1258
Langmuir 2007, 23, 1258-1263
Monitoring the Formation of Self-Assembled Monolayers of Alkanedithiols Using a Micromechanical Cantilever Sensor Swapnil Kohale,†,‡ Sara M. Molina,‡,§ Brandon L. Weeks,† Rajesh Khare,† and Louisa J. Hope-Weeks*,§ Department of Chemistry and Biochemistry, Texas Tech UniVersity, Lubbock, Texas 79409, and Department of Chemical Engineering, Texas Tech UniVersity, Lubbock, Texas 79409 ReceiVed August 17, 2006 Using a micromechanical cantilever device, the surface stress induced during the growth of alkanedithiol (HS(CH2)nSH) monolayers on gold in solution is continuously monitored and reported. Adsorption of alkanedithiols of varying chain lengths is observed and compared to each other, as well as to the adsorption of hydroxyalkanethiols (HS(CH2)nOH) and alkanethiols (HS(CH2)nCH3). The results have revealed a significant change in surface stress on the basis of the chain length of the alkanedithiol. The long-chain (n > 10) alkanedithiol adsorption imposes a tensile stress on the gold-coated surface of the cantilever rather than the compressive stress exhibited by both alkanethiols and short-chain dithiols. Our results suggest a phenomenon in which the two thiols of the alkanedithiol adsorb onto the gold surface forming a loop inducing a tensile stress on the cantilever for long chain lengths. This study shows that micromechanical cantilever sensors can be very valuable tools in the exploration and characterization of selfassembled monolayers.
Introduction Micromechanical cantilever sensors are highly flexible and sensitive devices capable of detecting differences in deflections at the pico- to nanogram range with a response time on the order of milliseconds.1 Differences in forces as small as a piconewton and displacements as small as an angstrom are detectable.2 The cantilever behaves as a transducer translating a physical quantity into a measurable electric signal.3 When the reactive surface of the cantilever is exposed to an analyte, the cantilever mechanically responds by bending due to a change in surface stress. The deflection of the cantilever produced by the adsorption of an analyte is detected and measured with the use of the optical lever technique. This technique uses a laser diode beam focused on the apex of the cantilever and is reflected onto a split photodiode (position-sensitive detector). When the cantilever bends, the reflected light moves on the photo detector and its movement is proportional to the cantilever deflection.1 The self-assembly of alkanethiols is easily detected with the use of a gold-coated microcantilever due to the affinity of the sulfur atom to the gold surface.4 The adsorption of the alkanethiol molecules on the surface, if not compensated by an equal stress on the opposite side, will result in a deflection of the cantilever. The deflection imposed is the result of either an increase (compressive stress) or decrease (tensile stress) in the surface area. In the case of a compressive stress, the cantilever bends downward, away from the self-assembled monolayer (SAM) as opposed to an upward bend (toward the SAM) for tensile stress.4 SAMs of alkanethiols as well as alkanedisulfides have proven * To whom correspondence should be addressed. E-mail:
[email protected]. † Department of Chemical Engineering. ‡ These authors contributed equally to this work. § Department of Chemistry and Biochemistry. (1) Berger, R.; Gerber, Ch.; Lang, H. P.; Gimzewski, J. K. Microelectron. Eng. 1997, 35, 373. (2) Thundat, T.; P. I. Oden, P. I.; Warmack, R. J. Microscale Thermophys. Eng. 1997, 1, 185. (3) Raiteri, R.; Grattarola, M.; Berger, R. Sens. Actuators, B 2001, 79, 115. (4) Berger, R.; Delamarche, E.; Lang, H. P.; Gerber, Ch.; Gimzewski, J. K.; Meyer, E.; Guntherodt, H. J. Science 1997, 276, 2021.
to self-organize into well-ordered, densely packed films on gold.5 They represent a model molecular system for controlling surface properties that can be adjusted by changing the chemical nature of their terminal groups.6 SAMs provide a unique link between the science of organic surfaces and technologies that seek to exploit their adaptable character. The ability to tailor the monolayer offers a simple approach to the enhancement of various applications ranging from biological sensing7-11 to micro/ nanofabrication.12,13 The properties of alkanethiol monolayers have been studied extensively through both experimental and computational means.14-20 The bond interaction between the sulfur and gold surface is strong and estimated to be approximately 40-45 kcal/ mol.21 The alkane chains adopt a predominantly all-trans conformation and tilt uniformly at approximately 30-40° from (5) Ulman, A. Chem. ReV. 1996, 96, 1533. (6) Witt, D.; Klajn, R.; Barski, P.; Grzybowski, B. A. Curr. Org. Chem. 2004, 8, 1763. (7) Wu, G.; Ji, H.; Hansen, K.; Thundat, T.; Datar, R.; Cote, R.; Hagan, M. F.; Chakraborty, A. K.; Majumdar, A. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 1560. (8) Hansen, K. M.; Ji, H.; Wu, G.; Datar, R.; Cote, R.; Majumdar, A.; Thundat, T. Anal. Chem. 2001, 73, 1567. (9) Fritz, J.; Baller, M. K.; Lang, H. P.; Rothuizen, H.; Vettiger, P.; Meyer, E.; Guntherodt, H. J.; Gerber, Ch.; Gimzewski, J. K. Science 2000, 288, 316. (10) Gfeller, K. Y.; Nugaeva, N.; Hegner, M. Appl. EnViron. Microbiol. 2005, 71, 2626. (11) Nelson, K. E.; Gamble, L.; Jung, L. S.; Boeckl, M. S.; Naeemi, E.; Golledge, S. L.; Sasaki, T.; Castner, D. G.; Campbell, C. T.; Stayton, P. S. Langmuir 2001, 17, 2807. (12) Joseph, S. T. S.; Ipe, B. I.; Pramod, P.; Thomas, K. G. J. Phys. Chem. B 2006, 110, 150. (13) Deng, W.; Yang, L.; Fujita, D.; Nejoh, H.; Bai, C. Appl. Phys. A 2000, 71, 639. (14) Edinger, K.; Golzhauser, A.; Demota, K.; Woll, Ch.; Grunze, M. Langmuir 1993, 9, 4. (15) Wilbu, J. L.; Biebuyck, H. A.; MacDonald, J. C.; Whitesides, G. M. Langmuir 1995, 11, 825. (16) Schnelder, T. W.; Buttry, D. A. J. Am. Chem. Soc. 1993, 115, 12391. (17) Witt, D.; Klajn, R.; Barski, P.; Grzybowski, B. A. Curr. Org. Chem. 2004, 8, 1763. (18) Hautman, J.; Klein, M. L. J. Chem. Phys. 1989, 91, 4994. (19) Jiang, S. Mol. Phys. 2002, 100, 226. (20) Alexiadis, O.; Daoulas, K.; Mavrantzas, V. G. To be submitted for publication. (21) Schonenberger, C.; Jorritsmal, J.; Sondaq-Huethorst, J. A. M.; Fokkink, L. G. J. J. Phys. Chem. 1995, 99, 3259.
10.1021/la062441n CCC: $37.00 © 2007 American Chemical Society Published on Web 11/30/2006
Self-Assembled Monolayers of Alkanedithiols
Langmuir, Vol. 23, No. 3, 2007 1259
Figure 1. Adsorption of hydroxythiols and a long-chain alkanethiol on a gold-coated cantilever illustrating typical alkanethiol surface stress kinetics (compressive stress).
the surface normal. The organization and self-assembly of these monolayers on a gold surface are consistent, not only for alkanethiols but for functionalized alkanethiols as well. Variation of the chain-terminating functional groups can have a dramatic effect on the structure of the SAM created.22 There are two distinct phases involved in the formation of an alkanethiol monolayer in solution. The first phase is the adsorption of the S-headgroups on the Au surface (which takes only a few minutes), and the second phase occurs gradually over time (several hours) as the hydrocarbon tails orient themselves in an all-trans ordered fashion.5 The interest in self-assembled monolayers of alkanedithiols (HS(CH2)nSH) comes predominately from applications where the monolayers are used as bridging molecules, or linkers, creating molecular junctions between metal particles such as nanorods and metal substrates, forming multilayered structures.12,23,24 Studies of alkanedithiol monolayer formation are not as common and frequent as studies of alkanethiols or disulfides on gold.25 The literature that is available is scattered and oftentimes conflicting.25-30 Several studies on the adsorption of dithiols report that only one thiol bonds to the gold surface, similar to the orientation of alkanethiols on gold. Cyclic voltammeric studies and impedance studies, in the case of 1,8-octanedithiol and 1,6-hexanedithiol, revealed that these SAMs adsorb on the Au(111) surface with only one sulfur atom on the gold.26 It was also reported in the same study that the adsorption rate of 1,8-octanedithiol was higher than that of its alkanethiol homologue because it has the choice of bonding with either of the two terminal-SH groups onto the gold substrate, which increases the probability of bond formation.26 X-ray photoelectron spectroscopy (XPS) studies of 1,8octanedithiol describe the dithiol adsorption to be in the headdown, tail-up orientation rather than a “loop” on the surface, where both sulfurs are bound to the gold substrate.25,27 Ellipsometry measurements of the 1,4-, 1,6-, 1,8-, and 1,10-dithiols (22) Schreiber, F. Prog. Surf. Sci. 2000, 65, 151-256. (23) Deng, W.; Yang, L.; Fujita, D.; Nejoh, H.; Bai, C. Appl. Phys. A 2000, 71, 639. (24) Li, X.; He, J.; Hihath, J.; Xu, B.; Lindsay, S. M.; Tao, N. J. Am. Chem. Soc. 2006, 128, 2135. (25) Vance, A. L.; Willey, T. M.; Nelson, A. J.; van Buuren, T.; Bostedt, C.; Terminello, L. J.; Fox, G. A. Langmuir 2002, 18, 8123. (26) Sur, U. K.; Subramanian, R.; Lakshiminarayanan, V. J. Colloid Interface Sci. 2003, 266, 175. (27) Riely, H.; Kendalss, G. K.; Zemicael, F. W.; Smith, T. L.; Yang, S. Langmuir 1998, 14, 5147. (28) Engelkes, V. B.; Beebe, J. M.; Frisbie, C. D. J. Am. Chem. Soc. 2004, 126, 14287. (29) Nishitani, R.; Tateishi, Y.; Arakawa, H.; Kasuya, A.; Sumiyama, K. Acta Phys. Pol., A 2003, 104 (3-4), 269. (30) Niklewski, A.; Azzam, W.; Strunskus, T.; Fischer, R. A.; Woll, Ch. Langmuir 2004, 20, 8620.
Figure 2. Adsorption of alkanedithiols: (a) 2.0 mM 1,6-hexanedithiol, SH(CH2)6SH, injected at approximately 420 s; (b) 2.0 mM 1,8-octanedithiol, SH(CH2)8SH, injected at approximately 900 s; (c) 2.0 mM 1,10-decanedithiol, SH(CH2)10SH, injected at approximately 900 s.
reported their adsorption behavior to be comparable to that of their analogous alkanethiol monolayers thus indicating that these molecules are not “hairpinned” to the surface but are standing upright.28 However, scanning tunneling microscopy (STM)-induced luminescence studies showed a spectral change of Au(111) due to the adsorption of octanedithiol molecules but not for that of decanethiol molecules. The spectral change is reportedly due to the different adsorption geometry between the two, the standing molecular chain for the decanethiol and the molecular chain flat on the surface for octanedithiol. The change of electronic states in octanedithiol is due to two S-Au bonds instead of one S-Au bond in the case of decanethiol molecules and also the interaction of the alkyl chain with the gold surface.29 Another study reported that to fabricate a SAM with a free thiol on the surface, a protecting group had to be used to prevent both thiols from adsorbing on the surface.30 A study by Carot et al. showed that dithiol monolayers have a higher stability than alkanethiol monolayers due to the formation of disulfide bonds between the terminal surface atoms, with the alkyl chains arranged in an antiparallel fashion.31 There have also been reports on the formation of alkanethiol monolayers forming disulfides between headgroups. This phenomenon exists through a gauche bond occurring when one sulfur atom sits in the 3-fold hollow site and the other sits near the bridge site.32,33 It is thus apparent that the properties of these monolayers are very versatile and the nature of the exact structure of these monolayers on the gold surface is still unclear, especially for the case of the dithiol monolayers. Experiments that use microcan(31) Carot, M. L.; Esplandiu, M. J.; Comelto, F. P.; Patrito, E. M.; Macagno, V. A. J. Electroanal. Chem. 2005, 579, 13. (32) Kluth, G. J.; Carraro, C.; Maboudian, R. Phys. ReV. B 1999, 59, R10 449. (33) Fenter, P.; Eberhardt, A.; Eisenberger, P. Science 1994, 266, 1216.
1260 Langmuir, Vol. 23, No. 3, 2007
Kohale et al.
Figure 3. (a) Schematic of an ordered monolayer phase. (b) Schematic of a disordered monolayer phase resulting from the presence of chain loops on the Au surface of the cantilever. (c) Internal coordinates of the chain: bond length (b), bond angle (θ), and torsional angle (φ). (d) Eulerian angles associated with the chain. Axes x and y are in the plane of the surface. ψ1 is the angle formed with the x axis by the projection of the S-C bond on the xy plane. ψ2 is the angle between S-C bond and the z axis. ψ3 is the dihedral angle formed by the z axis, the S-C bond, and the C-C bond (second bond of the chain).
tilevers to study the self-assembly of alkanethiols on gold have shown that monolayer formation induces a compressive surface stress on the microcantilever.34-37 This information is consistent for the adsorption of long-chain alkanethiols and hydroxyterminated alkanethiols (HS(CH2)nOH). This paper will focus on the self-assembly of alkanedithiols (HS(CH2)nSH) on a goldcoated microcantilever. The response of the cantilever to surface stress changes are monitored in real time as the formation of the monolayers occurs. These observations are used to elucidate the effects of both the presence of two thiol groups and the chain length of the hydrocarbon part of the dithiol on the structure of the adsorbed monolayer. Furthermore, the structural features of the dithiol monolayers suggested by the experimental data are validated by carrying out Monte Carlo simulations of the conformations of dithiol molecules that are adsorbed on a planar surface. Experimental Section Microcantilever Sensor Experiments. The surface stress induced by alkanethiol monolayer formation was monitored with a silicon nitride, gold-coated cantilever approximately 115 µm in length, 25 µm in width, and 0.6 µm thick (Veeco MLCT-AUHW-A). The cantilever is mounted in a Digital Instruments MultiMode head equipped with a diode laser and a linear split photodiode detector. The signal produced by the cantilever’s response to the adsorption of the alkanethiols was amplified and processed through the interface of a DAQ (data acquisition) pad to interpret the signal and a computer with Labview to record the voltage (measuring the degree of deflection relevant to its original position on the photodiode) vs time in seconds. All experiments were performed in a continuous flow environment controlled by a flow-regulating syringe pump. A 6-port valve injector is also incorporated into the system to provide a smooth transition between solvent and alkanethiol injections with a sample loop of ∼1 (34) Li, J.; Liang, K. S.; Scoles, G.; Ulman, A. Langmuir 1995, 11, 4418. (35) Berger, R.; Delamarche, E.; Lang, H. P.; Gerber, Ch.; Gimzewski, J. K.; Meyer, E.; Guntherodt, H. J. Appl. Phys. A 1998, 66, S55. (36) Godin, M.; Williams, P. J.; Tabard-Cossa, V.; Laroche, O.; Beaulieu, L. Y.; Lennox, R. B.; Grutter, P. Langmuir 2004, 20, 7090. (37) Itakura, A. N.; Berger, R.; Narushima, T.; Kitajima, M. Appl. Phys. Lett. 2002, 80, 3712.
mL. Prior to injection and after injection, only the solvent was flowed through the cell. The cantilever is clamped at one end to the main body of the holding plate, and the other end is suspended in the fluid cell for accurate measurements of deflection. Prior to exposure of the cantilever to samples of alkanethiol and dithiol, several tests were conducted to test the signal for accuracy and to select optimum solvent and flow rate parameters for the system. Control experiments were also performed to monitor the cantilever’s response to air and solvent (ethanol) at room temperature, confirming thermal drift to be negligible relative to the signal obtained from thiol adsorption. Materials. Gold-coated silicon nitride cantilevers were supplied by Veeco. The cantilever is coated on one side with approximately 60 nm of gold, promoting the adsorption of the alkanethiols on one side of the cantilever and thereby promoting a bend in the cantilever which creates a detectable signal. All solutions were prepared with pure HPLC grade ethanol. The alkanedithiols and alkanethiols, namely, 1,6-hexanedithiol (HS(CH2)6SH), 1,8-octanedithiol (HS(CH2)8SH), 1,10-decanedithiol (HS(CH2)10SH), 11-mercapto-1undecanol (HS(CH2)11OH), and 11-undecanethiol (HS(CH2)10CH3), were purchased from VWR, and 8-hydroxy-1-octanethiol (SH(CH2)8OH) was purchased from Dojindo Laboratories. All dithiols were supplied under an argon environment to minimize disulfide formation. All chemicals were used without further purification. Analysis and Data Acquisition. All experiments were performed at room temperature at a constant flow rate of 0.10 mL/min. The concentrations of alkanethiol in solution were held constant at 2 mM in ethanol. A data point relative to the deflection of the cantilever is collected every 1 s over the entire course of each experiment. The system is purged with ethanol for approximately 10 min to establish a baseline prior to injection of the thiol-containing analyte. Data shown are the entire run, including the first solvent purge defined at t ) 0. The injection time for each run is included in the data. Initial studies were performed on different cantilevers giving similar results to the data presented here. To minimize noise due to changes in spring constant and gold thickness, the data presented are for the same cantilever. The cantilever is cleaned between alkanethiol adsorption experiments with a 1:1:5 cleaning solution of hydrogen peroxide, ammonia, and water, which removes organic materials and does not etch gold like piranha. The cantilever is soaked in the cleaning solution for several hours and is then rinsed with ethanol. This cleaning procedure allows us to use the same cantilever for
Self-Assembled Monolayers of Alkanedithiols
Langmuir, Vol. 23, No. 3, 2007 1261
Table 1. Percentage of Chains Looping Back to the Planar Surface for Various Chain lengths and over the Range of ψ3a chain length
a
ψ3
5
6
7
8
9
10
11
12
13
14
15
-180 -170 -120 -50 0 50 120 170 180
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0.2 0.2 0.1 0 0 0 0.1 0.2 0.2
0.2 0.2 0.5 0.1 0.1 0.1 0.5 0.2 0.2
3.5 3.5 2.3 0.7 0.3 0.7 2.3 3.5 3.5
5.3 4.4 4 1.3 1 1.3 4 4.5 5.3
6.4 6.3 5.5 3 2 3 5.5 6.3 6.3
8.8 8.3 7.1 4 3 4 7.1 8.4 8.8
10.7 10.4 9.3 5.6 4.6 5.6 9.2 10.3 10.6
The results are shown for ψ1 ) 33° and ψ2 ) 15° and by considering the first torsional angle to be always in the trans state.
each experiment. Each experiment is repeated at least three times, and the data set for each molecule being studied is reproduced. The data were collected in random order with dithiols and monothiols interchanged on each run. Results are shown as graphs of cantilever deflection as a function of time for alkanethiol and alkanedithiol exposure. Two different hydroxythiols (HS(CH2)nOH) and one long-chain alkanethiol (HS(CH2)nCH3), comparable in chain length to the alkanedithiols, were also analyzed to investigate the phenomenon of long-chain alkanedithiol adsorption further. The hydroxy terminal end of the alkanethiol molecules is chemically similar to that of a thiol, except for the ability of a thiol to bind to gold. Therefore, these studies provide a baseline on the behavior of long-chain monothiols with similar functionality and chain length. By obserervation of the hydroxyalkanethiols, the importance of the dithiol’s contribution to the response of the cantilever is apparent. The two hydroxy alkanethiols studied were 8-hydroxy-1-octanethiol (HS(CH2)8OH) and 11-mercapto-1-undecanol (HS(CH2)11OH); see Figure 1. Also included in this figure is the adsorption of a long-chain alkanethiol, 11-undecanethiol (HS(CH2)10CH3). The adsorption of these thiols on the cantilever surface illustrates typical alkanethiol phase kinetics: a compressive stress on the surface which complies with previous vapor studies in the literature.12,23,24,26 Once exposed to the alkanethiols, the cantilever shows a gradual increase in deflection, with the deflection eventually reaching a steady state. Each alkanedithiol was introduced following a steady flow of ethanol through the system and was monitored in real time for approximately 1 h. The limiting factor on time was the volume in the syringe pump. Figure 2 shows the results of the cantilever deflection for alkanedithiols of varying length. In Figure 2a, the adsorption of 1,6-hexanedithiol (HS(CH2)6SH) on the gold surface creates a typical compressive stress on the cantilever. However, when exposed to the 1,8-octanedithiol (HS(CH2)8SH), Figure 2b, the cantilever initially shows a response similar to that of the 1,6dithiol (in a compressive manner) but gradually begins to bend the opposite way after approximately 2000 s. In Figure 2c, the adsorption of the 1,10-decanedithiol (HS(CH2)10SH) induces a tensile stress on the cantilever during its adsorption, thus resulting in a bend of the cantilever in the opposite direction. The cantilever’s response to the adsorption of the 1,6-hexanedithiol (Figure 2a) is consistent with the behavior of the adsorption of the hydroxythiols and the 11-undecanethiol (Figure 1). The 1,6-dithiol profile illustrates a steady state after adsorption relative to the ordering and assembly of the monolayer and shows a compressive stress on the cantilever. For 1,8-dithiol adsorption, the cantilever undergoes a compressive stress and then gradually begins to deflect the opposite way. This effect was reproducible through all of the data sets collected. In the first phase, normal kinetics of alkanethiol adsorption is observed. However, in the second phase, a steady state is never observed. It appears that the monolayer begins to undergo a change in conformation, thus creating a pull on the surface and resulting in the gradual bend of the cantilever in the opposite direction which was not observed for short chain dithiols or for monothiols of similar length. The 1,8-dithiol does not reach a steady state during the time
period studied. Data collection was not performed for long time durations due to the limitations in the volume of the syringe pump system. In the case of 1,10-dithiol adsorption, the change in conformation occurs at a much faster rate and results in a tensile stress on the surface during adsorption. The only difference between the alkanedithiols studied in this work is their alkyl chain length; the experimental data thus indicate that the self-assembly characteristics of these alkanedithiols are dependent on their chain length. In particular, the data suggest the possibility that the above-mentioned “pull” on the cantilever exerted by the longer dithiols results from the formation of a loop of the dithiols on the surface. On the other hand, shorter dithiols are unable to form such loops and consequently cannot exert a similar pulling force on the cantilever. To test this hypothesis, we have carried out Monte Carlo simulations of conformations of dithiol chains that are adsorbed on a planar surface to determine if there is a possibility of forming a loop structure. Monte Carlo Simulation of Conformations of Adsorbed Alkanedithiol Chains. The Monte Carlo simulations were used in this work with the primary purpose of determining whether a single alkane dithiol would be able to conform in a geometry which would allow it to bond at both thiol groups to the surface thus forming a loop. The model focuses on the conformational properties of a single chain that is attached to a surface; i.e., the model corresponds to zero coverage of the surface by the alkane. The schematic in Figure 3 shows two possible types of conformations resulting from the adsorption of alkanedithiols on a planar surface. Specifically, Figure 3a shows an ordered monolayer phase where the alkanedithiols orient in such a way that one thiol group is at the free surface of the monolayer and the other thiol group is attached to the gold surface. On the other hand, Figure 3b represents a monolayer consisting of mixed alkanedithiol orientation: some of the chains are oriented similar to those orientations in Figure 3a, whereas the others have formed a loop where both thiol end groups of the chain have formed a bond with the gold surface. Monte Carlo simulations were carried out to determine the fraction of adsorbed dithiol chains that can potentially fold back onto the surface, as is required if both thiol groups were to form bonds with the gold surface. It is well-known that chemical connectivity imposes strong constraints on the values of the bond lengths, bond angles, and the torsional (dihedral) angles of an alkane chain,38,39 and the reluctance of polymer chains to fold back on to themselves has been previously mentioned in the literature.40 In this work, conformations of dithiol chains that are adsorbed on a planar surface were generated in a bond-by-bond fashion by accounting for these constraints of the bond lengths (b), bond angles (θ), and the torsional angles (φ) as shown in Figure 3c, and the fraction of the adsorbed dithiol chains that can form a loop on the planar surface was quantified. (38) Flory, P. J. Statistical Mechanics of Chain Molecules; John Wiley & Sons: New York, 1969. (39) Mattice, W. L.; Suter, U. W. Conformational Theory of Large Molecules; John Wiley & Sons: New York, 1994. (40) Eichinger, B. E. ACS Symposium Series on Elastomers and Rubber Elasticity; American Chemical Society: Washington, DC, 1982; Vol. 193, p 243.
1262 Langmuir, Vol. 23, No. 3, 2007
Kohale et al.
Table 2. Percentage of Chains Looping Back to the Planar Surface for Various Chain Lengths over the Range of ψ3 chain length
a
ψ3
5
6
7
8
9
10
11
12
13
14
15
-180 -170 -120 -50 0 50 120 170 180
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
2 1 0 0 0 0 0 1 2
2 2 1 0 0 0 1 3 2
4.7 4 2.4 0.3 0 0.3 2.4 4 4.7
11.5 12 8.3 1.1 0.5 1.1 8.3 12 11.5
13.5 13.9 9.7 2.2 0.5 2.2 9.7 13.9 13.5
15.7 16.3 11.7 3.3 1.7 3.2 11.7 16.4 15.7
18.2 18.5 13.5 4.9 2.6 4.9 13.5 18.6 18.1
20.9 21 15.8 6.7 4.2 6.7 15.7 21.1 20.9
22.9 23.2 17.4 8.3 5.3 8.4 17.3 23.2 22.8
The results are shown for ψ1 ) ψ2 ) 33° and by considering the first torsional angle to be in the trans state.
As shown in Figure 3c and 3d, the coordinates of the chain atoms are specified by the three Eulerian angles (ψ1-ψ3),41 as well as the chain internal coordinates, namely, bond lengths, bond angles, and the set of torsional angles φ1, φ2, φ3, ... φn.38 The Eulerian angles characterize the orientation of the chain with respect to an external coordinate system such as the one defined in the context of the planar adsorbing surface. For the problem of interest here (fraction of chains folding back onto the surface), results are independent of the value of the angle ψ1. Previous work on methanethiolate42 has shown that the preferred values for the angle between the S-C bond and the surface normal (i.e. the angle ψ2) is either 15 or 33°. The entire range of values of ψ3 is considered in this work. Once the values of the Eulerian angles are set, the coordinates of the chain atoms are completely specified by the values of the internal coordinates: bond lengths (1.54 Å for C-C bond and 1.82 Å for the C-S bond); bond angles (112° for the C-C-C angle); torsional angles φi. In this work, chain conformations are generated by using the rotational isomeric state (RIS) theory38,39 for determining the values of the successive torsional angles φi. In RIS model for polyethylene (alkane), a chain is represented by assigning one of the 3 possible discrete states, trans, gauche+, or gauche-, to each of the torsional angles. These three torsional angle states correspond to the three minima on the potential energy surface of the chain. The model has been very successful in representing structural and conformational properties of a number of polymers.38,39 Given the energy differences between these three conformational states, the RIS model gives the conditional probability qξη;i for bond i being in state η given that the previous bond (i - 1) is in state ξ. As mentioned above, for an alkane chain, the possible values for the states η and ξ are trans, gauche+, or gauche-. In our calculations, we start with the literature values43 for the relative statistical weights of the three states trans, gauche+, or gauche- for an alkane chain. Using the procedure described by Mattice and Suter,39 RIS theory is then used to develop the values of the conditional probabilityqξη;i, given these statistical weights. All of the calculations were done at a temperature of 300 K. The conditional probability values qξη;i were used in a Monte Carlo procedure39 to generate chain configurations over a range of dithiol chain lengths from N ) 5 to 15 (e.g. for N ) 10, the chain consisted of 8 carbon atoms and 2 sulfur atoms in the terminal locations). For each chain configuration, the first sulfur atom was attached to the planar surface and the rest of the chain was then grown in a bond-by-bond fashion using the conditional probabilities qξη;i as described above. For each chain length, 2 000,000 different configurations were generated using the RIS probabilities and the fraction of the chain configurations that fold back to the surface was monitored. For this purpose, a chain was considered to cross the surface if the new bond being added to the growing chain intersects the plane of the surface. In few cases, the results were checked for 10 000 000 chain configurations and were found to be identical with the results obtained for 2 000 000 configurations within the statistical uncertainties. We note that such a Monte Carlo procedure makes two assumptions: (41) Theodorou, D. N.; Suter, U. W. Macromolecules 1985, 18, 1467. (42) Yourdshahyan, Y.; Rappe, A. M. J. Chem. Phys. 2002, 117, 825. (43) Abe, Y.; Flory P. J. J. Chem. Phys. 1970, 52, 2814.
(1) The modification of the chain conformations (probabilities qξη;i) by the presence of the surface are ignored. (2) The second end of the chain can form a bond only at specific locations on the surface. These geometric considerations are ignored in the current calculation procedure. These two effects will partially compensate each other for an attractive surface: relaxing the former assumption will increase the fraction of chains that form a loop while relaxing the latter will decrease the fraction of chains that form a loop. We consider both of these effects to be secondary (geometric and energetic considerations for chain internal coordinates being the primary effects) and take these assumptions to be reasonable for an initial analysis of loop formation ability of alkanedithiols.
Computational Results The principal result consists of the percentage of adsorbed dithiol chains that loop back to the planar surface. Tables 1 and 2 show these percentages as a function of the Eulerian angle ψ3 for chains of different lengths and for ψ2 ) 15° and ψ2 ) 33°, respectively.42 As can be seen from the tables, for chain lengths of up to 8 (1,6-hexanedithiol), hardly any of the chains fold back onto the surface. This is a direct consequence of the geometrical restrictions from the constraints of constant bond lengths and bond angles and the energetic restrictions on the values of the chain torsional angles. The fraction of chains that can loop back starts to increase sharply for chain lengths higher than 10 (1,8-octanedithiol). Note that these results were obtained by using the RIS theory based Monte Carlo chain generation procedure and assuming the first bond to be always in the trans state as has been reported in the literature.5 If the constraint of keeping the first angle in the trans state is relaxed, we still find that the percentage of chains that loop back is close to zero for N < 6 and begins increasing for longer chains. Tables 1 and 2 also show that the fraction of chains that loop back increases with the value of ψ2. If ψ2 is increased to 45°, the fraction of chains that loop back remains to be zero at shorter chain lengths but starts to increase after N ) 6. Finally, we once again point out that these computational results were obtained with a single-chain simulation, i.e., for a zero coverage model. The conformations of the chain are governed by both intramolecular and intermolecular interactions. The former are accounted for in our model, and the results show that the shorter chains cannot form a loop on the surface. For higher chain coverage of the surface, these intramolecular interactions will continue to be in effect and the conclusion that the shorter alkane chains are not able to form a loop will still be applicable. In addition, the crowding of chains in a monolayer formed at a higher surface coverage will lead to intermolecular interactions
Self-Assembled Monolayers of Alkanedithiols
between the chains. These interchain interactions are expected to cause the chains to form parallel arrays on the surface and thus further decrease the fraction of the longer alkane chains that are able to form a loop on the surface (but are not expected to completely preclude loop formation in the same manner as the geometrical constraints).
Discussion The computational results in conjunction with the experimental data suggest the following picture for the adsorption of dithiols on gold surface: (1) The percentage of chains looping back on the surface increases with alkyl chain length. (2) For dithiols shorter than 1,6-dithiol (N ) 8), the probability of the adsorbed chain forming a loop on the surface is negligible (Tables 1 and 2). Therefore, the monolayer will assemble on the surface such that one thiol group will form a bond to the surface and the alkane chains will uniformly tilt to create an ordered monolayer (Figure 3a), leaving the other thiol group at the free surface of the monolayer. (3) For slightly longer dithiols such as 1,8-dithiol, the probability of loop formation by the dithiol on the surface starts becoming nonzero. In the initial stages of the dithiol adsorption, most of the chains get adsorbed on the surface with one thiol group being bonded to the surface. However, with increased time, a small fraction of the chains may undergo a conformational change causing them to form a loop on the surface. (4) For much longer dithiols such as 1,10-dithiol, a significant fraction of the adsorbed dithiol chains can readily form a loop on the surface. Previous studies by Engelkes et al. have indicated that film thickness measurements of alkanethiols are indistinguishable from
Langmuir, Vol. 23, No. 3, 2007 1263
their complimentary alkanedithiols.28 However, ellipsometry data does not account for each molecule individually but describes the surface as a whole. It is thus possible that a fraction of the 1,10-dithiols (see Tables 1 and 2) and an even smaller fraction of the 1,8-dithiols get attached by both ends to the surface, and due to the sensitivity of a cantilever sensor, it is highly possible that the “molecular pull” on the surface is being detected due to a small percentage of “looped” alkanedithiols on the gold surface.
Conclusions In conclusion, we report experimental data and a molecular simulation study describing monolayer formation of alkanedithiols of varying lengths on a gold surface. Our work shows microcantilever sensors to be very sensitive indicators of chain conformations on an adsorbing surface. The results indicate that dithiols shorter than 1,6 hexanedithiol adsorb to the gold surface whereby only one thiol group forms a bond with the surface. This results in a compressive stress on the cantilever, very similar to that observed for the adsorption of thiols on a gold surface. On the other hand, for chain lengths N g 8, both thiol groups of the alkanedithiol molecules may get adsorbed on the gold surface, thus forming a chain “loop” on the surface. Monte Carlo simulations indicate that probability of this double attachment (loop formation) of dithiols to the surface increases significantly with the alkyl chain length. Formation of such a loop on the surface is the most likely explanation for the tensile stress observed with the cantilever for long-chain alkane dithiols. Acknowledgment. This work was supported by Texas Tech University Seed funds and the Department of Energy. LA062441N
