Adsorption Kinetics of 1-Alkanethiols on Hydrogenated Ge (111)

Langmuir 2004, 20, 835-840

835

Adsorption Kinetics of 1-Alkanethiols on Hydrogenated Ge(111) Madhava R. Kosuri, Roya Cone, Qiming Li, and Sang M. Han* Department of Chemical and Nuclear Engineering, The University of New Mexico, Albuquerque, New Mexico 87131

Bruce C. Bunker and Thomas M. Mayer Sandia National Laboratories, Albuquerque, New Mexico 87185 Received August 18, 2003. In Final Form: November 8, 2003 We have investigated the liquid-phase self-assembly of 1-alkanethiols (HS(CH2)n-1CH3, n ) 8, 16, and 18) on hydrogenated Ge(111), using attenuated total reflection Fourier transform infrared spectroscopy as well as water contact angle measurements. The infrared absorbance of C-H stretching modes of alkanethiolates on Ge, in conjunction with water contact angle measurements, demonstrates that the final packing density is a function of alkanethiol concentration in 2-propanol and its chain length. High concentration and long alkyl chain increase the steady-state surface coverage of alkanethiolates. A critical chain length exists between n ) 8 and 16, above which the adsorption kinetics is comparable for all long alkyl chain 1-alkanethiols. The steady-state coverage of hexadecanethiolates, representing long-chain alkanethiolates, reaches a maximum at approximately 5.9 × 1014 hexadecanethiolates/cm2 in 1 M solution. The characteristic time constant to reach a steady state also decreases with increasing chain length. This chain length dependence is attributed to the attractive chain-to-chain interaction in long-alkyl-chain selfassembled monolayers, which reduces the desorption-to-adsorption rate ratio (kd/ka). We also report the adsorption and desorption rate constants (ka and kd) of 1-hexadecanethiol on hydrogenated Ge(111) at room temperature. The alkanethiol adsorption is a two-step process following a first-order Langmuir isotherm: (1) fast adsorption with ka ) 2.4 ( 0.2 cm3/(mol s) and kd ) (8.2 ( 0.5) × 10-6 s-1; (2) slow adsorption with ka ) 0.8 ( 0.5 cm3/(mol s) and kd ) (3 ( 2) × 10-6 s-1.

Introduction Self-assembled monolayers (SAMs) on semiconductor surfaces1-7 present new possibilities to probe functional activities of biomolecules on synthetically created surfaces. This is similar to how SAMs on Au have been utilized. For instance, functionalized alkylthiolate SAMs on Au8,9 have provided an ideal test bed to study protein adsorptions and immunointeractions.10-13 In a parallel approach, we have demonstrated that 1-alkanethiols form a densely packed SAM on hydrogenated Ge(111) at room temper* To whom correspondence should be addressed. E-mail: [email protected]. (1) Linford, M. R.; Fenter, P.; Eisenberger, P. M.; Chidsey, C. E. D. J. Am. Chem. Soc. 1995, 117, 3145-3155. (2) Bansal, A.; Li, X. L.; Lauermann, I.; Lewis, N. S.; Yi, S. I.; Weinberg, W. H. J. Am. Chem. Soc. 1996, 118, 7225-7226. (3) Bergerson, W. F.; Mulder, J. A.; Hsung, R. P.; Zhu, X. Y. J. Am. Chem. Soc. 1999, 121, 454-455. (4) Linford, M. R.; Chidsey, C. E. D. J. Am. Chem. Soc. 1993, 115, 12631-12632. (5) Allara, D. L.; Parikh, A. N.; Judge, E. J. Chem. Phys. 1994, 100, 1761-1764. (6) He, J.; Lu, Z.-H.; Mitchell, S. A.; Wayner, D. D. M. J. Am. Chem. Soc. 1998, 120, 2660-2661. (7) Han, S. M.; Ashurst, W. R.; Carraro, C.; Maboudian, R. J. Am. Chem. Soc. 2001, 123, 2422-2425. (8) Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990, 112, 558-569. (9) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109, 3559-3568. (10) Feldman, K.; Ha¨hner, G.; Spencer, N. D.; Harder, P.; Grunze, M. J. Am. Chem. Soc. 1999, 121, 10134-10141. (11) Mrksich, M.; Grunwell, J. R.; Whitesides, G. M. J. Am. Chem. Soc. 1995, 117, 12009-12010. (12) Jung, L. S.; Nelson, K. E.; Stayton, P. S.; Campbell, C. T. Langmuir 2000, 16, 9421-9432. (13) Pe´rez-Luna, V. H.; O’Brien, M. J.; Opperman, K. A.; Hampton, P. D.; Lo´pez, G. P.; Klumb, L. A.; Stayton, P. S. J. Am. Chem. Soc. 1999, 121, 6469-6478.

ature.7 The resulting SAM exhibits a high degree of orientational ordering with the alkyl chains tilted at approximately 20° with respect to the surface normal. The thiolates are covalently bonded to the Ge surface by Ge-S(CH2)n-1CH3 linkage, where n represents the number of methylene groups in the alkyl chain. The study has also shown that the thiolate SAM is thermally stable up to 450 K. Herein, we focus on the adsorption kinetics of 1-alkanethiols on hydrogenated Ge(111). The purpose is to increase our understanding of adsorption kinetics and to prepare reliably and predictably high-quality thiolate SAMs on Ge. Utilizing water contact angle measurements, we have determined the effect of concentration and chain length of 1-alkanethiols on their adsorption kinetics and ultimately their steady-state surface coverage. We have also measured the adsorption and desorption rate constants of 1-hexadecanethiol (0.1 M in 2-propanol) on hydrogenated Ge(111), using attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIRS). The IR technique provides submonolayer sensitivity to surface adsorbates and proves useful in monitoring the adsorption kinetics. The technique also reveals the chemical nature of the molecules adjacent to the ATR crystal surface by their signature IR absorbance peaks.14,15 Experimental Methods Sample Preparation and Monolayer Formation. The asreceived Ge(111) samples are first sonicated in acetone for 5 min to dissolve organic contaminants. The samples are then immersed (14) Vigano, C.; Manciu, L.; Buyse, F.; Goormaghtigh, E.; Ruysschaert, J.-M. Biopolymers 2000, 55, 373-380. (15) Grimard, V.; Vigano, C.; Margolles, A.; Wattiez, R.; Veen, H. W. v.; Konings, W. N.; Ruysschaert, J.-M.; Goormaghtigh, E., Biochemistry 2001, 40, 11876-11886.

10.1021/la035521p CCC: $27.50 © 2004 American Chemical Society Published on Web 12/30/2003

836

Langmuir, Vol. 20, No. 3, 2004

Kosuri et al.

Figure 1. Experimental setup to monitor the liquid-phase adsorption kinetics of 1-alkanethiols on HF-treated Ge(111) using attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIRS). in H2O2 solution (30 wt % in H2O) for 30 s for surface oxidation16,17 and rinsed with deionized (DI) H2O for 30 s to dissolve the oxidation layer.17-19 The samples are subsequently treated with HF solution (49 wt % in H2O) for 30 s to hydrogenate the Ge surface.7,20 After each solution treatment, the samples are blowdried with N2. The H2O2 and HF treatments with intermediate DI H2O rinse steps are repeated three times to remove carbon contaminants and to obtain a smooth Ge surface.17 The smoothness of the Ge surface is independently verified by atomic force microscopy. The root-mean-square surface roughness of Ge prior to alkanethiolate SAM formation is approximately 7 Å. Note that the final HF treatment is conducted in a dilute solution (10 wt % in H2O) for 10 min to fully hydrogenate the Ge surface.20 The DI H2O rinse step is omitted after the final HF treatment to preserve the hydrogen termination of Ge(111). In our independent investigation, the alkanethiolate SAM prepared with the H2O rinse loses hydrophobicity, suggesting that dehydrogenation occurs during H2O rinse. The HF-treated samples are blow-dried with N2 before they are immersed in a 1-alkanethiol solution (10-4-1 M in 2-propanol). The container holding the thiol solution is sealed with Parafilm to minimize the evaporation of 2-propanol. After the monolayer formation is complete, the samples are sonicated in 2-propanol for 2 min to remove physically adsorbed alkanethiols and blow-dried with N2. The above preparation steps are performed in the ambient air (38% relative humidity) at room temperature (18-23 °C). Attenuated Total Reflection Fourier Transform Infrared Spectroscopy (ATR-FTIRS). Figure 1 shows a crosssectional view of the ATR-FTIRS experimental setup to probe the self-assembly of organic molecules occurring on semiconductor surfaces. A rectangular Teflon trough holds the 1-alkanethiol solution in 2-propanol. The Ge ATR crystal is placed at the bottom of the trough and sealed with a Kalrez ring to hold the solution. The IR beam from a spectrometer is focused onto a 45° beveled edge of a trapezoidal Ge(111) ATR crystal via flat and off-axis paraboloid IR mirrors. The length, width, and thickness of a typical Ge ATR crystal are 50, 10, and 2 mm, respectively. This geometry results in 12 internal reflections on the top surface. Figure 1 also conceptually shows how 1-alkanethiol molecules adsorb to the Ge top surface. The internal reflections create an evanescent electric field, E(z), at the Ge-SAM interface. The evanescent field permeates into the 1-alkanethiol solution and interacts with IR-active species (e.g., CH3 and CH2 of alkanethiolates adsorbed on Ge and 1-alkanethiols in the solution). Each reflection amplifies the IR absorbance and provides sub-monolayer detection sensitivity to the alkanethiolates. The thickness of organic SAMs is typically less than 40 Å and substantially less than the penetration depth of the evanescent wave (0.2-1.5 µm). Thus, the IR absorbance is not corrected for the exponentially decaying strength of the evanescent wave. Note that the adsorbed (16) Prabhakarana, K.; Ogino, T.; Hull, R.; Bean, J. C.; Peticolas, L. J. Surf. Sci. 1994, 316, L1031-L1033. (17) Okumura, H.; Akane, T.; Matsumoto, S. Appl. Surf. Sci. 1998, 125, 125-128. (18) Prabhakaran, K.; Ogino, T. Surf. Sci. 1995, 325, 263-271. (19) Zhang, X.-J.; Xue, G.; Agarwal, A.; Tsu, R.; Hasan, M.-A.; Greene, J. E.; Rockett, A. J. Vacuum Sci. Technol., A 1993, 11, 2553-2561. (20) Choi, K.; Buriak, J. M. Langmuir 2000, 16, 7737-7741.

Figure 2. Water contact angles on alkanethiolate-terminated Ge(111) as function of concentration in 2-propanol and alkyl chain length. alkanethiolate and the evanescent wave strength are not drawn to scale in Figure 1. After the entire waveguide is traversed, the IR beam exits the opposite beveled edge, and a HgCdTe IR detector measures the outgoing beam intensity. A background IR spectrum averaged over 300 scans at 2 cm-1 resolution is taken before the final HF treatment. After the 1-alkanethiol solution is poured, a series of sample spectra are taken with 2 cm-1 resolution during SAM formation. Each sample spectrum is averaged over 300 scans. Contact Angle Analysis. VCA-2000 instrument from AST Products, Inc., is used to measure static contact angles of sessile water drops on thiolated Ge. The contact angle measurements are made with DI H2O within a day of sample preparation. The contact angle is found to be independent of the droplet volume in the 1-20 µL range, and the measurement reproducibility is (3°.

Results and Discussion Figure 2 shows water contact angles measured on Ge(111) surface passivated by 1-octadecanethiol (ODT), 1-hexadecanethiol (HDT), and 1-octanethiol (OT). ODT and HDT represent long alkyl chain alkanethiols while OT represents short alkyl chain alkanethiols. The alkanethiol concentration (Cs) ranges from 10-4 to 1 M in 2-propanol, spanning 4 decades. The hydrogenated Ge samples are immersed in the alkanethiol solutions for 16 h. At each concentration, this exposure time reproducibly results in maximum achievable contact angles. That is, the upper bound on the characteristic time constant to reach a steady-state alkanethiolate surface coverage is 16 h for the concentration range explored in this experiment. Our later discussion will illustrate that increasing concentration and chain length shorten this time period. The water contact angle quantifies the hydrophobicity of an alkanethiolate-terminated Ge surface as an indirect measurement of alkanethiolate surface coverage. Figure 2 therefore reveals how the alkanethiol concentration and the alkyl chain length govern the steady-state surface coverage of alkanethiolates. Increasing concentration results in increasing contact angle, suggesting increasing surface coverage. The IR absorbance of C-H stretching vibrational modes8,9,21 of alkanethiolate SAM prepared at various concentrations (Figure 3) provides corroborating evidence that the steady-state surface coverage increases with increasing alkanethiol concentration. The steadystate absorbance spectra are taken after the solution is removed from the Teflon trough in the ATR-FTIRS setup, (21) Walczak, M. M.; Chung, C.; Stole, S. M.; Widrig, C. A.; Proter, M. D. J. Am. Chem. Soc. 1991, 113, 2370-2378.

1-Alkanethiols on Ge(111)

Langmuir, Vol. 20, No. 3, 2004 837

Figure 3. Steady-state infrared absorbance spectra of hexadecanethiolate self-assembled monolayer (SAM) prepared in various concentrations. Increasing concentration results in increasing surface coverage of hexadecanethiolates. The inset shows red shifting C-H stretching vibrational modes with increasing concentration and therefore increasing surface coverage. The red-shifted peak positions correspond to those of crystalline polymethylene chain.

following the 16-h exposure and after the Ge ATR crystal is sonicated in neat 2-propanol. The IR absorbance monotonically increases with increasing concentration, qualitatively verifying the concentration dependence. In the inset of Figure 3, the observed peak positions of C-H stretching vibrational modes8,9,21 of CH2 also red shift from 2928 and 2854 cm-1 to 2921 and 2851 cm-1, respectively, as the concentration increases. The hexadecanethiolate SAM formed in a 10-4 M solution exhibits peak positions that are similar to those of the liquid-state hexadecanethiol, whereas the hexadecanethiolate SAM formed in a 1 M solution exhibit peak positions of the crystalline polymethylene chain.9,22 The increasing surface coverage of alkanethiolates is known to cause such structural ordering in the SAM.9,22 For quantitative analysis, the IR absorbance integrated over 2980-2830 cm-1 can be converted to the absolute surface coverage (N) of alkanethiolates by

1 nRnC-H

∫νν

2

1

A(ν) dν ) RN

(1)

where A, nR, nC-H, ν1 to ν2, and R denote the IR absorbance, the number of internal reflections, the number of C-H bonds per 1-alkanethiol molecule, the limits of integration in wavenumbers (cm-1), and the absorptivity of C-H stretching mode, respectively. N is given in alkanethiolates/cm2. For the absolute hexadecanethiolate coverage, the values corresponding to nR, nC-H, and R are 12, 33, and 2.1 × 10-18 cm/C-H bond. The value for R is based on the absorptivity reported for pentane.23 Figure 4 plots the absolute steady-state coverage (Ns.s.) of hexadecanethiolates calculated from eq 1. The corresponding scale is labeled on the right-hand-side ordinate. The absolute coverage increases with increasing concentration, validating the qualitatively observed concentration dependence. However, the concentration dependence is less pronounced in dilute and concentrated regions near 10-4 and 1 M. The absolute coverage reaches a maximum level (Nmaxs.s.) at 5.9 × 1014 hexadecanethiolates/cm2 for the SAM (22) Bensebaa, F.; Voicu, R.; Huron, L.; Ellis, T. H.; Kruus, E. Langmuir 1997, 13, 5335-5340. (23) Aubuchon, C. M.; Davison, B. S.; Nishimura, A. M.; Tro, N. J. J. Phys. Chem. 1994, 98, 240-244.

Figure 4. Steady-state fractional and absolute surface coverages (θSAMs.s. and Nmaxs.s.) of hexadecanethiolates as a function of concentration. The model fit (- - -) traces the experimental data with a correlation coefficient (R) equal to 0.94.

formed in 1 M solution. This surface coverage translates to approximately 80% of Ge atom density on Ge(111) surface, exceeding the absolute saturation coverage of alkanethiolate SAMs on Au24 and that of alkane SAMs on Si.25-27 The calculated Nmaxs.s. of hexadecanethiolate SAM on Ge(111) appears to be an overestimation, likely due to the inaccuracy associated with R. To extract the kinetic parameters that govern the alkanethiol adsorption on Ge, the steady-state fractional surface coverage (θSAMs.s.) of alkanethiolates is first calculated by

θSAMs.s. ) Ns.s./Nmax s.s.

(2)

A mathematical equivalent is the steady-state integrated absorbance (Iabss.s.) normalized by its maximum value (Iabs,maxs.s.)

θSAMs.s. ) Iabss.s./Iabs,maxs.s.

(3)

For the hexadecanethiolate SAM, Iabs,maxs.s. corresponds to the C-H IR absorbance integrated over 2980-2830 cm-1 for the SAM formed in 1 M solution. Equations 2 and 3 essentially convert Ns.s. to θSAMs.s. in Figure 4 with the corresponding scale on the left-hand-side ordinate. We have assumed a first-order Langmuir isotherm for the 1-alkanethiol adsorption on Ge(111), much the same way 1-alkanethiols adsorb on Au28-31 with the only exception that the adsorption requires surface H. We speculate that the overall reaction follows ka

z RS - Ge(s) + H2(solv) Ge - H(s) + RS - H(solv) y\ kd (4) Since the exact elemental steps of alkanethiol adsorption on hydrogenated Ge(111) remain unknown, we employ (24) Poirier, G. E.; Tarlov, M. J. Langmuir 1994, 10, 2853-2856. (25) Sieval, A. B.; Hout, B. v. d.; Zuilhof, H.; Sudho¨lter, E. J. R. Langmuir 2000, 16, 2987-2990. (26) Sieval, A. B.; Hout, B. v. d.; Zuilhof, H.; Sudho¨lter, E. J. R. Langmuir 2001, 17, 2172-2181. (27) Zhang, L.; Wesley, K.; Jiang, S. Langmuir 2001, 17, 6275-6281. (28) Poirier, G. E.; Pylant, E. D. Science 1996, 272, 1145-1148. (29) Karpovich, D. S.; Blanchard, G. J. Langmuir 1994, 10, 33153322. (30) Hu, K.; Bard, A. J. Langmuir 1998, 14, 4790-4794. (31) Thomas, R. C.; Sun, L.; Crooks, R. M. Langmuir 1991, 7, 620622.

838

Langmuir, Vol. 20, No. 3, 2004

Kosuri et al.

the following rate expression for the fractional surface coverage (θSAM)29

dθSAM/dt ) kaCs(1 - θSAM) - kdθSAM

(5)

where ka and kd denote adsorption and desorption rate constants, respectively. Integrating eq 5 with respect to time (t) results in an analytical expression for θSAM30

(

θSAM(t) ) 1 +

)

kd 1 ka Cs

-1

{1 - exp[-(kaCs + kd)t]} (6)

The steady-state fractional surface coverage (θSAMs.s.) as t f ∞ is therefore the pre-exponential factor in eq 6

(

θSAMs.s. ) θSAM(t f ∞) ) 1 +

kd 1 ka Cs

)

-1

(7)

Equation 7 describes how θSAMs.s. behaves as a function of kd/ka ratio and Cs. For a known set of θSAMs.s. and Cs, the kd/ka ratio can be used as a fitting parameter to model the experimental results in Figure 4. The best fit with a correlation coefficient (R) equal to 0.94 results from

kd/ka ) 3.5 × 10-3

(8)

The best fit is shown by the dotted line in Figure 4. Equation 8 will be used in a later discussion to estimate ka and kd. The only noticeable deviation from the model appears when the concentration is near 10-4 M. The origin of this deviation remains suspect. However, the nonzero θSAMs.s. at low concentration entails that the kd/ka ratio in eq 8 sets an upper limit on the estimate. That is, the extracted ka, in relation kd, can be higher than the calculated value. The overall functional behavior of θSAMs.s. and its dependence on kd/ka indicate that a subtle equilibrium between adsorption and desorption of alkanethiolates determines θSAMs.s.. In addition to the concentration dependence, the contact angles in Figure 2 indicate that short-alkyl-chain OT requires higher concentrations than long-alkyl-chain ODT and HDT to reach the same level of surface coverage. This chain length dependence is attributed to the chain-tochain interaction that increases with increasing chain length. This attractive interaction reduces the kd/ka ratio, resulting in increased packing density for long-chain alkanethiolates. Therefore, the kd/ka ratio for OT adsorption is expected to be greater than 3.5 × 10-3 of HDT. Figure 5 supports this argument by showing that the average water contact angle of HDT SAM increases at a faster rate than that of OT SAM as a function of exposure length in 1 M solution. The water contact angle for HDT SAM reaches 108° within 5 min, compared to 16 h for OT SAM. A later discussion will reveal that this rise time (5 min) is consistent with the characteristic time constant calculated from (kaCs + kd)-1 in eq 6. Note also that the chain length dependence is strongly pronounced between n ) 8 and 16 in the Ge system. Similar chain-length dependence is observed for 1-alkanethiol adsorption on Au.9,32 The chain-length dependence in an Au system is most pronounced between n ) 5 and 11.9,32 The shifted range (5-10 for Au to 8-16 for Ge) of chain length for pronounced increase in hydrophobicity and film structure suggests that the kd/ka ratio on Ge is greater than that on Au and that the Ge-S bond is relatively weaker than the Au-S bond. (32) Bain, C. D.; Troughton, E. B.; Tao, Y.-T.; Evall, J.; Whitesides, G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321-335.

Figure 5. Water contact angles measured on alkanethiolateterminated Ge(111) prepared in various exposure lengths in 1 M solution. The long-alkyl-chain 1-hexadecanethiol exhibits faster adsorption kinetics than the short-alkyl-chain 1-octanethiol.

The ATR technique is used to determine ka and kd of HDT SAM formation on hydrogenated Ge(111). The HDT SAM is representative of long chain 1-alkanethiols where n g 16. A background spectrum is collected after the final H2O2 treatment. A time series of IR absorbance spectra are taken while Ge(111) surface is immersed in 0.1 M 1-hexadecanethiol solution. The 0.1 M concentration is chosen to set the characteristic time constant (kaCs + kd) - 1 on the order of hours, such that the changes in IR absorbance can be observed with reasonable time resolution while the packing density of the resulting SAM approaches near maximum value. Such time series IR spectra consist of contributions from surface hexadecanethiolates, 1-hexadecanethiol, and 2-propanol. To monitor the changes in IR absorbance solely due to the surface hexadecanethiolates, the IR absorbance of neat 2-propanol is first subtracted from the time series spectra until the IR absorbance by C-O stretching vibration mode33 of 2-propanol near 1180-1075 cm-1 becomes zero. The resulting spectra yet contain contributions from hexadecanethiolates on the Ge surface and 1-hexadecanethiol molecules in the solution. The solution contribution is approximated by the first absorbance spectrum in the series. This is an overestimation since the first spectrum is taken over 3 min during which hexadecanethiolate self-assembly progresses to an extent. The solution contribution is subsequently subtracted from the remaining time series spectra to isolate the IR absorbance solely due to the hexadecanethiolates on Ge(111). The subtraction method is an approximation in lieu of multivariate analysis.34,35 The resulting time series spectra taken over an 8-h period are shown in Figure 6. The asymmetric C-H stretching vibrational mode8,9,21 of CH3 appears near 2958 cm-1, while asymmetric and symmetric C-H stretching vibrational modes8,9,21 of CH2 appear near 2919 and 2850 cm-1, respectively. Similar to the peaks shown in Figure 3, these peak positions reflect the structure of a crystalline polymethylene chain.9 The integrated absorbance of the time series spectra is converted to N according to eq 1 and to θSAM according to eq 3 where Iabss.s. is replaced with (33) Roeges, N. P. G., Normal Vibrations and Absorption Regions of CHX. In Guide to the Complete Interpretation of Infrared Spectra of Organic Structures; John Wiley & Sons: New York, 1994; p 129. (34) Johnson, R. A.; Wichern, D. W. Applied Multivariate Statistical Analysis, 5th ed.; Prentice Hall: Englewood Cliffs, NJ, 2002. (35) Malinowski, E. R., Factor Analysis in Chemistry, 2nd ed.; John Wiley & Sons: New York, 1991.

1-Alkanethiols on Ge(111)

Langmuir, Vol. 20, No. 3, 2004 839 Table 1. Calculated Adsorption and Desorption Kinetic Constants (ka and kd) for the Two-Stage Adsorption Process of 1-Alkanethiols on Ge(111)a ka, cm3/(mol s) kd, s-1

stage 1

stage 2

2.4 ( 0.2 (8.2 ( 0.5) × 10-6

0.8 ( 0.5 (3 ( 2) × 10-6

a Fast adsorption occurs during the first stage, followed by slow adsorption with ordering of straightened alkyl chains during the second stage.

Figure 6. Infrared absorbance spectra in the region of C-H stretching vibrational modes of hexadecanethiolate on Ge(111) taken in situ and in real time during self-assembly.

exponential expression for θSAM in eq 6 for the initial stage of adsorption process. As an additional test of the proposed first order Langmuir-Hinshelwood reaction mechanism described by eq 5, the theoretical time constant for 1 M solution is calculated to be 7 min, which compares well with the experimentally observed 5-min rise time in Figure 5. That is, the characteristic time constant or the rise time decreases with increasing concentration. The observed rate constants, kobs1 ) 0.88 h-1 for the first 2 h and kobs2 ) 0.3 h-1 for the remaining 6 h, give rise to model fits to θSAM shown by the dotted lines in Figure 7. The model fits trace the experimental data with a correlation coefficient (R) equal to 0.92 for the first stage and 0.68 for the second stage. We suspect that the adsorption kinetics during the second stage is quite more complicated than the simple first order LangmuirHinshelwood reaction kinetics, thus resulting in reduced goodness of fit for the second stage. The large uncertainty in rate constants for the second stage is also due to the experimental scatter in the IR absorbance during the second stage. Conclusions

Figure 7. Fractional surface coverage of hexadecanethiolates on Ge(111) as a function of time. A first order Langmuir isotherm describes the adsorption process. The inset demonstrates how kinetic parameters can be extracted from the ln[1 - (1 + (kd/ ka)(1/Cs))θSAM] vs t plot.

Iabs(t). Figure 7 shows the resulting N and θSAM, and the inset shows a plot of ln[1 - (1 + (kd/ka)(1/Cs))θSAM] vs t. Two sets of experimental data demonstrate the level of reproducibility. Equation 8 is substituted into ln[1 - (1 + (kd/ka)(1/Cs))θSAM] for the kd/ka ratio, and 0.1 M is used for Cs. The slope of the plot in the inset gives -(kaCs + kd) according to eq 6. We will hereafter denote kaCs + kd as kobs for convenience. Two different slopes (kobs1 and kobs2) are evident in the inset, indicating that the adsorption takes place in two stages. Previous investigations28,30,32,36 on Au analogously suggest that alkanethiolates on Ge undergo (1) rapid adsorption with strongly entangled alkyl chains during the first stage and (2) slow adsorption with ordering of straightened alkyl chains during the second stage. Linear regressions on the experimental data in the inset result in statistically averaged kobs1 and kobs2 at 0.88 ( 0.061 and 0.3 ( 0.2 h-1 for the first and second stages, respectively. These kobs1 and kobs2 values (i.e., kaCs + kd) in conjunction with eq 8 yield ka and kd for the two stages of adsorption. Table 1 summarizes the outcome in cgs units. Using ka and kd for the first stage and 0.1 M for Cs, the characteristic time constant, (kaCs + kd) - 1, for the HDT SAM formation is 1.1 h. The experimental data in Figure 7 show that θSAM is 0.64 at 1.1 h, which virtually agrees with 1 - 1/e. This agreement further validates the (36) Ha¨hner, G.; Wo¨ll, C.; Buck, M.; Grunze, M. Langmuir 1993, 9, 1955-1958.

We have employed water contact angle measurements as well as ATR-FTIRS to investigate the adsorption kinetics of 1-alkanethiols on hydrogenated Ge(111). The contact angle measurements reveal that the alkanethiol concentration in 2-propanol and the alkyl chain length play a critical role in the adsorption kinetics, ultimately controlling the steady-state fractional surface coverage of alkanethiolates (θSAMs.s.). On the basis of a first order Langmuir-Hinshelwood mechanism where the adsorption and desorption rates (ka and kd) compete, θSAMs.s. is a function of both kd/ka ratio and alkanethiol concentration (Cs). The experimentally measured θSAMs.s. (Figure 4) substantiates this functional dependence. Equation 7 stemming from our model also explicitly states this functionality. Thus, θSAMs.s. increases with increasing Cs. Increasing chain length also results in increasing θSAMs.s., indicated by the increasing water contact angles. This chain length dependence is attributed to the attractive chain-to-chain interaction in long alkyl chain SAMs. The favorable interaction reduces the kd/ka ratio, resulting in increased θSAMs.s. at a given concentration. For 1-hexadecanethiol that represents long-chain alkanethiols, the kd/ka ratio calculated from a model fit is 3.5 × 10-3 at room temperature. On the basis of the Langmuir isotherm experimentally measured by ATR-FTIRS, the adsorption of 1-hexadecanethiols on hydrogenated Ge(111) undergoes two stages: (1) fast adsorption with ka ) 2.4 ( 0.2 cm3/ (mol s) and kd ) (8.2 ( 0.5) × 10-6 s-1 during the first stage followed by (2) slow adsorption with ordering of alkyl chains with apparent ka ) 0.8 ( 0.5 cm3/(mol s) and kd ) (3 ( 2) × 10-6 s-1 during the 2nd stage. At 0.1 M, the fast adsorption occurs over a 2-h period, following an exponential Langmuir isotherm. The ensuing chain ordering

840

Langmuir, Vol. 20, No. 3, 2004

takes 6 h. The observed two-stage kinetic behavior is very similar to that of 1-alkanethiols on Au. Acknowledgment. The authors thank University of New Mexico SEED, Sandia University Research Program (Contract Number 13575), and National Science Foundation Nanoscale Exploratory Research Program (Award

Kosuri et al.

Number CTS-0304237) for generous financial support. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy under Contract DE-AC0494AL85000. LA035521P