Adsorption and Thermal Decomposition of Alkanethiols on Cu (110)

Synchrotron Radiation Research Center, Hsin-Chu Science-Based Industrial .... of Methanethiolate and Atomic Sulfur at the Cu(111) Surface: A Computati...
8 downloads 0 Views 135KB Size
Download PDF
5438

J. Phys. Chem. B 2002, 106, 5438-5446

Adsorption and Thermal Decomposition of Alkanethiols on Cu(110) Ying-Huang Lai and Chuin-Tih Yeh Department of Chemistry, National Tsing-Hua UniVersity, Hsin-Chu 300, Taiwan

Shu-Hua Cheng Department of Applied Chemistry, National Chi-Nan UniVersity, Pu-Li 545, Taiwan

Pammy Liao and Wei-Hsiu Hung* Synchrotron Radiation Research Center, Hsin-Chu Science-Based Industrial Park, Hsin-Chu 300, Taiwan ReceiVed: December 31, 2001; In Final Form: March 26, 2002

Adsorption and thermal decomposition of alkanethiols (RSH, R ) CH3, C2H5, and C4H9) on a Cu(110) surface have been studied by means of temperature-programmed desorption (TPD) and X-ray photoemission spectroscopy (XPS) with synchrotron radiation. At a small coverage, CH3SH and C2H5SH dissociate to form surface thiolates and hydrogen, whereas C4H9SH adsorbs molecularly on the surface at 100 K; adsorbed C4H9SH begins to deprotonate at ∼170 K. All of these alkanethiolates can decompose to evolve hydrocarbon via scission of the C-S bond, resulting in deposition of sulfur on the surface. CH3 generated from CH3S reacts with surface hydrogen to evolve CH4, but at a large coverage can also undergo coupling to form C2H6. The thermal reaction of surface C2H5 formed from C2H5S produces C2H6 through hydrogenation and C2H4 through β-hydride elimination, the latter with desorption of H2; the ratio of products C2H4 and C2H6 varies with the adsorption site of surface C2H5S. The C4H9 group of surface C4H9S undergoes exclusively elimination of β-hydride to form C4H8. To a small extent the alkyl moiety also undergoes dehydrogenation, resulting in deposition of carbon on the surface. Three CH4 desorption states are proposed to correspond to CH3S intermediates with distinct adsorption sites after decomposition of CH3SH, whereas two C2H5S sites and one C4H9S site are proposed for cases of C2H5SH and C4H9SH, respectively. The distribution of desorption products depends on the nature of the adsorption site of a particular alkanethiolate.

Introduction Adsorption and thermal decomposition of alkanethiols on various transition metals have industrial importance. The thermal reaction of alkanethiols on surfaces is associated with catalytic poisoning and hydrodesulfurization in the petroleum industry, as such thiols occur commonly in crude oil.1,2 Another aspect pertains to preparation of a self-assembled monolayer (SAM) on a metallic surface, which can modify the chemical and physical properties of that surface and is potentially adaptable for such applications as wetting, lubrication, and corrosion.3-5 Alkanethiol molecules can form well-ordered thiolate adlayers on noble transition metals such as Au and Ag through deprotonation of alkanethiols in solution.3-14 It has also been reported that an alkanethiolate SAM can be obtained to act as a barrier film to retard oxidation of a copper surface.15-17 However, a copper surface is more reactive than that of Au or Ag, and might facilitate scission of the C-S bond to form sulfur and alkyl species on the surface.18-24 As a consequence, the adlayer of alkanethiolate on the copper surface is not as densely packed as in the case of gold. The thermal stability and chemical composition of a thiolate adlayer on a Cu surface, derived from analogy with the reaction of alkanethiols on metallic surfaces, is incompletely understood. Investigation of the dissociation of * Corresponding author. Fax: +886-3-5789016. E-mail: whung@ srrc.gov.tw.

alkanethiols on reactive metal surfaces can therefore provide critical insight into mechanisms of overlayer formation and film breakdown because the thermal reaction and stability of an adlayer affects the subsequent fabrication and application of a SAM.4,25 As CH3SH is the simplest sulfur-containing organic compound, much work in relation to adsorption and decomposition of CH3SH on various transition-metal surfaces is reported.18,26-36 Except on an Au surface, which is inert, CH3S intermediate is formed on thermal decomposition of CH3SH via deprotonation on a transition-metal surface; this intermediate can occupy various surface sites and can display various geometries, such as on Cu (100), Cu(111), Ni(111), Ru(0001), and W(001), dependent on coverage and surface temperature.18,26,27,34-37 It is generally found that at high temperatures and large coverage thermal decomposition of CH3S leads to evolution of CH4, leaving atomic sulfur. We report here an investigation of the mechanism of adsorption and decomposition of alkanethiols on a Cu (110) surface, utilizing temperature- programmed desorption (TPD) and X-ray photoelectron spectroscopy (XPS) techniques; such spectral examination elucidates pathways of this surface reaction. Our spectra indicate that, on decomposition of CH3SH, CH3S forms and resides at three distinct sites, which exhibit varied reactive behavior. In principle, the length of the carbon chain of an alkanethiol might exert a significant influence on the thermal stability and reaction mechanism due to the presence

10.1021/jp0146869 CCC: $22.00 © 2002 American Chemical Society Published on Web 05/04/2002

Adsorption of Alkanethiols on Cu(110)

J. Phys. Chem. B, Vol. 106, No. 21, 2002 5439

of β-hydrogen and the interaction between a metallic surface and alkyl groups.38-41 We thus undertook a comparison of thermal reactivity and reaction products for alkanethiols with varied length of alkyl chain (i.e., RSH with R ) CH3, C2H5, and C4H9), thus probing the formation and thermal stability of alkanethiolate adlayer on the Cu surface, so as to provide mechanistic insight into deposition of sulfur on a catalytic surface of Cu and to serve as a useful reference for adsorption of alkanethiols with longer chains. Experimental Section Experiments were performed in an ultrahigh vacuum (UHV) chamber with a base pressure of 2 × 10-10 Torr. The system was equipped with a quadrupole mass filter (EPIC, Hiden), LEED, and an electron-energy analyzer (HA100, VSW). A Cu(110) sample was mounted on a button heater that was in turn mounted on a copper block; the Cu sample could be cooled to 100 K with liquid nitrogen via conduction through the copper block and heated with the button heater. The surface temperature was measured with a thermocouple (K-type) inserted into a pinhole located at the edge of the Cu sample. A cold-cathode ion gun (AG5000, VG Microtech) provided sputtering ions to clean the surface. Gaseous Ar was directly fed into the ion gun via a leak valve so that the system operated at pressures smaller than with back fill. The Cu surface was cleaned with 1 keV Ar+ at an incident angle 45° and annealed to ∼820 K for 1 min. The cleanliness of the Cu surface was verified with LEED and XPS measurements. Before use, C2H5SH (>99%, Merck) and C4H9SH (>98%, Merck) liquids were subjected to several freeze-pump-thaw cycles. CH3SH (>99.5%, Matheson) was used without further purification. Alkanethiols were introduced onto the Cu surface via a stainless steels tube with a pinhole (dia. 250 µm). During dosing, partial pressures of alkanethiols were controlled at 5 × 10-10 Torr and the sample surface was placed ∼2 cm in front of the doser pinhole to minimize contamination of the UHV system with thiols. XPS were measured at the wide-range beamline of SRRC (Synchrotron Radiation Research Center, Taiwan); the incident angle of photons was 55° from the surface normal. Emitted photoelectrons were collected with the electron analyzer normal to the sample surface in an angle-integrated mode. Calibration of the binding energy was performed by locating the Fermi edge of a sputtering-cleaned Au foil at 0.0 eV. The overall instrumental resolution for the XPS measurement was determined from the width of the Fermi edge, which was dependent on the photon source and the parameter setting of the electron energy analyzer. For measurements at varied temperature, the sample was heated to a desired temperature at a rate of 1 K/s and cooled immediately to 100 K, at which XPS spectra were recorded. Collected spectra were numerically fitted with the Gaussianbroadened Lorentzian function after Shirley background subtraction with a third-order polynomial to each side of the peak. In the curve fitting of the S 2p spectra, the spin-orbit splitting and the intensity ratio of p1/2 to p3/2 were kept at 1.18 eV and 0.5, respectively. The full widths at half-maximum (fwhm) of the 2p spin-orbit doublets were varied in the range of 0.6-0.9 eV, depending on the surface species. The quadrupole mass filter rendered analysis of desorption products in the TPD measurement. The mass analyzer was enclosed in a differentially pumped cylinder, at the end of which is a skimmer with an entrance aperture (diameter 2 mm). For TPD measurement, the sample surface was placed about 2 mm before the aperture and in line of sight of the ionizer of the

Figure 1. XPS spectra of S 2p and C 1s taken from a Cu surface for varied duration of exposure to CH3SH at 100 K. Dots represent data collected after background subtraction; solid lines are fitted curves, and various components are shown in dashed lines. The photon energy used to collect these spectra is 400 eV.

mass spectrometer; TPD scans were recorded on ramping the sample at a linear rate of ∼2 K/s. Results and Discussion Surface species upon adsorption and thermal fragmentation of CH3SH on a Cu surface were identified chemically using X-ray photoelectron spectroscopy. Figure 1 shows spectra in S 2p and C 1s regions collected from a Cu(110) surface exposed to CH3SH for varied duration. At small exposure, the S 2p3/2 binding energy is observed predominantly at 162.4 eV, which is attributed to surface methyl thiolate (CH3S) and discussed further below. Two additional S 2p3/2 signals develop subsequently at 164.5 and 165.2 eV with increasing duration of exposure. The intensity of the former signal reaches a maximum at an exposure duration of 80 s, whereas the intensity of the latter signal does not saturate upon extended exposure of CH3SH. These S 2p3/2 components are thus assigned to CH3SH molecules chemisorbed and physisorbed on the surface, respectively. XPS data indicate that at initial adsorption CH3SH molecules dissociate mostly to form surface CH3S and hydrogen (CH3SH f CH3S + H). CH3SH can also adsorb molecularly on the surface at large coverage because the activity of a Cu surface for deprotonation of CH3S-H is gradually weakened with increasing coverage of CH3S. For a saturated monolayer, about 20% of adsorbed CH3SH deprotonates to form thiolate at adsorption temperature 100 K. CH3SH molecules can condense on the surface after saturation of adsorption sites but the sticking coefficient is greatly decreased. Parallel to these S 2p spectra, there are three corresponding C chemical states of methyl groups with 1s binding energies at 284.1, 285.5, and 286.3 eV, which are attributed to dissociative CH3S and chemisorbed and physisorbed CH3SH molecules, respectively. The assignment of XPS peaks is consistent with the observation that intensities of S 2p and C 1s signals due to CH3S and chemisorbed CH3SH become attenuated by physisorbed CH3SH molecules on protracted exposure (>80 s). Chemisorbed CH3SH and CH3S are

5440 J. Phys. Chem. B, Vol. 106, No. 21, 2002

Lai et al.

Figure 2. Composite temperature-programmed desorption scans collected from a Cu(110) surface exposed to CH3SH at 100 K for 25 s.

Figure 4. Temperature-programmed desorption scans of CH4 (m/e ) 16) collected from Cu(110) as a function of duration of exposure to CH3SH.

Figure 3. Composite temperature-programmed desorption scans collected from Cu(110) exposed to CH3SH at 100 K for 200 s.

bound to the metallic surface through the sulfur atom.42 From interpretation of ultraviolet photoelectron spectra, it has been proposed that 3d is principally involved in adsorbate bonding.19 Apart from decreased intensity, there is no significant alteration of Cu 3p core-level binding energy after the Cu surface is exposed to CH3SH, which is located in the range of photon energy of the beam used in this work. Figure 2 shows TPD scans taken from a Cu(110) surface exposed to CH3SH at 100 K for 25 s. To detect possible products evolved on thermal decomposition of CH3SH on the surface we followed several fragments. Hydrogen (m/e ) 2) and CH4 (m/e ) 16) are major desorption products after decomposition of CH3SH. Desorption of CH4 exhibits two maximum signals at 430 and 330 K, whereas desorption of only H2 is observed at 310 K. Although CH3 ions (m/e ) 15) are also detected during TPD measurement, their TPD feature resembles that of CH4. A CH4 molecule can become fragmented to produce a CH3 ion on electron bombardment in the ionization zone of the mass spectrometer; for this reason the signal at m/e ) 15 results from fragmentation of CH4, rather than direct desorption of CH3 after dissociation of CH3-S bonds. Figure 3 shows the composite TPD scans for a Cu surface on long exposure (200 s) to CH3SH. Molecular desorption of

CH3SH (m/e ) 48) is observed with an intense and sharp signal at 120 K and a shoulder at 175 K. The intensity of the former desorption signal does not become saturated upon increased exposure; this signal is thus attributed to physisorbed CH3SH in multiple layers. The signal at 175 K attains maximum intensity on exposure duration for ∼80 s and is thus assigned to molecular desorption of chemisorbed CH3SH, but most chemisorbed CH3SH undergoes thermal decomposition. In addition to CH4 and H2, C2H6 (m/e ) 30) shows a maximum signal at 400 K; desorption of C2H6 can occur via coupling of two surface CH3 groups each of which is generated from dissociation of the bond S-CH3.43 Desorption features of CH4 on prolonged exposure differ significantly from those observed for brief exposure. Only a signal due to desorption of hydrogen is observed at ∼315 K for small exposure, as shown in Figure 2; its onset temperature is similar to that observed for hydrogen on a Cu(110) surface.44 Hence, this H2 desorption is attributed to a combinative reaction of surface hydrogen produced by dissociation of CH3S-H. Figure 3 shows that at large exposure of CH3SH an additional hydrogen desorption appears in the lowtemperature region with a maximum signal at 175 K. We deduce that not all hydrogen atoms produced on dissociation of CH3S-H can transfer onto the surface because most adsorption sites are occupied by the thiolate (CH3S) at large exposure. Thus, some released hydrogen is desorbed from the surface in a form of molecular hydrogen at the temperature at which the CH3S-H bond dissociates, implying that chemisorbed CH3SH begins to dissociate to form H and CH3S at ∼175 K. As shown in Figures 2 and 3, desorption products and features vary with CH3SH exposure: the reaction pathway is thus influenced by the coverage of CH3SH on a Cu surface. Figure 4 shows the dependence of CH4 desorption on exposure to CH3SH. At small exposure, two distinct CH4 desorptions have signals at 355 and 415 K; their intensities increase with increased exposure and reach maxima at ∼25 s. With greater exposure, the intensity of the low-temperature signal remains nearly constant, whereas the intensity of the high-temperature signal decreases with increased exposure. Desorption of CH4 at high

Adsorption of Alkanethiols on Cu(110)

Figure 5. (a) Consecutive TPD spectra of CH4 (m/e ) 16) for the Cu(110) surface exposed to CH3SH for 80 s at 100 K in each cycle of adsorption and desorption. (b) Four Gaussian peaks (dashed lines) are used to highlight the desorption features in the first-cycle TPD spectrum of CH4.

temperature occurs in a region in which CH4 is produced from surface CH3 on Cu surface.7,43,45 Desorption of CH4 thus occurs via a dissociative and associative reaction disproportionation of surface CH3 generated after dissociation of CH3-S and transferred onto the surface (CH3 + CHx f CH4 + CHx-1, e3). Careful scrutiny of TPD scans shows that the intensity of H2 desorption rapidly decreases at temperatures exceeding 310 K, at which point desorption of CH4 at low-temperature commences. This coincidence of temperatures indicates that this formation of CH4 is associated with hydrogen on the surface. The signal for desorption at low temperature is thus due to reduction (hydrogenolysis) of CH3 moiety in CH3-S with surface hydrogen (CH3-S + H f CH4 + S). Our XPS data also show that dissociation of the CH3-S bond commences at this temperature, to be discussed later. Desorption of H2 is not discernible from the background level at temperatures above 310 K. Hence H on the surface tends (thermodynamically or kinetically) to react with CH3, rather than to undergo combinative desorption of H2. Additional states of desorption of CH4 appear in a temperature region between the signals at low and high temperatures (355415 K) for exposure exceeding 25 s. A possible reason is that CH3S can occupy various sites of adsorption available on the Cu surface at large exposure when the most stable adsorption sites are mostly occupied. Thus, additional states of CH4 desorption are attributed to originate from CH3S intermediates with adsorption sites or conformations different from that at small exposure. To distinguish the desorption states of CH4, we present TPD scans in a series in Figure 5a, for which the Cu surface is subjected to the cycles of CH3SH exposing at 100 K for 200 s and annealing to 670 K. Except for the first cycle of CH3SH

J. Phys. Chem. B, Vol. 106, No. 21, 2002 5441 exposing and annealing, the Cu surface is treated with no cleaning procedure before the next exposure of CH3SH. This figure shows that desorption at 340 K disappears and the hightemperature desorption is greatly attenuated at the second cycle. As described above, the former desorption is attributed to a contribution of CH3SH that is adsorbed at the most stable site and the latter desorption is due to disproportionation of surface CH3. Only two well-resolved desorption signals are observed at 375 and 415 K in the second cycle; these desorption states might originate from CH3SH adsorbed at sites different from that for desorption at low temperature. After the first cycle of CH3SH exposing and sample annealing, the atomic sulfur atom produced by the thermal decomposition of CH3SH occupies the most stable adsorption site. In subsequent cycles of exposure, CH3SH adsorbs on only empty sites remaining on the surface, which provide smaller energy of adsorption. After five cycles of exposing and annealing, no desorption of CH4 is observed because sites on the surface become saturated with atomic sulfur and some atomic carbon; a CH3SH molecule is consequently unable to adsorb chemically on the surface. The profile of the TPD spectrum is determined by the reaction mechanism and the kinetic process which are complicated by the various effects, such as adsorbate-absorbate and adsorbatesurface interactions, surface diffusion, etc. We make no attempt to numerically deconvolute the TPD spectrum with a function which embodies the detailed information about the reaction mechanism and the kinetic process. However, as shown in Figure 5b, the TPD spectrum is resolved into components of a Gaussian function, simply to highlight four possible desorption states with maximum signals at temperatures of 340, 375, 415, and 450 K (pertaining to desorption states A, B, C, and D, respectively). State D at high temperature is attributed to a dissociative/associative process of desorption of surface CH3. Desorption states A, B, and C are due to hydrogenation of the CH3 moiety of CH3S adsorbed on distinct adsorption sites. Four possible adsorption sites for CH3S on a Cu(110) surface are on top, with short or long bridge, and a hollow site.46 Adsorption with larger coordination numbers generally produces a greater adsorption energy and lower temperature of decomposition.34 The CH3S species adsorbed at various sites decompose to evolve CH4 at distinct temperatures. With increased exposure to CH3SH, desorption states A, B, and C shift to lower temperature, exhibiting a process of second kinetic order, consistent with an assignment that this desorbing CH4 is formed by combination of surface CH3 and hydrogen. Desorption state D moves to slightly higher temperatures with greater exposures. As described above, this desorption is attributed to disproportonation of CH3, for which the rate of reaction is determined by formation of hydrogen via cleavage of the C-H bond. Electronegative sulfur is known to play a role in deactivating metallic surfaces for catalytic reaction.47-50 As the activity of a Cu surface for dissociation of C-H is weakened by surface sulfur, the desorption temperature of CH4 increases with increased CH3SH exposure. C2H6 is a desorption product formed on thermal decomposition of CH3SH. Figure 6 shows desorption of C2H6 as a function of CH3SH exposure. Desorption of C2H6 is observed at exposure duration greater than 25 s, at which states B and C of CH4 desorption begin to appear. As desorption of C2H6 occurs in the range 350-450 K of desorption of B and C, this desorption may originate from the same CH3S intermediates corresponding to B and C states of CH4 desorption. We identified the reaction intermediates during thermal decomposition of CH3SH through measurement of X-ray

5442 J. Phys. Chem. B, Vol. 106, No. 21, 2002

Lai et al.

Figure 6. Temperature-programmed desorption scans of C2H6 (m/e ) 30) collected from Cu(110) as a function of duration of exposure to CH3SH.

Figure 8. XPS spectra of S 2p for the Cu(110) surface exposed to CH3SH at 100 K for 200 s and subsequently annealed to the indicated temperatures.

Figure 7. XPS spectra of S 2p and C 1s for a Cu surface exposed to CH3SH for 25 s at 100 K and subsequently heated to the indicated temperatures.

photoelectron spectra. Figure 7 shows the S 2p and C 1s spectra for the Cu surface exposed to CH3SH at 100 K for 25 s and subsequently heated to indicated temperatures. At 100 K, two S 2p3/2 features are observed at 164.3 and 162.4 eV attributed to chemisorbed CH3SH and CH3S, respectively. The dissociation of chemisorbed CH3S-H to form CH3S begins at ∼150 K and is complete at 270 K. Between 320 and 370 K, the S 2p component due to CH3S gradually decreases and a new S 2p3/2 signal appears at 161.5 eV, attributed to surface sulfur.26 Surface CH3-S species evidently further dissociate to form atomic sulfur and CH3. Both the intensity of the C 1s signal and its binding energy decrease as the surface temperature increases. The decreased C 1s intensity corresponds to desorption of CH4 formed through a reaction between CH3 moiety and surface hydrogen and is maximum at 330 K, as shown in Figure 2. The decrease of C 1s binding energy reflects formation of

surface CH3, which exhibits a smaller C 1s bonding energy than CH3S. A CH3 group generated through CH3-S dissociation can either react with surface hydrogen to desorb CH4 or transfer into an empty adsorption site. Upon annealing the sample to 470 K, the C1s signal decreases in intensity and decreases to 283.5 eV; this signal is attributed to atomic carbon at the surface.46 The surface CH3 eventually undergoes disportionation leading to formation of surface carbon and CH4, as previously reported for trimethylphosphine and methyl iodide on Cu.51,52 This CH4 formation corresponds to the desorption signal at 430 K in TPD scans. Figures 8 shows S 2p and C 1s spectra collected from a Cu surface exposed to CH3SH at 100 K for 200 s, thus becoming saturated with chemisorbed CH3SH and CH3S, and then annealed. The variations of S 2p and C 1s spectral features are similar to those observed at small exposure. Deprotonation of chemisorbed CH3SH to form CH3S is nearly complete upon annealing the sample to 270 K. Between 270 and 470 K, the signal of S 2p3/2 at 162.4 eV due to CH3S gradually disappears due to dissociation CH3-S bonds. At the same time, the C 1s intensity decreases as the temperature increases, consistent with TPD results that show desorption of CH4 in this temperature range. At 470 K, all CH3S species decompose into atomic sulfur on the surface and atomic carbon in small proportion. Assuming that the coverage of a carbon-containing species is proportional to its C 1s intensity, the ratio of the integrated areas of peaks at 150 and 470 K indicates that 10-15% of chemisorbed CH3S decomposes and dehydrogenates to form surface carbon. Other 85-90% chemisorbed CH3S decomposes to form the desorbed CH4 and C2H6 via different reaction pathways (hydrogenation and coupling reaction). On the basis of the TPD results corrected by the relative sensitivity of ionization, the ratio of CH4 to C2H6 products is about 20. Comparison of changes in XPS spectra shown in Figures 7 and 8 provides further details about occupation of sites of

Adsorption of Alkanethiols on Cu(110)

J. Phys. Chem. B, Vol. 106, No. 21, 2002 5443

Figure 9. Temperature-programmed desorption scans collected from Cu(110) exposed to C2H5SH at 100 K for 10 s.

adsorption of CH3S intermediates at small and large coverage. In contrast for small exposure, at large exposure much CH3S is still present on the surface at 370 K. These CH3S species surviving at 370 K differ from that at small exposure, corresponding to states B and C of desorption of CH4, consistent with our argument that CH3SH molecule can adsorb at multiple sites for large exposure. These surface thiolates at distinct adsorption sites are indistinguishable in XPS data, even with a fine instrumental resolution of 0.15 eV at the photon energy (400 eV) used to acquire the S 2p spectra. The fwhm of the S 2p spin-orbit doublet obtained for small exposure is somewhat smaller than that for large exposure. More carbon remains on the surface after decomposition of CH3SH at small exposure, indicating that at small coverage the CH3 moiety generated by CH3-S dissociation mostly transfers onto an empty surface site and eventually undergoes disproportionation. On the basis of TPD and XPS results, the reaction mechanism of CH3SH decomposition on Cu(110) is summarized as follows:

(a) CH3SH(ad) f CH3S(ad) + H(ad) 100-270 K (b) CH3S(ad) + H(ad) f CH4 + S(ad) 320-470 K (c) CH3S(ad) f CH3(ad) + S(ad) 300-470 K (d) CH3(ad) + CH3(ad) f C2H6 350-450 K (e) CH3(ad) + CHx(ad) f CH4 + CHx-1(ad) (x e 3) 380-470 K (f) CHx(ad) f C(ad) + x/2H2(g) (x e 3) 380-470 K Some atomic carbon remaining on the surface after thermal decomposition of CH3SH might result from reactions e and f. Reaction f is dehydrogenation of CHx to evolve H2, which is commonly observed in thermal decomposition of hydrocarbons on metallic surfaces.45 As desorption of H2 is not discernible above the background in the temperature range 380-470 K, dehydrogenation of CHx might proceed primarily via reaction e, but reaction f cannot be excluded. To understand the effect of length of alkyl chain on the reactions of thiols on a Cu surface, we also studied adsorption of C2H5SH and C4H9SH. Figure 9 shows TPD scans for a Cu surface exposed to C2H5SH for 10 s. C2H4 (m/e ) 27) and H2 (m/e ) 2) are products of desorption of C2H5SH decomposition

Figure 10. Temperature-programmed desorption scans collected from Cu(110) exposed to C2H5SH at 100 K for 200 s. Dashed lines serve to highlight desorption features.

with a maximum at 340 K, whereas desorption of C2H6 (m/e ) 30), a hydrogenation product, is not observed. The intensity of a signal at m/e ) 27 is recorded for C2H4 desorption, instead of at m/e ) 28, to avoid background interference from residual gas (e.g., CO and N2) in the chamber. C2H4 is formed via β-hydride elimination of surface C2H5 groups generated on decomposition of C2H5SH (C2H5 f C2H4 + H).43,45 Desorption of hydrogen is due to recombination of surface hydrogen produced on dissociation of the C2H5S-H bond and β-hydride elimination of C2H5 groups. Analogous to the case of CH3SH, features of TPD scans for C2H5SH vary with exposure. Figure 10 shows composite TPD spectra taken for a Cu surface with prolonged duration of exposure to C2H5SH (200 s). A sharp and intense signal (m/e ) 62) due to molecular desorption at 120 K is attributed to C2H5SH in multiple layers. Although a small fraction of chemisorbed C2H5SH is desorbed intact at 170 K, most chemisorbed C2H5SH undergoes thermal decomposition as the sample temperature increases. Besides H2 and C2H4, C2H6 (m/e ) 30) is a desorption product at large exposure to C2H5SH. The desorption features of C2H4 and C2H6 can be deconvoluted into two components with the same signals at 335 and 365 K. These two desorption states might originate from surface C2H5S adsorbed at distinct sites. Hydrogenation to form C2H6 and β-hydride elimination to form C2H4 are competing reactions in thermal decomposition of surface C2H5 generated on cleavage of the C2H5-S bond. The variation in the ratio of C2H4 and C2H6 products indicates that the reaction pathway of C2H5S depends on the site or conformation of adsorption. The adsorption site corresponding to desorption at 365 K favors hydrogenation to form C2H6 rather than that for the high-temperature desorption. The rate of hydrogenation of C2H5 is determined by the availability of surface hydrogen which decreases with increasing surface temperature. Contrary to coupling of CH3, a corresponding formation of C4H10 via coupling of C2H5 is below the detection limit at all exposures. For decomposition of C2H5SH, desorption of hydrogen occurs in two temperature ranges, 120-200 K and 250-400 K. The former desorption originates from direct recombination of hydrogen formed by C2H5S-H scission without surface adsorption, as observed for the case of CH3SH. The latter desorption

5444 J. Phys. Chem. B, Vol. 106, No. 21, 2002

Figure 11. XPS spectra of S 2p and C 1s for a Cu surface exposed to C2H5SH for 80 s at 100 K and subsequently heated to the indicated temperatures.

Figure 12. Temperature-programmed desorption scans collected from Cu(110) exposed to C4H9SH at 100 K for 200 s.

comprises three components with maxima at 315, 335 and 365 K. The former is due to recombinative desorption of surface hydrogen. The latter two desorptions at 335 and 365 K accompany formation of C2H4; their relative intensities and fwhm resemble those for C2H4. These hydrogen desorptions are attributed to originate from β-hydride elimination. Figure 11 shows the S 2p and C 1s XPS spectra of a Cu(110) surface saturated with chemisorbed C 2H5S and C2H5SH (80 s) as a function of surface temperature. The S 2p spectrum at 100 K contains two S 2p3/2 features at 164.2 and 162.4 eV due to adsorbed C2H5SH and C2H5S. C2H5SH to a small proportion (5-10%) decomposes into C2H5S on the Cu surface at the temperature (100 K) of exposure; C2H5SH seems less reactive than CH3SH. Between 100 and 170 K, the intensity of S 2p3/2 at 162.4 eV gradually increases, indicating deprotonation of chemisorbed C2H5SH. Atomic sulfur with S 2p3/2 at 161.5 eV begins to appear at 270 K and C2H5S completely

Lai et al.

Figure 13. XPS spectra of S 2p and C 1s for a Cu surface exposed to C4H9SH for 80 s at 100 K and subsequently heated to the indicated temperatures.

decomposes into atomic sulfur at 470 K. Accompanying the change of S 2p spectra, intensity of C 1s spectra decrease as C2H4 and C2H6 form and desorb in a temperature range 270-470 K, consistent with TPD results shown in Figure 10. Upon annealing to 470 K, atomic carbon is left on the surface with 1s binding energy at 283.6 eV because surface C2H5 undergoes dehydrogenation to a slight extent. Figure 12 shows composite TPD spectra for a Cu surface exposed to C4H9SH for 200 s at 100 K. Hydrogen (m/e ) 2) and C4H8 (m/e ) 56) are desorption products through thermal decomposition of C4H9SH. Desorption signals of hydrogen at 175 and 305 K are due to recombination of surface hydrogen generated on cleavage of C4H9S-H bonds, as observed for both CH3SH and C2H5SH. Desorptions of H2 and C4H8 have a common feature at 365 K, attributed to originate from β-hydride elimination of C4H9 groups produced on decomposition of C4H9SH. Desorption signals for m/e ) 56 at 130 and 175 K reflect fragmentation of desorbed C4H9SH molecules which physisorb and chemisorb on the surface at low temperatures, respectively. Desorption of C4H10 due to hydrogenation of surface C4H9 is not discernible above the background level. Figure 13 shows a general outline of thermal evolution of S 2p and C 1s XPS spectra following saturation exposure of chemisorbed C4H9SH on Cu(110) at 100 K. Only a S 2p3/2 chemical state is observed at 164.3 eV, indicating that C4H9SH molecularly adsorbs on the Cu surface at 100 K. Upon heating the sample to 170 K, the adsorbed C4H9SH begins to deprotonate to form C4H9S with S 2p3/2 binding energy at 162.4 eV. The temperature of decomposition of C4H9SH is greater than those of CH3SH and C2H5SH, consistent with a general observation that RS-H with a longer alkyl chain exhibits weaker acidity, that is, a stronger S-H bond. Between 270 and 470 K, C4H9S further decomposes to produce surface C4H9 and atomic sulfur with S 2p3/2 at 161.5 eV. The C 1s intensity rapidly decreases due to desorption of C4H8 in this temperature region as shown in the TPD result in Figure

Adsorption of Alkanethiols on Cu(110) 12. At 470 K, C4H9SH completely decomposes into surface sulfur. The C 1s peak at 283.6 eV is reflects formation of atomic surface carbon because some surface C4H9 undergoes dehydrogenation. The coupling of alkyl groups is not observable during thermal decomposition of C2H5SH and C4H9SH, in contrast with CH3SH: longer alkyl groups might diffuse along the surface with greater difficulty, or the molecular orientation for coupling might be more restrictive. Hydrogenation of an alkyl group is not favored for a longer carbon chain. CH4 and C2H6 are formed in the decomposition of CH3SH and C2H5SH, respectively, whereas hydrogenation is not observed for C4H9 produced on decomposition of C4H9SH. The long C4H9 group might lie on the surface preferentially to undergo β-hydride elimination to form C4H8, in which Cu can act as a catalyst for scission of the C-H bond.45 We introduced an adlayer of thiolate onto the surface through vapor deposition in UHV conditions, which may differ significantly from a SAM adlayer prepared in solution. Adsorption and thermal reaction of gaseous thiols on a surface can be used to model the initial adsorption and deposition of a SAM adlayer in solution. The adsorption site for alkanethiolate depends on coverage and length of alkyl chain, which subsequently determines the decomposition temperature. Our TPD and XPS data show that alkanethiolate (CH3S, C2H5S, and C4H9S) decomposes to some extent to form atomic surface sulfur at room temperature. From a thermodynamic point of view, a Cu surface, at least Cu(110), is reactive for alkanethiol molecules and might be a poor substrate for formation of densely packed alkanethiolate SAM because surface sulfur is present. At room temperature, even though a thiol with a long alkyl chain likely decomposes into an alkyl chain and surface sulfur, the resulting adlayer is composed of a mixture of alkanethiolate and sulfide on the copper surface, consistent with a structural analysis of saturated alkanetholate on Cu(110).23 Conclusion Our results on adsorption and thermal reaction of alkanethiols provide information about the thermal stability of alkanethiols on the Cu surface. The thiol molecule with a shorter alkyl chain is more reactive and deprotonates on the surface. At 100 K, CH3SH and C2H5SH molecules dissociate to form surface thiolate and hydrogen at small coverage and then chemisorb molecularly with increased exposure. C4H9SH adsorbs molecularly on the surface at 100 K and begins to deprotonate at 170 K. TPD data indicate that CH3SH adsorbs at three sites, but two for C2H5SH and one for C4H9SH. The number of adsorption sites available for thiols thus depends on the length of the alkyl chain. Thiolate species decompose into atomic surface sulfur and alkyl groups by cleavage of C-S bonds. Decomposition of C2H5S and C4H9S occurs at lower temperatures than that for CH3S. A direct interaction between the alkyl group of thiolate and the Cu surface might be involved in cleavage of the C-S bond. The surface alkyl group becomes removed on formation of alkane via hydrogenation or alkene via β-hydride elimination. The β-hydride elimination is favored for the longer alkyl chain. C4H8 is the only desorption product of hydrocarbon in the thermal decomposition of C4H9SH. Hydrogenation products CH4 and C2H6 are observed on decomposition of CH3SH and C2H5SH. An alternative pathway for CH3 group is a coupling that occurs at large coverage, resulting in formation of C2H6. Coupling of alkyl groups is not observed for thermal decomposition of higher alkyl homologues C2H5 and C4H9.

J. Phys. Chem. B, Vol. 106, No. 21, 2002 5445 Using TPD and XPS, we find that alkanethiol can decompose to some extent to form an alkyl chain and atomic sulfur on a copper surface at room temperature. This may indicate that a copper surface is a poor substrate for formation of densely packed alkanethiolate SAM due to the presence of atomic sulfur on the surface. Acknowledgment. This work was supported by SRRC and National Science Council under Grant No. NSC90-2113-M-213014. References and Notes (1) Schuman, S. C.; Shalit, H. Catal. ReV.-Sci. Eng. 1970, 4, 245. (2) Weigand, B. C.; Friend, C. M. Chem. ReV. 1992, 92, 491. (3) Ulman, A. Self-Assembled Monolayers of Thiol Thin Films; Academic Press: San Diego, 1998; Vol. 24. (4) Ulman, A. Chem. ReV. 1996, 96, 1533. (5) Whitesides, G. M.; Laibinis, P. E. Langmuir 1990, 6, 87. (6) Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir-Blodgett to Self-Assembly; Academic Press: New York, 1991. (7) Nuzzo, R. G.; Allara, D. L. J. Am. Chem. Soc. 1983, 105, 4481. (8) Laibinis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y.-T.; Parikh, T. T.; Nuzzo, R. G. J. Am. Chem. Soc. 1991, 113, 7152. (9) Walzcak, M. M.; Chung, C.; Stole, S. M.; Widrig, C. A.; Porter, M. D. J. Am. Chem. Soc. 1991, 113, 2370. (10) Ihs, A.; Liedberg, B. Langmuir 1994, 10, 734. (11) Sellers, H.; Ulman, A.; Shnidman, Y.; Eilers, J. E. J. Am. Chem. Soc. 1993, 115, 9389. (12) Whitesides, G. M.; Laibinis, P. E. Langmuir 1990, 6, 87. (13) Bain, C. D.; Troughton, E. B.; Tao, Y.-T.; Evall, J.; Whitesies, G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321. (14) Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G. J. Chem. Phys. 1993, 98, 678. (15) 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. (16) Laibinis, P. E.; Whitesides, G. M. J. Am. Chem. Soc. 1992, 114, 9022. (17) Ron, H.; Cohen, H.; Matlis, S.; Rappaport, M.; Rubinstein, I. J. Phys. Chem. B 1998, 102, 9861. (18) Sexton, B. A.; Nyberg, G. L. Surf. Sci. 1986, 165, 251. (19) Anderson, S. E.; Nyberg, G. L. J. Electron Spectrosc. Relat. Phenom. 1990, 52, 735. (20) Agron, P. A.; Carlson, T. A.; Dress, W. B.; Nyberg, G. L. J. Electron Spectrosc. Relat. Phenom. 1987, 42, 313. (21) Bao, S.; McConville, C. F.; Woodruff, D. P. Surf. Sci. 1987, 187, 133. (22) Rieley, H.; Kendall, G. K.; Chan, A.; Jones, R. G.; Ludecke, J.; Woodruff, D. P.; Cowie, B. C. C. Surf. Sci. 1997, 392, 143. (23) Vollmer, S.; Witte, G.; Woll, C. Langmuir 2001, 17, 7560. (24) Loepp, G.; Vollmer, S.; Witte, G.; Woll, C. Langmuir 1999, 15, 3767. (25) Dubois, L. H.; Nuzzo, R. G. Annu. ReV. Phys. Chem. 1992, 43, 437. (26) Kariapper, M. S.; Grom, G. F.; Jackson, G. J.; McConville, C. F.; Woodruff, D. P. J. Phys. Condens. Matter 1998, 10, 8661. (27) Rufael, T. S.; Huntley, D. R.; Mullin, D. R.; Gland, J. L. J. Phys. Chem. 1995, 99, 11472. (28) Albert, M. A.; Lu, J. P.; Bernasek, S. L.; Cameron, S. D.; Gland, J. L. Surf. Sci. 1988, 206, 348. (29) Nuzzo, R. G.; Zegarski, B. R.; Dubois, L. H. J. Am. Chem. Soc. 1987, 109, 733. (30) Castro, M. E.; White, J. M. Surf. Sci. 1991, 257, 22.; Castro, M. E.; Ahkter, S.; Golchet, A.; White, J. M. Langmuir 1991, 7, 126. (31) Benziger, J.; Preston, R. E. J. Phys. Chem. 1985, 89, 5002. (32) Koestner, R. J.; Stohr, J.; Glnad, J. L.; Kollin, E. B.; Sette, F. Chem. Phys. Lett. 1985, 120, 285. (33) Huntly, D. R. J. Phys. Chem. 1989, 93, 6156. (34) Mullins, D. R.; Lyman, P. F. J. Phys. Chem. 1993, 97, 12008. (35) Mullins, D. R.; Lyman, P. F. J. Phys. Chem. 1993, 97, 9226. (36) Wiegand, B. C.; Uvdal, P.; Friend, C. M. Surf. Sci. 1992, 279, 105. (37) Jackson, G. J.; Woodruff, D. P.; Jones, R. G.; Singh, N. K.; Chan, A. S. Y.; Cowie, B. C. C.; Formoso, V. Phys. ReV. Lett. 2000, 84, 119. (38) Parker, B.; Gellman, A. Surf. Sci. 1993, 292, 223. (39) Robert, J. T.; Friend, C. M. J. Am. Chem. Soc. 1987, 109, 4423. (40) Friend, C. M.; Roberts, J. T. Acc. Chem. Res. 1988, 21, 394. (41) Meagher, K. K.; Bocarsly A. B.; Bernasek, S. L.; Ramanarayanan, T. A. J. Phys. Chem. B 2000, 104, 3320. (42) Blyholder, G. W.; Cagle, G. W. EnViron. Sci. Technol. 1971, 5, 158.

5446 J. Phys. Chem. B, Vol. 106, No. 21, 2002 (43) Chiang, C. M.; Wenzlaff, H.; Jenks, C. J.; Bent, B. E. J. Vac. Sci. Technol. A 1992, 10, 2185. (44) Baddorf A. P. Ph.D. Thesis, University if Pennsylvania, Department of Physics, 1987. (45) Bent, B. E. Chem. ReV. 1996, 96, 1361. (46) Lai, Y. H.; Yeh, C. T.; Lin, H. J.; Chen, C. T.; Hung, W. H. J. Phys. Chem. B 2002, 106, 1722. (47) Kiskinova, M. P. Surf. Sci. Rep. 1988, 8, 359.

Lai et al. (48) Kiskinova, M.; Goodman, D. W. Surf. Sci. 1981, 108, 64. (49) Hardegree, E. L.; White, J. M. Surf. Sci. 1986, 175, 78. (50) Moon, D. W.; Bernasek, S. L.; Lu, J. P.; Gland, J. L.; Dwyer, D. Y. Surf. Sci. 1987, 184, 90. (51) Chiang, C. M.; Wentzlaff, T. H.; Bent, B. E. J. Phys. Chem. 1992, 96, 1836. (52) Jenks, C. J.; Bent, B. E.; Zaera, F. J. Phys. Chem B 2000, 104, 3017.