Vertical Positioning of Internal Molecular Scaffolding within a Single

Yi Yang, Yunfeng Lu, Mengcheng Lu, Jinman Huang, Raid Haddad, George Xomeritakis, Nanguo Liu, Anthony P. Malanoski, Dietmar Sturmayr, Hongyou Fan, ...
0 downloads 0 Views 203KB Size
9550

J. Phys. Chem. B 1998, 102, 9550-9556

Vertical Positioning of Internal Molecular Scaffolding within a Single Molecular Layer Henning Menzel,† Mark D. Mowery, Mei Cai, and Christine E. Evans* Department of Chemistry, UniVersity of Michigan, Ann Arbor, Michigan 48109-1055 ReceiVed: July 13, 1998; In Final Form: September 9, 1998

Three-dimensional design within a two-dimensional molecular layer is demonstrated by varying the vertical positioning of internal molecular scaffolding. Monolayer structures are successfully fabricated using photopolymerization of adjacent diacetylene-containing molecules with controlled position of the resultant polymer backbone. Regardless of diacetylene position within the monolayer, polymerization is shown for all molecular architectures. The effective conjugation length of the resulting monolayer polymer, however, appears to be dependent on the molecular architecture. Although no significant expansion or contraction is expected upon polymerization, changes in hybridization must be accommodated within the monolayer structure. While the spatial constraints for polymerization are considerable, the longest conjugation lengths are exhibited by molecular architectures with some disordered component within the alkyl side chains. The excellent longrange order in these monolayers, as demonstrated by heterogeneous electron-transfer measurements, indicates little disordering within interstitial regions at grain boundaries. As a result, it is likely that strain created by the hybridization shift is accommodated by degrees of freedom within the side chains. When the architecture prevents such strain release, the conjugation length of the polymer backbone is compromised. However, this strain is not sufficient to disrupt the long-range monolayer order, and no correlation exists between polymer conjugation length and inhibition of electron transfer. Indeed, a remarkable degree of long-range order is observed for all molecular architectures, demonstrating the viability of fabricating robust monolayer structures with varying positions of internal molecular scaffolding.

Introduction Nanoscale design and fabrication of monolayer assemblies are increasingly important in wide-ranging applications including adhesion, optoelectronics, and sensors.1-3 The formation of molecular scaffolding by linking adjacent molecules within a single molecular layer provides a versatile means to fabricate such interfacial structures. Formed whenever internal functionalities within the monolayer associate laterally, molecular scaffolding creates an extended network at the monolayer level. As a result, a broad range of interaction mechanisms, including hydrogen bonding,4,5 π-stacking,6,7 dipole coupling,8 and covalent attachment,9-14 may be utilized in fabricating these singlelayer assemblies. Manipulating the monolayer molecular architecture in this fashion affords considerable control over the chemical and physical properties that are central to interfacial design. However, the choice of interconnection mechanisms must not compromise the structural integrity of the monolayer assembly. In addition, this interconnected assembly is likely to require surface attachment to stabilize the resulting monolayer structure. This combination of requirements are often mutually exclusive, hindering the advancement of this design approach. In this study, these difficulties are overcome by creating molecular scaffolding using photocoupling of adjacent diacetylene molecules within a spontaneously assembled monolayer. Using precursors of the form

[CH3(CH2)nCtCsCtC(CH2)mS-]2 the tail (n) and spacer (m) lengths may be systematically varied † Institute for Macromolecular Chemistry, University of Hannover, Am Kleinen Felde 30, 30167 Hannover, Germany.

to form molecular scaffolding at varying vertical positions within the monolayer structure. After spontaneous assembly onto a gold substrate accompanied by surface attachment via Au-S bonds, covalent bonding of adjacent molecules is accomplished by photopolymerization. As illustrated in Figure 1, solid-state studies indicate that the spatial constraints for polymerization are considerable, but no significant expansion or contraction of the resulting structure upon cross-linking is anticipated.15-17 Moreover, no mass transport is required for polymerization because light is used to induce polymerization and no byproducts are formed during the process. As a result, minimal disruption of the monolayer structure by the interconnection process is expected. Indeed, the feasibility of forming such polymerized monolayers has been recently demonstrated.10-14 These films have also been shown to be quite robust and capable of withstanding a wide range of chemical, temperature, and electrical conditions.13 In this study, the fabrication of architectures with different vertical positioning of molecular scaffolding is investigated. Both the tail (n ) 7, 11, 15) and spacer (m ) 4, 6, 9) lengths are varied to fabricate monolayer structures where the polymer backbone is selectively positioned within the monolayer assembly. An n,m-PDA designation is used to identify polymerized monolayer architectures, as for a methylene tail length of 15 and spacer length of 9 (15,9-PDA). Since previous studies have shown a strong dependence on surface topography,14 all monolayers are fabricated on high-temperature evaporated gold substrates. These substrates are expected to have atomically flat domains of >100 nm and not significantly inhibit the formation of long conjugation length polymers (ref 14 and references therein). In all measurements, direct comparison is made to octadecanethiolate monolayers formed under identical conditions.

10.1021/jp9830023 CCC: $15.00 © 1998 American Chemical Society Published on Web 10/30/1998

Vertical Positioning of Internal Molecular Scaffolding

Figure 1. Schematic diagram of spatial contraints for polymerization.

Experimental Methods Chemicals. Diheneicosa-10,12-diyn disulfide (7,9-DA), dipentacosa-10,12-diyn disulfide (11,9-DA), ditetracosa-5,7-diyn disulfide (15,4-DA), dihexacosa-7,9-diyn disulfide (15,6-DA), and dinonacosa-10,12-diyn disulfide (15,9-DA) were synthesized according to a general procedure which has been previously described.18 The starting material for the synthesis of 15,6DA is not commercially available and was synthesized according to Brown19 by isomerization of 2-octyn-1-ol using the potassium salt of 1,3-diaminopropane.20 The final products were purified using a silica gel column with hexanes/dichloromethane (4:1) as solvent. All resultant diacetylenes were a white, crystalline material that showed a single spot by TLC (hexanes/DCM 4:1) and no extraneous peaks by 1H NMR. In contrast with analogous thiols, these diacetylene-containing disulfides are quite robust and can be stored for extended periods (>12 months) without degradation or polymerization. Monolayer Fabrication. A custom-built UHV deposition system was utilized to prepare gold films. Mica (ASTM V-2; Asheville-Schoonmaker Mica Co.) was cleaved on both sides immediately before insertion into the chamber. Mica substrates were suspended 21 cm above the source using a tantalum mask and baked out at 380 °C for 12-24 h using two 500 W halogen lamps positioned approximately 8 cm below the substrate. Gold (99.99%) was vapor deposited from a K-Cell (Oxford Instruments) onto the heated substrate (250 °C) at a rate of 0.03 Å/s to a final thickness of ∼2000 Å (Leybold Inficon Inc.). During the deposition process, the pressure was always less than 1 × 10-7 Torr. Substrates were subsequently annealed for 3 h at the deposition temperature and then allowed to cool to room temperature (P ) 2 × 10-10 Torr). Finally, the chamber was backfilled with dry nitrogen and the substrates were removed. The resulting gold films were immediately immersed in a 1 mM chloroform solution (Aldrich, >99%) of the diacetylene disulfide or octadecanethiol and allowed to equilibrate at room temperature for 40-48 h. Strict light control was maintained during the preparation and storage of the diacetylene solutions and monolayer films. The substrates were then removed and rinsed extensively with chloroform (Aldrich, >99%) and deionized water (Model UV Plus Milli-Q, Millipore; >18 MΩ) and dried under nitrogen. The resulting diacetylene monolayers were polymerized under nitrogen with a low-intensity UV lamp (model UVG-11; Ultra-Violet Products Inc.; λ ) 250-260 nm) at a distance of 2 cm. It is important to note that no evidence is observed for photooxidation under these irradiation conditions.

J. Phys. Chem. B, Vol. 102, No. 47, 1998 9551 Contact Angle Measurements. Advancing contact angles were measured using the tilted plate method.21 A drop of the liquid was placed on the sample, which was tilted until just before the drop moves. The angle was measured using a custom-built instrument. The measurements with hexadecane were carried out under ambient conditions and those with water in a chamber to control the humidity. Raman Spectroscopy. Resonance Raman spectra were obtained at two excitation frequencies using two different systems. For 633 nm excitation measurements, the imaging system consists of a microscope objective (10×, 0.25 NA), a spectrograph (Holoscope f/1.8; VPT; Kaiser Optical System), and a charge-coupled device (CCD) detector (TK1024AB, Photometrics). The 633 nm line from a He-Ne laser (model 05-LHP-991, Melles Griot) was utilized for excitation. For 532 nm excitation spectra, the imaging system utilized a microscope objective (5×, 0.13 NA), a spectrograph (Holoscope f/1.8i; VPT; Kaiser Optical System), and a CCD camera (EEV1511, Princeton Instruments). A diode-pumped, frequency-doubled Nd: YVO4 operated at 532 nm was used for excitation. For both systems, the laser power incident on the samples was typically 10 mW, and CCD detectors were cooled with liquid nitrogen to -110 °C. Both systems were calibrated using emission lines of known wavelength from a neon lamp. Fourier Transform Infrared Spectroscopy. External reflection FTIR experiments were accomplished using a nitrogenpurged spectrometer with a liquid nitrogen cooled MCT detector (Nicolet 550 Magna IR). A grazing-incidence angle of 85° (Spectra-Tech Inc.) and p-polarized light were utilized for all measurements. Spectra shown are the average of 1024 scans and referenced against an unmodified gold film. Due to possible reference surface contamination with time, a fresh gold film was utilized for each series of measurements to minimize spectral bias. All spectra were recorded with a resolution of 2 cm-1. Electrochemistry. Cyclic voltammetry experiments utilized a standard three-electrode cell with a double junction Ag/AgCl reference electrode (saturated KCl internal solution) and a coiled platinum wire counter electrode. All measurements were conducted at room temperature (22 ( 1 °C) with a potentiostat (CV-27, Bioanalytical Systems) and an XY recorder (HewlettPackard). Two complete scans were acquired for each sample, with up to 10 scans showing no discernible effect. The area of the working electrode was 0.95 cm2, as defined by an inert elastomer O-ring, and all solutions were prepared immediately before use with ultrahigh-purity water. Results and Discussion The formation of molecular scaffolding within a single molecular layer provides a considerable degree of control over the chemical and physical properties of the interfacial region. In this study, such molecular scaffolding is formed by the covalent attachment of neighboring monomers within a surfaceattached monolayer assembly. The position of the polymer backbone within the monolayer is controlled by synthesis of precursors with varying alkyl chain length in the spacer and tail regions. The range of molecular architectures examined in this study are illustrated in Figure 2. The influence of the spacer length is examined, while maintaining the distance from the interface to the polymer backbone constant, in 15,4-PDA, 15,6PDA, and 15,9-PDA. In contrast, the distance from the interface to the polymer backbone is varied within the 15,9-PDA, 11, 9-PDA, and 7,9-PDA series, keeping the spacer length constant. These systematic variations in the vertical position of the

9552 J. Phys. Chem. B, Vol. 102, No. 47, 1998

Figure 2. Molecular architectures investigated and their nomenclature.

TABLE 1: Phase Characteristics in Langmuir-Blodgett Multilayer Polydiacetylene Films

polymer backbone afford a range of molecular scaffolding configurations. Polymerization. As illustrated in Figure 1, the spatial constraints for polymerization are considerable for these diacetylene assemblies.15-17 However, previous studies of spontaneously assembled thiol and disulfide analogues demonstrate the feasibility for polymerizing these surface-attached monolayer structures.10-14 Upon polymerization, the backbone is expected to be fully conjugated in the form illustrated in Figure 1. Previous studies of multilayer assemblies fabricated using the Langmuir-Blodgett (LB) technique have shown the existence of two predominant phases and the occasional occurrence of an additional phase.15,16,22-25 As shown in Table 1, previous thin-film and multilayer studies have identified these phases based on their color. The so-called blue phase shows an absorption transition centered at 640 nm, with the purple and red phase centered at 600 and 540 nm, respectively. These transitions result from the excitonic absorption of the polymer backbone, where the conjugation length of the red form is diminished relative to the blue form. It is important to emphasize that the effective conjugation length may or may not be the actual length of the polymer backbone. Any twisting or misalignment of the polymer backbone will result in a conjugation length less than the actual polymer length. The polymerization process and measurement of the polymerized form have been historically monitored using UV/vis absorption spectroscopy. We have undertaken absorption measurements of the polymerization process for these surfaceattached monolayers in both reflection and transmission geometries. In both cases, a broad band centered at 650 nm evolves upon irradiation, similar to that reported for Langmuir-Blodgett PDA films. However, in contrast with previous reports for surface-attached PDA on gold substrates,11,12 this transition is not ascribed to the formation of the PDA backbone. This disagreement is based on control experiments in our laboratory

Menzel et al. that indicate the same peak is observed when irradiating an alkanethiolate monolayer formed on a gold substrate. Moreover, upon extended exposure, this broad transition does not shift to shorter wavelengths as might be expected for PDA.15,22,23 Although the exact nature of this transition is not well understood, the frequency range corresponds with plasmon absorption transitions.26 As a result, the direct measurement of monolayer polymerization by UV/vis absorption is not feasible. Resonance Raman spectroscopy provides a good alternative approach due to the selective measurement of the conjugated polymer backbone region. The increased Raman cross section upon conjugation allows monitoring of the polymerization process without interference from low-probability, nonresonant transitions. Indeed, alkyl-based monolayers and nonpolymerized diacetylenes exhibit no signal under the measurement conditions utilized in this study. Based on previous absorption studies of LB monolayer and multilayer assemblies, Raman excitation at 633 nm is expected to be in resonance with the blue form of the polymer nearly exclusively.22-25 In contrast, multilayer studies22-25 indicate that spectral overlap does not allow independent measurement of the red polymer form, and excitation at 532 nm is expected to probe both red and blue forms. As a result, resonance Raman spectroscopy provides a means for evaluating polymer formation as well as the state of the polymer. As illustrated in Figure 3, resonance Raman spectra of the polymerized monolayer assemblies exhibit transitions assigned to stretching vibrations in the double (1450-1500 cm-1) and triple (2080-2100 cm-1) bond regions. These transitions are considerably shifted relative to the transitions expected for the isolated bonds (1620 and 2260 cm-1, respectively).23 This decrease in the transition energy is consistent with the increased conjugation expected upon polymerization.27 Transitions in the 700 and 1100-1400 cm-1 regions have not been conclusively assigned in the literature and are conjectured to arise from the in-plane stretching vibrations of the backbone and the rocking/ wagging modes of methylenes adjacent to the backbone.10,28,29 Relative intensities of the alkene and alkyne vibrational transitions at a given excitation wavelength are expected to correspond with the degree of polymerization into the specific polymer form. However, intensity comparison between spectra of differing excitation wavelength is not feasible due to differing experimental configurations. Nonetheless, multiple frequency resonance Raman affords a clear measurement of polymerization for these monolayer assemblies. As illustrated in Figure 3, polymerization is observed for all molecular architectures investigated here. Excitation at 633 nm indicates a significant concentration of the blue form of the polymer for 15,9-PDA, with lower conversion indicated for 11,9-PDA and 7,9-PDA. The transition frequency for the alkene stretch at 1459 cm-1 is in good agreement with multilayer studies of the blue phase (ν ) 1455 cm-1).22,24 Excitation at 532 nm, expected to probe both the blue and red forms, shows polymer for all molecular configurations. While the precursor analogues with a nine-carbon spacer indicate polymer formation, the presence of the blue form in these monolayers does not allow unambiguous determination of the red form. However, measurements with lower wavelength excitation show a shift in both alkene and alkyne transitions to higher frequencies, indicative of a decreased conjugation length consistent with the red form. None of these films reach the frequency expected for the alkene transition of the red form based on multilayer studies (ν ) 1514 cm-1).22,24 Indeed, as the tail length decreases (15,9-PDA f

Vertical Positioning of Internal Molecular Scaffolding

J. Phys. Chem. B, Vol. 102, No. 47, 1998 9553

Figure 3. Resonance Raman spectra for two excitation frequencies as a function of molecular architecture.

11,9-PDA f 7,9-PDA), the alkene transition decreases from 1503 to 1495 to 1484 cm-1, respectively. This frequency decrease is mirrored for the alkyne transition, moving from 2102 to 2098 to 2088 cm-1, respectively. These transitions correspond more directly with an intermediate conjugation length polymer form cited in some LB multilayer assemblies (ν ) 1486 cm-1).22,24 This so-called purple form may represent a transition state between the blue and red form or may be an entirely distinct phase. Temperature perturbation studies are presently underway to assess the nature of this intermediate state. Based on these observations, however, these three architectures exhibit resonance Raman spectra consistent with both the extended conjugation length, blue form, and a shorter conjugation length form. At present, it is not clear whether these two conjugation length polymers are present within a single domain or represent differing domain structures. In contrast, as the spacer chain is decreased (15,9-PDA f 15,6-PDA f 15,4-PDA), the blue form polymer decreases markedly until no signal is measured for the shortest spacer. As discussed above, the 15,9-PDA monolayer shows both the blue and red form of the polymer. Upon decreasing the spacer chain to six carbons, the conversion to the blue form is diminished approximately 50 times. Moreover, when excited at 532 nm, the 15,6-PDA exhibits an alkene transition at 1481 cm-1, consistent with the intermediate conjugation length purple form. Finally, the shortest spacer chain architecture (15,4-PDA) shows no blue form at 633 nm, and therefore, the signal at 532 nm can be attributed solely to the shorter conjugation length. Only a small signal is observed at the shorter conjugation length transition (ν ) 1483 cm-1), indicating that the 15,4-PDA monolayer has only a small degree of polymerization into the purple form. Together, these observations indicate that the position of the diacetylene within the monolayer assembly has a direct impact on the form of the polymer obtained. That is, the polymer phase is dependent on the molecular architecture. Addressing the origin of this phase formation is more complex. The formation of the red and purple phases results from a decreased conjugation length. This form may arise when the domain size limits the polymer length and therefore the conjugation length. Studies are presently underway to assess the domain behavior as a function of molecular architecture.

Alternatively, strain on the backbone created by the polymerization process may impact the effective conjugation length. Although no significant expansion or contraction of the layer is expected upon polymerization, a change in hybridization from sp to sp2 occurs above and below the polymer backbone (Figure 1). The strain created by this hybridization change is likely translated into the alkyl chain portion of the monolayer in the form of induced twisting and/or tilting.16,24 In contrast with LB films, the bond between the gold surface and the sulfur atom is much stronger, and as a result, a greater influence of the substrate on the polymer formation is expected. In light of this realization, the longer spacer chain may act primarily to accommodate the alignment of adjacent diacetylene moieties leading to more efficient polymerization. Concomitantly, the longer spacer chain offers increased degrees of freedom to reduce any lattice strain created by the hybridization shift. As the spacer chain length decreases, it is reasonable to conjecture that both the initial alignment affecting polymer length and the ability to reduce lattice strain may be significantly diminished. As a result, the shorter spacer may be expected to yield red or purple phase because the strain is transferred to the polymer backbone, decreasing the conjugation length. These observations are in general agreement with LB multilayer studies where the position of the polymer backbone was varied. When the distance between the headgroup and the backbone is small, studies show a decrease in the overall polymerization efficiency and corresponding decrease in the effective conjugation length.24 Studies are underway to assess the role of the actual polymer chain length by direct measurement as a function of molecular architecture. Although the polymer length and the decreased ability to release lattice strain may be the key factors, it is important not to overlook the possible influence of the odd/ even effect in the spacer region. That is, the short spacer chains in these studies are both even (m ) 4, 6) and the longer spacer is odd (m ) 9). The differing strain leading to the shorter conjugation length may arise from the odd/even character of the spacer chain. However, multilayer studies with even spacers indicate that a spacer of m ) 2 yields the red form whereas m ) 8 yields the blue form. Although these multilayer studies indicate no significant odd/even effect, degrees of freedom within the multilayer structure may act to minimize this impact.

9554 J. Phys. Chem. B, Vol. 102, No. 47, 1998

Menzel et al. TABLE 2: Band Assignment for the C-H Stretching Region of the Alkyl Monolayers30-34 peak position, cm-1 2851 2855 2878 2918 2928 2937 ∼2898 (br) 2956 2965

mode assignmenta νs(CH2) νs(CH2) νs(CH3, FR) νa(CH2) νa(CH2) νs(CH3, FR) νs(CH2, FR) νa(CH3, ip) νa(CH3, op)

order cryst liq cryst liq

a Mode assignment nomenclature: νa ) asymmetric, νs ) symmetric, FR ) Fermi resonance, ip ) in-plane, op ) out-of-plane.

Figure 4. Advancing contact angles for water and hexadecane (HD) for various molecular architectures. Error bars are shown for triplicate measurements on each film, and the literature value for C18 is excerpted from ref 30.

With direct surface attachment, such degrees of freedom are considerably limited in the monolayer investigated here. As a result, the impact of odd/even spacers on the polymer form cannot be eliminated and are the subject of further studies. Wettability. The overall quality of these polymer assemblies is assessed using the contact angle of a drop of solvent on the monolayer surface. Although a large lateral area is measured, interactions probe only the outermost methylene region of the monolayer polymer. Indeed, polymerization has been shown to not have a significant impact on the measured contact angle.12 As illustrated in Figure 4, advancing contact angles for water on the longest tail monolayers (15,9-PDA; 15,6-PDA; 15,4PDA) are nearly identical and in good agreement with monolayers formed from octadecanethiol (C18). In contrast with the nonwetting nature of water on alkyl-based interfaces, hexadecane offers a more rigorous test of the chain ordering in the outermost regions. The deviation between hexadecane contact angles on the PDA and C18 is notable, in this case, indicating the outermost chains are not as well ordered for the PDA monolayers. Again, however, the position of the polymer backbone does not appear to significantly affect the contact angle measured for the longest tail PDA monolayers. In contrast, both water and hexadecane contact angles exhibit a significant diminution as the length of the tail region is decreased. This result is likely due to an increased disorder in the short tail chain, exposing more methylene groups to the surface as has been observed for short chain alkanethiols.21 These contact angle values indicate an overall high film quality where the tail region influences outer chain ordering. It is important to note that no apparent correlation is seen between the phase behavior measured by resonance Raman and contact angle measurements. This interesting observation indicates that if strain induces a perturbation in the conjugation length of the backbone, it is not sufficient to impact the structure of the outermost chains to a significant extent. The lack of correlation with the phase behavior is also a preliminary indication that the lower conjugation length phase does not exist at grain boundaries large enough to affect the contact angle measurements.

Alkyl Chain Crystallinity. The crystallinity of the alkyl side chains has often been implicated in the formation of differing polymer phases. In multilayer and thin-film studies, increased chain disorder has been correlated with the decreased conjugation length leading to the red phase. This correlation may arise from a decrease in the polymer length caused by misalignment of neighboring diacetylenes with more disordered alkyl chains. Equally likely, disordered alkyl chains may place strain on the polymer backbone, decreasing the conjugation length. Structure within the alkyl chain regions in these monolayer polymers is assessed using grazing-angle infrared spectroscopy. The CH stretching region of the infrared spectrum provides a measure of alkyl chain crystallinity based on the vibrational transition frequency (Table 2). A highly crystalline, all-trans configuration is expected to exhibit a transition at 2918 cm-1 for the methylene asymmetric stretch and 2850 cm-1 for the methylene symmetric stretch.31-35 With increased gauche defect sites, the transition frequency for both the symmetric and asymmetric vibrations will be shifted to higher values. In general, monolayer structures formed from long chain thiols on gold exhibit transitions consistent with highly crystalline structures. With the possible exceptions of interstitial regions and the termination of the tail region, these monolayers are expected to be quite homogeneous. In contrast, polydiacetylene structures have two distinct alkyl chain regions within the monolayer structure. Infrared spectroscopic measurements will probe both the spacer and tail regions simultaneously, leading to a vibrational transition that is a convolution of the crystallinity in the spacer and tail regions. As illustrated in Figure 5, the crystallinity of alkyl chains in PDA varies considerably with molecular architecture. The longest tail structures of 15,9-PDA, 15,6-PDA, and 15,4-PDA show highly ordered transitions from asymmetric and symmetric stretching vibrations centered at 2919 and 2851 cm-1, respectively. As the spacer chain is increased from four to nine carbons, however, an increasing transition appears as a higher frequency shoulder on the νa transition. This transition is ascribed to an increasing disordered component within the alkyl chains. Although the methyl Fermi resonance band (νs ) 2937 cm-1) complicates this spectral region, the polymerized monolayer appears to retain a highly crystalline character with an increasing disordered fraction. This apparent nonhomogeneity with increasing spacer length may arise from domain structure changes creating disordered interstitial regions in the presence of highly ordered domains. Alternatively, variation of the spacer length may lead to distinctly different crystallinity in the spacer and tail regions. This observation may indicate that the spacer region is becoming less well ordered as the carbon number increases or in shifting from an even to an odd carbon number. Both the domain structure and spacer/tail ordering hypotheses are consistent with resonance Raman measurements showing a

Vertical Positioning of Internal Molecular Scaffolding

Figure 5. Fourier transform infrared spectra of the CH stretching region for various molecular architectures obtained at grazing incidence.

marked decrease in polymer conversion with decreasing spacer length. This disordered component becomes more predominant as the tail length is decreased from 15,9-PDA to 11,9-PDA and 7,9-PDA. A corresponding shift in the center frequency from 2919 to 2924 and 2926 cm-1 is observed with decreasing tail length. In this case, the spacer length is constant and only the tail length is varied. As a result, an increasing disorder with decreasing chain length within the tail region is implicated. This conjecture is consistent with wettability measurements which showed a decreased contact angle as the tail length decreased. In contrast with their shorter spacer analogues, the differing tail length architectures examined here all indicate some degree of polymerization into the blue phase. Indeed, these monolayers exhibit mixed phase behavior, showing both the presence of blue phase together with the shorter conjugation length red and purple phases. Infrared results, together with resonance Raman spectra, demonstrate that high crystallinity in the alkyl chain regions does not always yield long conjugation length polymer. A high degree of crystalline character is observed when the polymer form is predominantly blue (15,9-PDA) and when no blue polymer is observed (15,4-PDA). For the range of architectures examined here, some chain disorder is always present when the longer conjugation length blue form is observed. These results are consistent with the conjecture that strain created by hybridization changes may be transferred into the alkyl chains. That is, increased length or disorder within the spacer region may allow strain to be transferred to the spacer upon hybridization shift (15,9-PDA). Indeed, blue polymer is observed even in structures with a considerable degree of disorder within the alkyl regions (11,9-PDA and 7,9-PDA). However, when the strain cannot be transferred away from the backbone, shorter

J. Phys. Chem. B, Vol. 102, No. 47, 1998 9555 conjugation lengths and lower overall conversion efficiency results. Characterization of the change in properties upon polymerization for these architectures is consistent with this strain mechanism.36 Nonetheless, the formation of varying length polymers according to the molecular architecture must also be considered. It is feasible that the shorter conjugation lengths observed are due to a decrease in the actual length of the polymer backbone. Studies are presently underway to discern these two models by evaluating the interconversion between the red and blue phases, as well as directly measuring the polymer length. Nonetheless, either mechanism indicates that the molecular architecture may be used to control the polymer phase. Long-Range Order. Finally, the question of the long-range order of these monolayer polymers remains. Heterogeneous electron transfer is used to assess spatial defects and overall film permeability as a function of molecular architecture. If a significant number of defect sites permits electron transfer through these monolayer films, the hypothesis regarding the presence of highly ordered domains within a less ordered structure is likely. Alternatively, if the films have good blocking characteristics, less ordered interstitial regions between domains may not be an accurate picture, and disordered alkyl chains within the tail or spacer region are implicated. Inhibition of electron transfer is evaluated by the shape and magnitude of cyclic voltammetric current response. As illustrated in Figure 6, an unmodified gold substrate shows the current response corresponding with unimpeded mass transfer of ferricyanide ions to the gold surface. Consistent with previous measurements,31 the octadecanethiolate monolayer exhibits excellent blocking characteristics. Significantly inhibited electron transfer is observed for the short spacer chain monolayers (15,4-PDA and 15,6-PDA). However, blocking is not quite complete, with the plateau in current response indicating possible radial diffusion to small, independent defect sites.37,38 For these shorter spacer chains, the domain structure may create small spatial defects without significantly affecting the alkyl chain ordering. In contrast, the longer spacer chain shows outstanding inhibition for all tail length architectures (15,9-PDA; 11,9-PDA; 7,9-PDA). This result is somewhat surprising given the significant degree of alkyl chain disorder in the shorter tail monolayers. Based on these results, no significant defect sites are observed within the monolayer structure, and it is unlikely that significant chain disorder occurs at interstitial grain boundaries. Moreover, these combined observations are inconsistent with the hypothesis attributing nonhomogeneity in alkyl chain order to disordered interstitial regions in the presence of highly ordered domains. The actual role of the polymer backbone on electron-transfer inhibition has yet to be clarified, and studies are underway to assess the impact of both domain structure and polymer character. Nonetheless, these results demonstrate that monolayer structures with longrange order can be fabricated in a range of molecular architectures with differing polymer forms. Conclusions A robust means for manipulating molecular scaffolding within single molecular layers is successfully demonstrated using surface-attached monolayers with internally linked structures. Systematic variation in polymer backbone position yielded highquality monolayer structures with a considerable degree of longrange order. Based on these observations, the formation of

9556 J. Phys. Chem. B, Vol. 102, No. 47, 1998

Menzel et al. Acknowledgment. The authors acknowledge financial support from the National Institutes of General Medical Sciences, National Institutes of Health (GM52555-01). H.M. thanks the Fulbright Commission for a travel grant. The assistance of Jerilyn Timlin in the research group of Prof. Michael Morris in obtaining the resonance Raman measurements at 532 nm is gratefully acknowledged. References and Notes

Figure 6. Cyclic voltammograms as a function of molecular architecture. A solution of 1.0 mM in Fe(CN)63- and 1.0 mM KCl was used at a sweep rate of 100 mV/s.

extended conjugation length polymer chains appears to be one of compromise. Adjacent diacetylenes must be well aligned to form extended polymer lengths, but sufficient flexibility near the backbone is needed to accommodate the shift in hybridization upon polymerization. When the alkyl chains are tightly constrained, both the conjugation length and the conversion efficiency exhibit a marked decrease. However, as the fraction of the disorder state within the side chains increases, a range of conjugation lengths are created. Studies are presently underway to discern the role of domain size on the resultant conjugation length obtained for different architectures. Nonetheless, the successful fabrication of a wide range of molecular architectures has been demonstrated, offering a significant degree of control over the interfacial properties without significant variation in chemical composition. Not only interesting from a fundamental perspective, such monolayer assemblies are central to the design of robust interfaces with well-characterized viscoelastic and optical properties.

(1) Lio, A.; Charych, D. H.; Salmeron, M. J. Phys. Chem. B 1997, 101, 3800. (2) Mandler, D.; Turyan, I. Electroanalysis 1996, 8, 207. (3) Dermody, D. L.; Crooks, R. M.; Kim, T. J. Am. Chem. Soc. 1996, 118, 11912. (4) Clegg, R. S.; Hutchinson, J. E. Langmuir 1996, 12, 5239. (5) Sabapathy, R. C.; Bhattacharyya, S.; Leavy, M. C.; Cleland, W. E.; Hussey, C. L. Langmuir 1998, 14, 124. (6) Dhirani, A.; Zehner, R. W.; Hsung, R. P.; Guyot-Sionnest, P.; Sita, L. R. J. Am. Chem. Soc. 1996, 118, 3319. (7) Sachs, S. B.; Dudek, S. P.; Hsung, R. P.; Sita, L. R.; Small, J. F.; Newton, M. D.; Feldberg, S. W.; Chidsey, C. E. D. J. Am. Chem. Soc. 1997, 119, 10563. (8) Evans, S. D.; Goppert-Berarducci, K. E.; Urankar, E.; Gerenser, L. J.; Ulman, A. Langmuir 1991, 7, 2700. (9) Peanasky, J. S.; McCarley, R. L. Langmuir 1998, 14, 113. (10) Batchelder, D. N.; Evans, S. D.; Freeman, T. L.; Ha¨ussling, L.; Ringsdorf, H.; Wolf, H. J. Am. Chem. Soc. 1994, 116, 1050. (11) Kim, T.; Crooks, R. M.; Tsen, M.; Sun, L. J. Am. Chem. Soc. 1995, 117, 3963. (12) Kim, T.; Ye, Q.; Sun, L.; Chan, K. C.; Crooks, R. M. Langmuir 1996, 12, 6065. (13) (a) Kim, T.; Chan, K. C.; Crooks, R. M. J. Am. Chem. Soc. 1997, 119, 189. (b) Mowery, M. D.; Menzel, H.; Cai, M.; Evans, C. E., unpublished results. (14) (a) Mowery, M. D.; Evans, C. E. J. Phys. Chem. B 1997, 101, 8513. (b) Mowery, M. D.; Menzel, H.; Cai, M.; Evans, C. E. Langmuir 1998, 14, 5594. (15) Lando, J. B. In Polydiacetylenes; Bloor, D., Chance, R., Eds.; Nijhoff: Dordrecht, The Netherlands, 1985. (16) Schott, M. W.; Wegner, G. In Nonlinear Optical Properties of Organic Molecules and Crystals; Chemla, J., Ed.; Academic Press: Orlando, 1987. (17) Cao, G.; Mallouk, T. E. J. Solid State Chem. 1991, 94, 59. (18) Mowery, M.; Evans, C. E. Tetrahedron Lett. 1997, 38, 11. (19) Brown, C. A.; Yamashita, A. J. Chem. Soc., Chem. Commun. 1976, 959. (20) Kimmel, T.; Becker, D. J. Org. Chem. 1984, 49, 2494. (21) (a) Ulman, A. An Introduction to Ultrathin Organic Films; Academic Press: New York, 1991 and references therein. (b) Berg, J. C. Wettability, Vol. 49 Surfactant Science Series; Marcel Dekker: New York, 1993. (22) Tieke, B.; Bloor, D. Makromol. Chem. 1979, 180, 2275. (23) Lieser, G.; Tieke, B.; Wegner, G. Thin Solid Films 1980, 68, 77. (24) Tieke, B.; Lieser, G. J. Colloid Interface Sci. 1982, 88, 471. (25) Mino, N.; Tamura, H.; Ogawa, K. Langmuir 1992, 8, 594. (26) Barker, A. S. Phys. ReV. B 1973, 8, 5418. (27) Lin-Vien, D.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. In The Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic Press: San Diego, CA, 1991. (28) Bower, D. I.; Maddams, W. F. The Vibrational Spectroscopy of Polymers; Cambridge University Press: Cambridge, 1989. (29) Angkaew, S.; Wang, H.-Y.; Lando, J. B. Chem. Mater. 1994, 6, 1444. (30) Laibinis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y.-T.; Parikh, A. N.; Nuzzo, R. G. J. Am. Chem. Soc. 1991, 113, 7152. (31) Hill, I. R.; Levin, I. W. J. Chem. Phys. 1979, 70, 842. (32) Snyder, R. G.; Scherrer, J. R. J. Chem. Phys. 1979, 71, 3221. (33) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109, 3559. (34) Bain, C. D.; Troughton, E. B.; Tao, Y.-T.; Evall, J.; Whitesides, G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321. (35) Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990, 112, 558. (36) Cai, M.; Mowery, M. D.; Menzel, H.; Evans, C. E. Langmuir, submitted. (37) Amatore, C.; Save´ant, J.-M.; Tessier, D. J. Electroanal. Chem. 1983, 147, 39. (38) Chailapakul, O.; Crooks, R. M. Langmuir 1993, 9, 884.