the monolayer is determined by balance between the two interactions.

Introduction In the field of nanotechnology, devices are being made with increasingly smaller features. Using lithography, complex structures can be made via a “top-down” approach, which recently has been reported to reach a patterning resolution of less than 5 nm.1 Alternatively, nanoscale devices can be made from a “bottom-up” approach by making use of the self-assembly (SA) of molecules. It is crucial to control the design and reproducibility of surface layers in order to make effective assembly templates. Understanding the mechanism of SA is essential to produce ordered surface structures at the molecular level. The many factors that control self-assembly fall into two main categories: molecule-substrate interactions and molecule-molecule interactions.2 This paper focuses on the influence of the moleculemolecule interactions, specifically hydrogen-bonding and van der Waals interactions between adsorbate molecules, on the morphology of self-assembled monolayers (SAMs). Compared to chemical bonds, hydrogen-bonding and van der Waals interactions are significantly weaker. Although the weakness of the interaction may be considered as a negative feature for various applications, a careful design based on the balance of interactions in the particular system may lead to the formation of highly stable SAMs. In addition, the relatively weak hydrogen-bonding and van der Waals interactions can lead to a self-correction mechanism to perfect the morphology of the film.3 Scanning tunneling microscopy (STM) has been used extensively to study the structures of monolayers of various organic molecules as it can be used to probe physisorbed monolayers on the nanoscale. Previous work has shown that, by using STM at the solid-liquid interface, information about SAMs can be obtained. For example, how varying functional groups on molecules affects SA was investigated.4 In addition, properties such as chirality have been extensively studied with STM. It was shown that in some cases enantiomorphous domains formed due to weak (1) Sreenivasan, S. V. MRS Bull. 2008, 33, 854. (2) Lindoy, L. F.; Atkinson, I. M. Royal Society of Chemistry, Self-assembly in supramolecular systems; Royal Society of Chemistry: Cambridge, UK, 2000. (3) Schneider, H. J.; Yatsimirsky, A. K. Principles and Methods in Supramolecular Chemistry; John Wiley: Chichester, UK, 2000. (4) De Feyter, S.; De Schryver, F. C. J. Phys. Chem. B 2005, 109, 4290. (5) Tao, F.; Bernasek, S. L. J. Phys. Chem. B 2005, 109, 6233.

Langmuir 2010, 26(23), 18155–18161

van der Waals interactions between molecules.5 Using STM to probe nanoscale structures has also been corroborated by ab initio calculations with good agreement between theoretical and experimental results.6 Self-assembled monolayers have been shown to create unique two-dimensional structures which have the potential to be used as molecular templates. Coadsorption of molecules led to the assembly of a stoichiometric two-component molecular mesh.7 Other two-dimensional structures have been observed such as organization of simple organic molecules such as long-chain alkanes, alcohols, and fatty acids,8 and also more complex molecules such as substituted porphyrin molecules.9,10 A recent study investigated the correlation between dipole interactions and SA, and concluded that while molecular physisorption is dominated by molecule-surface interactions, the assembly of molecules on flat surfaces such as highly ordered pyrolytic graphite (HOPG) is driven by weak intermolecular forces.11-13 Although there are numerous examples of complex self-assemblies observed by STM, the weak intermolecular forces that dominate SA are not well understood. This paper presents a systematic STM study of self-assembled monolayers formed by a series of substituted isophthalic acids, 5-octadecyloxyisophthalic acid (5-OIA), 5-decyloxyisophthalic acid (5-DIA), 5-hexyloxyisophthalic acid (5-HIA), and 5pentyloxyisophthalic acid (5-PIA) (Figure 1), on HOPG at the solid-liquid interface. These molecules have the same dicarboxylic acid headgroups, which allow for the same types of opportunities for hydrogen-bonding per molecule. Contributions from dipolar interactions11 between molecules will also be constant across the series, as all these molecules have ether linkages and carboxylic acid headgroups which will contribute similarly to the moleculemolecule interaction in these monolayer structures. However, the (6) Yang, T.; Berber, S.; Liu, J. F.; Miller, G. P.; Tomanek, D. J. Chem. Phys. 2008, 128, 124709. (7) Tao, F.; Bernasek, S. L. J. Am. Chem. Soc. 2005, 127, 12750. (8) Rabe, J. P.; Buchholz, S. Science 1991, 253, 424. (9) Oncel, N.; Bernasek, S. L. Appl. Phys. Lett. 2008, 92, 13305. (10) Oncel, N.; Bernasek, S. L. Langmuir 2009, 25, 9290. (11) Wei, Y. H.; Tong, W. J.; Zimmt, M. B. J. Am. Chem. Soc. 2008, 130, 3399. (12) Wei, Y. H.; Tong, W. J.; Wise, C; Wei, X; Armbrust, K; Zimmt, M. B. J. Am. Chem. Soc. 2006, 128, 13362. (13) Tong, W. J.; Wei, Y. H.; Armbrust, K; Zimmt, M. B. Langmuir 2009, 25, 2913.

Published on Web 11/10/2010

DOI: 10.1021/la103494g

18155

Article

Dickerson et al.

Synthesis of 5-Hexyloxyisophthalic Acid (5-HIA) and 5-Pentyloxyisophthalic Acid (5-PIA). To a solution of 1-

Figure 1. General chemical structure of the isophthalic acids used, where n = 18, 10, 6, and 5 (5-OIA, 5-DIA, 5-HIA, and 5-PIA, respectively).

molecules have tails with different hydrocarbon chain lengths. A comparative study of the morphology of the SAMs formed from these molecules is an ideal way to explore the balance between hydrogen-bonding and van der Waals interactions between adsorbate molecules during self-assembly in a system where molecule-substrate and molecule-molecule dipolar interactions remain similar across the series.

Experimental Section Scanning Tunneling Microscopy. A solution of each molecule dissolved in 1-phenyloctane (∼2 mM) was used. Phenyloctane (98%) and 5-OIA (98%) were purchased from Aldrich and used without further purification. The other isophthalic acids in the series were synthesized as described below. Samples were prepared by depositing the prepared solution on a freshly cleaved basal plane of HOPG and allowing enough time (∼5 min) for monolayer formation. Monolayer formation was assured in each case by measuring line profiles over the edge of the ordered domain onto the bare HOPG surface, determining that the layer was one molecule in thickness. All experiments were conducted under ambient conditions at the solid-liquid interface. Images were obtained in constant current mode and under various tunneling conditions (Vb = 0.8-1.2 V, It = 0.4-1.0 nA). All experiments were conducted using a Nanosurf AG EasyScan STM system (NanoScience Instruments, Liestal, Switzerland). The HOPG substrate (5 mm  5 mm, research grade) and wire for STM tips (Pt/Ir=80/20, 0.25 mm diameter) were purchased from NanoScience Instruments. Synthesis of 5-Decyloxyisophthalic Acid (5-DIA). To a solution of 1-bromodecane (Aldrich, 98%; 6 mmol; 1.24 mL) and K2CO3 (Aldrich, 98%; 6 mmol; 0.83 g) in 7.5 mL of acetone at room temperature under argon was added a solution of dimethyl5-hydroxy-isophthalate (Aldrich, 98%; 5 mmol; 1.05 g) in 7.5 mL of acetone via cannula. The resulting white suspension was heated to ca. 50 °C and stirred for 24 h. As after the 24 h period thin layer chromatography (TLC) indicated little conversion of starting material, NaI was added to facilitate the substitution reaction. After refluxing overnight, the reaction mixture was allowed to cool to room temperature, filtered through a plug of sand and Celite, and concentrated via rotary evaporator. Purification of the diester product was done by flash column chromatography on silica gel with a hexanes/acetone eluent. To a solution of the isolated diester in 14 mL of ethanol was added KOH (11.3 mmol; 3.7 mL of a 3 M solution), and the reaction was brought to reflux. After 2 h at reflux, TLC indicated no remaining starting material, and the reaction was allowed to cool to room temperature. The solvent was removed via rotary evaporator, and the product was diluted in water, acidified with 1 M HCl, extracted with 3  15 mL of diethyl ether, and dried with Na2SO4 overnight. The product was fully dried by connection to a high vacuum overnight. Solid was obtained in approximately 85% overall yield. 1H NMR (CD3OD) δ 8.10 (s, 1 H), 7.61 (s, 2 H), 3.95 (t, 2 H), 1.31 and 1.70 (m, 16 H), 0.98 (t, 3 H). EI-MS: m/z=322. mp=196-198 °C. IR (neat) significant peaks: 2954, 2921, 2851 (C-H), 1717, 1687 (CdO), 1593 (CdC) cm-1. 18156 DOI: 10.1021/la103494g

iodohexane or 1-iodopentane (Aldrich, 98%; 6 mmol; 0.79 mL) and K2CO3 (6 mmol; 0.83 g) in 7.5 mL of acetone at room temperature under argon was added a solution of dimethyl5-hydroxy-isophthalate (Aldrich, 98%; 5 mmol; 1.05 g) in 7.5 mL of acetone via cannula. The resulting white suspension was heated to ca. 50 °C. After stirring at 50 °C for 48 h, TLC indicated complete conversion of starting material. The reaction mixture was allowed to cool to room temperature, filtered through a plug of sand and Celite, and concentrated via rotary evaporator. Purification of the diester product was done by flash chromatography on silica gel with a hexanes/acetone eluent. The isolated diester was dissolved in 14 mL of ethanol, and to the resulting solution was added KOH (11.3 mmol; 3.7 mL of a 3 M solution). After refluxing the solution for 2 h, TLC indicated no remaining starting material and the reaction was allowed to cool to room temperature. The solvent was removed via rotary evaporator, and the product was diluted in water, acidified with 1 M HCl, extracted with 3  15 mL of diethyl ether, and dried with Na2SO4 overnight. The product was fully dried by connection to a high vacuum overnight. For 5-HIA: Solid was obtained in approximately 85% overall yield. 1H NMR (CD3OD) δ 8.23 (s, 1 H), 7.73 (s, 2 H), 4.07 (t, 2 H), 1.82, 1.52 and 1.70 (m, 8 H), 0.94 (t, 3 H). EI-MS: m/z= 266. mp = 199-201 °C. IR (neat) significant peaks: 2954, 2921, 2857 (C-H), 1707, 1687 (CdO), 1596 (CdC) cm-1. For 5-PIA: Solid obtained in approximately 83% overall yield. 1 H NMR (CD3OD) δ 8.10 (s, 1 H), 7.61 (s, 2 H), 3.95 (t, 2 H), 1.31 and 1.70 (m, 6 H), 0.98 (t, 3 H). EI-MS: m/z=252. mp=202-203 °C. IR (neat) significant peaks: 2932, 2871, 2958 (C-H), 1720, 1682 (CdO), 1594 (CdC) cm-1.

Results and Discussion Although hydrogen-bonding and van der Waals interactions are both weak intermolecular interactions, they are in different energetic ranges.2 Previous studies of the adsorption of alkanes and alkane thiols on gold surfaces have probed the energy value of van der Waals interactions between hydrocarbon chains, determining a value of approximately 6-7 kJ/mol per CH2 group.14,15 Based on previous research regarding carboxylic acids, the energy of a hydrogen-bond ranges from 15 to 25 kJ/mol.16,17 For the series of isophthalic acids examined, each molecule has nearly equal opportunities for hydrogen-bonding and varying possibilities for van der Waals interactions. 5-Octadecyloxyisophthalic Acid (5-OIA). As reported in a recent work,18 5-OIA self-assembles on HOPG and forms uniformly ordered lamellae (Figure 2). The average distance between repeating headgroups, indicated as d in Figure 2, was found to be 29 ( 2 A˚, which is comparable to the overall theoretical length of the molecule in all-trans configuration in the gas phase, 30 A˚ (as determined by the semiempirical method PM3). The headgroup to hydrocarbon chain angle, θ, was measured here to be 84 ( 5°, and the angle based on the model proposed for the adsorbed layer, θ0 , is seen to be 87°. This value is also comparable to the previously reported headgroup to chain angle of 88°.18 This assembly of uniformly ordered lamellae suggests that the dominant force in controlling self-assembly is interchain van der Waals interactions. The hydrocarbon chain groups interdigitate with each other, maximizing van der Waals interactions, as seen in (14) Lavrich, D. J.; Wetterer, S. M.; Bernasek, S. L.; Scoles, G. J. Phys. Chem. B 1998, 102, 3456. (15) Israelachvili, J. N. Intermolecular and Surface Forces; Academic Press Inc.: San Diego, CA, 1992. (16) Winkler, A.; Hess, P. J. Am. Chem. Soc. 1994, 116, 9233. (17) Pinto, S. S.; Diogo, H. P.; Guedes, R. C.; Cabral, B. J. C.; da Piedade, M. E. M.; Simoes, J. A. M. J. Phys. Chem. A 2005, 109, 9700. (18) Tao, F.; Bernasek, S. L. Surf. Sci. 2007, 601, 2284.

Langmuir 2010, 26(23), 18155–18161

Dickerson et al.

Article

Figure 2. (a) STM image of a monolayer of 5-octadecyloxyisophthalic acid dissolved in phenyloctane on HOPG (Vb =1.2 V, I=1.0 nA). Enlarged image is 5.513.2 nm2. (b) Proposed model for self-assembly of the 5-OIA system. Two unique opportunities for hydrogen bonding per molecule are indicated. The specific molecule is indicated as A, and the theoretical angle between headgroups is indicated as θ0 .

Figure 2. Based on the energy approximations for intermolecular forces, the total van der Waals interactions for each 5-OIA molecule would be ∼108-126 kJ/mol while the total hydrogenbonding contribution would be ∼30-50 kJ/mol. The hydrogenbonding estimate is based on an average of two unique hydrogenbonding opportunities per molecule as indicated in Figure 2. Therefore, the energetics suggest that the controlling factor for self-assembly here is van der Waals interactions. The 5-OIA molecules minimize their energy by aligning in uniform lamellae with interdigitated hydrocarbon tails, in order to maximize the van der Waals interactions. 5-Decyloxyisophthalic Acid (5-DIA). As seen in Figure 3a, 5-DIA also self-assembles into lamellar structures on HOPG. Two different conformations are seen in the images, indicating a deviation from the uniformly ordered lamellae witnessed in the case of the 5-OIA. The average distance between repeating headgroups, d, was measured for 5-DIA as well and was found to be 18 ( 2 A˚, which is comparable to the overall theoretical length of the all-trans configuration of the molecule in the gas phase, 20 A˚ (as determined by the semiempirical method PM3). The headgroup to chain angles, θ, were measured to be 76 ( 4° and 102 ( 6° in the two structures, and the angles based on the models shown in Figure 3b and c, θ0 , were determined to be 74° and 104°, respectively. Again, the structures involve an interdigitated packing of the hydrocarbon tails, maximizing van der Waals interactions. The presence of two different structures suggests closer competition between the intermolecular forces that control self-assembly. For 5-DIA, van der Waals forces are ∼60-70 kJ/mol and hydrogenbonding per molecule is the same as in 5-OIA, ∼ 30-50 kJ/mol. The van der Waals forces remain the most substantial energetically; however, the difference between the two types of interaction is significantly lower. Therefore, the chain-chain van der Waals interactions still control self-assembly, but hydrogen-bonding of the carboxylic headgroups makes up a larger portion of the total intermolecular forces. As seen in Figure 3a, the proposed models for this system consist of a row of 5-DIA molecules interacting with an adjacent row of one of the two faces of the molecule. This suggests that the energy barrier between the two different assemblies is low enough to allow for both conformers to be present. Conversely, previous work has shown a homogeneous enantiomorphous structure leads to the lowest total energy by maximizing van der Waals interactions.5 Langmuir 2010, 26(23), 18155–18161

5-Pentyloxyisophthalic Acid (5-PIA). Unlike 5-OIA and 5-DIA, 5-PIA forms self-assembled monolayers with hydrogenbonding as the main driving force. As seen in Figure 4a, the molecules organize on the surface into a cluster formation, with an average width of 27 ( 2 A˚. The proposed model (Figure 4b) consists of a hexamer formation with the carboxylic groups organized in a ring. The hydrocarbon tail groups may not necessarily be lying flat on the surface, but rather might project out of the plane of the sample. Due to the lower local density of states, these shorter tails are not visible in the STM images of 5-PIA. For 5-PIA, hydrogen-bonding becomes a more dominant force energetically. Based on the earlier energy estimates, van der Waals interactions for a five carbon hydrocarbon chain are worth ∼30-35 kJ/mol while the hydrogen-bonding is ∼23-38 kJ/ mol. The hydrogen-bonding estimate is based on nine interactions per each six molecule cluster, and thus an average of 1.5 hydrogen-bonds per molecule, as indicated in Figure 4b. This agrees well with the proposed structure due to the apparent optimization of hydrogen-bonding at the expense of van der Waals interactions. The cluster formation observed suggests that the length of the chain is not significant enough for stabilization from van der Waals interactions to affect the structure. Therefore, the preferred structure is the close packing of hexamers observed. In 5-OIA and 5-DIA, van der Waals interactions controlled the structure and the hydrogen-bonds provided secondary stabilization. In the shorter chain 5-PIA molecules, the hydrogen-bonds are the primary controlling force, thus resulting in a structure with more optimal hydrogen-bonding opportunities. Previous work has shown that hydrogen-bonding is the dominant force in controlling the assembly of some carboxylic acids.19-21 5-Hexyloxyisophthalic Acid (5-HIA). Similarly to 5-PIA, 5-HIA does not assemble into lamellar structures, as is seen in Figure 5a. Instead the molecules arrange in a two-dimensional structure dominated by hydrogen-bonding. In Figure 5b, the proposed model shows the alignment of the carboxylic acid headgroups and the hydrocarbon chain alignment perpendicular to the surface. Figure 5c illustrates the higher order structure formed by 5-HIA. This structure is similar to the observed (19) Cicoira, F.; Santato, C.; Rosei, F. Top. Curr. Chem. 2008, 285, 203–267. (20) Markus Lackinger, M.; Griessl, S.; Markert, T.; Jamitzky, F.; Heckl, W. M. J. Phys. Chem. B 2004, 108, 13652–13655. (21) Tao, F.; Bernasek, S. L. Langmuir 2007, 23, 3513.

DOI: 10.1021/la103494g

18157

Article

Dickerson et al.

Figure 3. (a) STM image of a monolayer of 5-decyloxyisophthalic acid dissolved in phenyloctane on HOPG. (Vb = 1.2 V, I = 1.0 nA). Image is 16.6  16 nm2. Colored bars indicate the two different headgroup to chain angles present. For 5-DIA, these two angles were experimentally found to be 102 ( 6° and 76 ( 4°. (b, c) Proposed models for the 5-DIA monolayer system with two unique opportunities for hydrogen bonding per molecule indicated for molecule A. For (b), θ0 = 104°; for (c), θ0 = 74°.

supramolecular self-assembly of trimesic acid,19 and also the carboxylic acid headgroup alignment is comparable to the zigzag pattern observed for the assembly of unsubstituted isophthalic acid.20 Energetically, for 5-HIA, hydrogen-bonding becomes a more dominant force. Based on the energy estimates, van der Waals interactions for a six carbon hydrocarbon chain are worth ∼3642 kJ/mol while the hydrogen-bonding is the same as in 5-OIA and 5-DIA, ∼30-50 kJ/mol. The hydrogen-bonding estimate is based on an average of two hydrogen-bonds per molecule, as indicated in Figure 5b. This agrees well with the proposed structure due to the apparent optimization of hydrogen-bonding at the expense of van der Waals interactions. As seen in Figure 5b, 18158 DOI: 10.1021/la103494g

the hydrocarbon chains appear to be interacting with adjacent chains. The slight contribution from van der Waals interactions between the chains could help stabilize this supramolecular structure, thereby lowering the total energy. Previous work revealed an odd-even effect in the ordering of carboxylic acids on HOPG leading to different chiral domains depending on the length of the hydrocarbon chain.22 However, the hydrogen-bonding dominated structures seen for both 5-PIA and 5-HIA suggest that the overall structure for these molecules is not affected by the parity of the hydrocarbon chain. The series of (22) Tao, F.; Goswami, J.; Bernasek, S. L. J. Phys. Chem. B 2006, 110, 4199–4206.

Langmuir 2010, 26(23), 18155–18161

Dickerson et al.

Article

Figure 4. (a) STM image of a monolayer of 5-pentyloxyisophthalic acid dissolved in phenyloctane on HOPG. (Vb = 1.1 V, I = 0.5 nA). Image is 17  29 nm2. The average width of a cluster, A, was measured to be 27 ( 2 A˚. (b) Proposed model for self-assembly of 5-PIA. The circles indicate opportunities for hydrogen-bonding in the six molecule cluster which leads to an average of 1.5 hydrogen-bonds per molecule. A simplified representation of this model is shown in the inset and overlaid on the STM image. The large center circle represents the ring of carboxylic groups, and the smaller circles represent the benzene rings with carboxylic acids. A 27 A˚ bar is drawn to scale for comparison purposes.

Figure 5. (a) STM image of a monolayer of 5-hexyloxyisophthalic acid dissolved in phenyloctane on HOPG. (Vb = 1.1 V, I = 0.4 nA). Image is 16  29 nm2. The average width between repeating units, A, was measured to be 19 ( 3 A˚, and the average size of the repeating unit, B, was measured to be 24 ( 3 A˚. (b) Proposed model for self-assembly of 5-HIA. Opportunities for hydrogen-bonding are indicated with an average of two hydrogen-bonds per molecule. A 19 A˚ bar is drawn to scale for comparison purposes. The inset illustrates the interaction of the hydrocarbon chains above the surface. This proposed model can arrange in a two-dimensional structure shown in (c), where the solid circles represent benzene rings and carboxylic acid dimers and the hollow circles represent chain-chain interactions. A simplified representation of this model is shown and overlaid on the STM image.

structures observed indicates that there is a transition between van der Waals controlled and hydrogen-bonding controlled assemblies between the 5-HIA and 5-DIA. Energetics Summary. The previously mentioned per molecule energy estimates clearly suggest van der Waals forces dominate the structure for 5-OIA and 5-DIA, while hydrogenbonding is the controlling force for 5-HIA and 5-PIA. Figure 6 shows STM images for each molecule with primitive unit cells Langmuir 2010, 26(23), 18155–18161

indicated. Corresponding Fourier transforms of each image confirm the proposed lamellar structures and headgroup to chain angles for 5-OIA and 5-DIA, with 5-DIA displaying streaking due to the different structures present on the surface. 5-HIA clearly shows a near uniform hexagonal lattice, while 5-PIA shows a distorted hexagonal lattice, correlating with the proposed models. Table 1 summarizes the energy estimates for each molecule based on interactions per molecule and per unit area. The observed DOI: 10.1021/la103494g

18159

Article

Dickerson et al.

Figure 6. STM images indicating the primitive unit cell and Fourier transform for (a) 5- OIA, (b) 5-DIA, (c) 5-HIA, and (d) 5-PIA. Lattice parameters a, b, and R are indicated for each. For (a), a = 30 A˚, b = 8 A˚, and R = 87°. For (b), a = 20 A˚, b = 8 A˚, and R = 102°. For (c), a = 27 A˚, b = 25 A˚, and R = 62°. For (d), a = 22 A˚, b = 27 A˚, and R = 58°.

structures all correlate well with the energy estimates, not only on the per molecule basis, but also when looking at the overall surface energy on a unit area basis. Comparing the contributions from the van der Waals interactions and hydrogen-bonding, it appears that these are closest in energy for the 5-DIA, suggesting 18160 DOI: 10.1021/la103494g

the transition between the lamellar and hexagonal structures is closer to 5-DIA than 5-HIA. The hydrogen-bonding dominated structures for 5-HIA and 5-PIA are further confirmed by noting the lower energy contributions from van der Waals interactions were they to form lamellar structures. Langmuir 2010, 26(23), 18155–18161

Dickerson et al.

Article

Table 1. Energy Contributions from Each Interaction per molecule

-2

per area (nm )

molecule vdWs (kJ/mol) H-B (kJ/mol) vdWs (kJ/mol) H-B (kJ/mol) 5-OIA 108-126 30-50 91-106 25-42 5-DIA 60-70 30-50 43-50 22-36 17-23 5-HIA 36-42 30-50 7-8a 61-101 5-PIA 30-35 23-38 6-7a a Estimates for van der Waals forces per area were based on a model unit cell where 5-HIA and 5-PIA were arranged in a lamellar formation as in longer chain molecules, for comparison purposes. All per area estimates are based on an averaging of each intermolecular interaction per area across multiple images.

Conclusions Self-assembly is controlled by the balance between intermolecular interactions. For the series of isophthalic acids studied, hydrogen-bonding and van der Waals interactions control the structure. Uniformly ordered lamellae observed for 5-OIA

Langmuir 2010, 26(23), 18155–18161

suggest that, for sufficiently long chain length, the van der Waals interactions control self-assembly. For the intermediate chain length, 5-DIA, ordered lamellae are still formed, but were not uniform, suggesting reduced chain-chain interactions and stronger competition between van der Waals and hydrogen bonding interactions. The formation of different ordered structures instead of ordered lamellae for 5-HIA and 5-PIA suggests hydrogenbonding controlled self-assembly. This behavior is consistent with energy estimates on a per molecular basis as well as per unit area. The structures and estimates suggest that the transition between the two different types of structures lies closer to the 5-DIA than the 5-HIA molecule. Acknowledgment. The authors would like to acknowledge Jared Allred for fruitful scientific discussions. This work was partially supported by the National Science Foundation, Division of Chemistry, CHE-0616457. P.N.D. acknowledges the support of NSF IGERT Award DGE-0903661.

DOI: 10.1021/la103494g

18161