Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

52 Langmuir 1985,1, 52-66

Taken together, the data show that closest packed, or- portant limitation to closest-packed, self-assembly on this
iented monolayers of n-alkanoic acids can be formed by substrate which is related directly to the length of the alkyl
adsorption from dilute solution. The formation of these chain. The data further show that oleophobicity is an
quasi-two-dimensional organic surface phases is charac- insufficiently sensitive probe to determine the formation
terized by complicated kinetics in which surface and/or of equilibrium closest packed structures. In a companion
monolayer defects, as well as impurities, seemingly play paper we describe these structures and their relationship
important roles. The structures so obtained are dynamic to many of the above observations in detail.
in nature in that they undergo rapid exchange with com-
parable ligands in solution. These results, and those Registry No. n-Cl9H&O2H, 506-30-9; aluminum oxide,
presented in the following paper, further suggest an im- 1344-28-1; stearic acid, 57-11-4; hydrogen, 1333-74-0.

Spontaneously Organized Molecular Assemblies. 2.


Quantitative Infrared Spectroscopic Determination of
Equilibrium Structures of Solution-Adsorbed n -Alkanoic
Acids on an Oxidized Aluminum Surface
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

David L. Allara*
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

Bell Communications Research, Murray Hill, New Jersey 07974


Ralph G. Nuzzo*
AT&T Bell Laboratories, Murray Hill, New Jersey 07974
Received August 16, 1984

Infrared reflection spectroscopy has been applied to the determination of the structures of adsorbed
monolayer films of n-alkanoic acids on oxidized aluminum substrates. Acids of 16-22 carbons, terminated
by methyl, vinyl, or propargyl groups, were absorbed from hexadecane solution at 25 O C , using immersion
times up to several days. These results,together with those for perdeuterated acids, indicate that close-packed
-
assemblies are formed with extended alkyl tails oriented with their chain axes tilted away from normal
to the surface with a value of 10” for the longer chains. Frequency shifts to higher values are observed
for CH stretching modes of the vinyl and propargyl terminal groups measured in monolayer films compared
to the bulk acids. These shifts are of similar magnitude to gas-liquid phase shifts for these groups and
imply the environment of the ambient-monolayer interface exhibits very diminished intermolecular in-
teractions for these terminal groups compared to the bulk. The spectra clearly show that chemisorption
occurs by proton dissociationto form carboxylatespecies. Examination of both line width and peak positions
of the carboxylate stretching modes suggests a variety of binding geometries of the carboxylate groups
to the surface exists. These results lend support to the general concept of forming stable, oriented and
ordered two-dimensional organic films by solution adsorption and show the utility of applying surface
vibrational spectroscopy to structural determination in these films.

1. Introduction cated molecular organic films. A very elementary problem


The structures, formation, and properties of organic is that most popular surface techniques require impinging
molecular monolayer films are the subjects of much current ions or electrons as probes and this usually causes damage
interest. One system of recent particular interest to us has to organic molecules. The techniques that have been
been spontaneously adsorbed n-alkanoic acid monolayer generally found useful for organic structures are those
films on ambient metal surfaces. In a previous paper’ we using photon probes, in particular, X-ray photoelectron,
presented extensive results on the formation, properties, infrared, and Wan spectroscopies. Additionally, inelastic
and dynamics of solution-adsorbed n-alkanoic acid films electron tunneling and electron energy loss spectroscopies
on oxidized aluminum substrates. Central to a compre- have had success for organic monolayers. Past studies of
hensive understanding of those results is the development the adsorption of medium- to long-chain n-alkanoic acids
of the details of molecular structure. Such detail includes on ambient metal surfaces have utilized electron diffrac-
the nature of the individual molecular species, e.g., ion t i ~ n ,inelastic
~ - ~ electron tunneling (IETS),
formation, surface orientations, and surface-adsorbate
bonding, and the nature of the total monolayer assembly,
(2) A review of earlier work can be found in: Bowden, F. P.; Tabor,
especially as regards ordered packing arrangements. While D. “The Friction and Lubrication of Solids”; Oxford University Press:
a number of surface analysis tools are now available due London, 1968; Part 11, Chapter 19.
to extensive developments in the surface physics commu- (3) Menter, J. W.; Tabor, D. Proc. R . SOC.London, Ser. A 1957,204,
514-524.
nity, in actual fact, few are applicable to the analysis of (4) Chapman, J. A,; Tabor, D. Proc. R. SOC.London, Ser. A 1957,242,
the subtle features of the chemical structures of compli- 96-107.
(5) Brockway, L. 0.; Karle, J. J. Colloid Sci. 1947,2, 277-287.
(6) Cass, D. A.; Straws, H. L.; Hansma, P. K. Science (Washington,
(1) A h a , D. L.; Nuzzo, R. G. Langmuir, preceding paper in this issue. D.C.)1976, 192, 1128-1130.

0743-7463/85/2401-0052$01.50/0 0 1985 American Chemical Society


Spontaneously Organized Molecular Assemblies. 2 Langmuir, Vol. 1, No. I , 1985 53

and infrared spectroscopygJO(IR). The latter two vibra-


tional spectroscopy techniques are quite sensitive, even for
- 2

submonolayer coverages and, in general, are quite inform-


ative with regard to the chemical nature of the adsorbed
species, in contrast to electron diffraction. Recently in-
frared spectroscopyhas been applied to the determination
of chain orientation" and ordering12of n-alkanoic acid salts
deposited by the Langmuir-Blodgett technique. Thus for
our initial studies we have selected IR spectroscopy as a
general probe of adsorbate structures. In our previous
paper,' we determined conditions for the formation of
stable monolayer assemblies of n-alkanoic, alkenoic, and
alkynoic acids on oxidized aluminum and measured such
selected properties as film thicknesses, contact angles, and
tendencies toward molecular exchange in a solution en- Figure 1. Schematic representation of a multiple-phase, par-
vironment. Since it was found that acids with greater than allel-layer sample having a total of m phases. Light approaches
approximately 12 carbons lead to more consistent results, the sample from infinite phase 1at an angle of incidence The
we have chosen to present in this paper only our results directions of the electric field for s (out of plane) and p (in plane)
on the longer chain acids. It is the purpose of this paper polarizations are shown for incoming and outgoing propagation
to apply quantitative IR reflection spectroscopy to the +
directions, designated by and - superscripts, respectively.
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

determination of molecular structure in these films. In


particular, we establish evidence to lend support to a model and equipped with stops and apertures to give an -f/15 beam
of assemblies of C16and longer alkanoic acids bound to the focusing to an -3-mm spot. The collected beam was detected
oxide substrates by multiple alkanoate ion species with by a mercury-cadmium-telluride liquid-nitrogen cooled detector
with a 1-mm2active area and a cutoff at -700 cm-'. Spectra were
strongly tilted and twisted head-group orientations and
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

taken at 2-cm-' resolution with a mirror speed of 1.4 cm/s. The


extended chains oriented about loo away from perpen- interferograms were collected in double precision and Fourier
dicular alignment to the surface. In addition, we show transformed with triangular apodization. Usually about 200-800
evidence for the environmental effects of the air-film in- scans were signal averaged for acceptable signal/noise ratios.
terface relative to the bulk molecular environment in the Reference spectra were obtained on freshly evaporated gold-co-
cases of terminally unsaturated chains. Since our approach vered silicon wafers. These samples were checked carefully for
is to be quantitative, a section of the paper is devoted to hydrocarbon contamination and cleaned further using hot
collecting together and presenting the known principles HZSo4/30%HzOz(5:1, v/v) if necessary to assure that no features
of macroscopic dielectric electromagnetic theory which we would be present in the final ratioed spectra. Dirty reference
samples can cause intensity errors and for sufficiently dirty
utilized for our structural interpretations drawn from the samples can result in "upside down" C-H stretching peaks in the
IR spectra. As far as we are aware, this particular detailed final sample/reference ratio spectra. Such artifacts can be checked
approach has not been used by others and, accordingly, by changing polarization manually or by high-frequency modu-
is presented here. l a t i ~ n . ' ~ , 'Our
~ references were checked by the former method
and no residual C-H features were found when our cleaning
2. Experimental Section procedure was carefully followed. Samples were mounted in a
separately purged compartment isolated from the spectrometer
2.1. Materials and Sample Preparation. Full details of the
system by KBr windows. The nitrogen purge gas was passed
materials used and sample preparations are given in the previous
through molecular sieve and charcoal filters.
paper.' The samples generally consisted of -200-nm films of
Spectra of bulk compounds were obtained using normal in-
evaporated aluminum (99.999%) on single-crystal silicon wafers
cidence transmission spectra of pressed KBr disks prepared in
polished to high optical quality. Adsorption of the alkanoic acids
a dry nitrogen atmosphere.
was done from dilute hexadecane solutions (generally 0.005 M)
*
thermostated at 25.0 0.3 "C. Most of the substrates are pre-
3. Theory of t h e Optical Measurements and
treated with acetic acid solutions, and the adsorbed acetate is
displaced by the chosen alkanoic acid. The IR spectra did not Calculations
appear different for these two methods although some difference
might have been expected because of the observation that not 3.1. Parallel Layer Model. The infrared experiments
all acetate is displaced in the formation of longer chain acid presented in this paper, as well as the ellipisometry mea-
monolayers.' Samples were generally immersed for 2 or 3 days surements in the previous paper,l involve the interaction
before analysis, a point at which the films appeared to exhibit of a light beam with a macroscopically smooth, reflective
no change in structure with time. Storage of the samples under solid material. The theory which relates the macroscopic
ambient conditions for extended periods did not induce changes variables of the experiment to the propagating incident
in the IR spectra. and reflected electric fields has been thoroughly developed
2.2. Infrared Measurements. IR spectra of adsorbates were in terms of boundary value solutions to Maxwell's equa-
taken by reflection of the incident beam at an angle of incidence tions for the cases of parallel layer samples where each
of 87' (3" off glancing) using p-polarized radiation. A nitro-
gen-purged Digilab 15-B Fourier transform spectrometer was used phase can be described by a spatially uniform dielectric
with modified optics to permit focusing outside the instrument f~nction.'"'~ Specific adaptations for reflection spec-
troscopy have been describedl8Jgas well as the importance

( 7 ) Brown, N. M. D.; Floyd, R. B.; Walmsby, D. G. J . Chem. SOC.,


Faraday Trans. 2 1979, 75, 261-270. (13) Dowrey, A.E.;Marcott, C. A. Appl. Spectrosc. 1982,36,414-416.
(8) Hall, J. T.; Hansma, P. K. Surf. Sci. 1978, 76, 61-76. (14) Golden, W. G.; Saperstein, D. D.; Severson, M. W.; Overend, J.
(9) Boerio, F.J.; Chen, S. L. J. Colloid Interface Sci. 1980,73,176-185. J. Phys. Chem. 1984,88, 574-581.
(10)Golden, W.G.; Snyder, C.; Smith, B. J. Phys. Chem. 1982, 86, (15) The general principles can be found in: Born, M.; Wolf, E.
4675-4678. "Principles of Optics", 5th ed.; Pergamon Press: New York, 1965.
(11) Allara, D. L.;Swalen, J. J . Phys. Chem. 1982, 86, 2700-2704. (16) Heavens, 0.S. "Optical Properties of Thin Solid Films"; Dover
(12) Rabolt, J. F.; Burns, F. C.; Schlotter, N. E.; Swalen, J. D. J. Chem. Publications: New York, 1965; Chapter 4.
Phys. 1983, 78, 946-952. Chollet, P. A. Thin Solid Films 1978, 52, (17) For a general treatment, see: Berremen, D. W. J . Opt. SOC.
Am.
343-360. 1972, 62, 502-510.
54 Langmuir, Vol. 1, No. 1, 1985 Allara and Nuzzo

of the application of these principies to quantitative is measured in the reflection experiments and the tran-
analysis of band shapes and intensities.*O A brief de- smissivity T (for unpolarized light) in the KBr pellet
scription will be given here to show the development of transmission measurements. The first quantity is easily
our calculations, particularly for the determination of calculated from the Fresnel coefficient:
adsorbate orientations and corrected spectral shifts from
RIP = rlprlp* (6)
IR data.
Figure 1 depicts the interaction of a parallel mono- where rlp* is the complex conjugate of rlp, which in turn
chromatic light beam of frequency i j (in wavenumbers) with is defined as the ratio of the reflected outgoing and in-
a parallel layer sample consisting of m phases each having coming p-polarized fields for phase 1,
complex optical functions 6, where j referes to the j t h
phase. The quantity tij is defined by eq 1in terms of the
A, = n, + ik, (1) r1p = (2)
real refractive index nj and the absorption constant k .. For
each polarization, the incoming, transmitted and redected Similarly, for normal incidence where Tp = T,,
complex electric fields, El+, E,+, and E<, respectively, can nm
be related to the experimental variables as shown in eq 2 T = - tt* (7)
nl
(Em-is required to be zero since light enters the sample
where t, the complex transmission Fresnel coefficient of
the multilayer sample, is defined as
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

t=@)

from only one direction). The matrix operators R for s and


Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

p polarization are defined as In this paper the experimental reflection spectra are com-
pared to theoretical spectra in order to calculate molecular
orientations. The theoretical spectra are calculated for a
given thickness of a flat, uniform, isotropic layer of bulk
material on the aluminum (oxide)/aluminum substrates
and are computer generated from eq 6 and 2. The values
of the optical constanb of the bulk overlayer material, (C0,
using Figure 1)are determined from transmission spectra
of KBr pellets as follows. As a first approximation for
transmission we can write
T/To = e-4rk'd,V (9)
where a, = fij cos $j and P; = f i j cos $j-l. The R matrix where d, is the pellet thickness and Torefers to the tran-
performs an operation which transforms the phase and smissivity of an identical KBr pellet with no absorber. The
direction of the E fields upon interaction with an interface. absorption index k'is defined as k C / C o where k refers to
The reflectivity of an interface is calculated by this the pure material and C and Coare the molar concentra-
quantity. The operators P are independent of polarization tions of the material in the KBr mixture and the pure
and defined as material phase, respectively. By use of these estimated

r e-ik2jd2j 0
1
values of k ( v ) , corresponding values of n(ij) can be calcu-
lated from the Kramers-Kronig transform:21

where kZjis the wave vector z projection, Brimj cos $j, and where vi refers to the ith point in the frequency set. This
d,, is the distance across phase j along the z axis. This
initial set of 6(e) values is then used in a more rigorous
matrix performs the operation of changing the phase of
calculation (eq 7 and 2) of a theoretical transmission
the E fields generated by propagation of the beam through
spectrum. Differences with the experiment are used to
the matrial phase j . From the relations of eq 2 the required
refine the 6(e) set. This process is iterated until conver-
ratios of incoming and outgoing fields can be calculated
for any reflection experiment. gence of calculated and experimentalspectra occurs. Many
3.2. Applications to Infrared Speutroscopy. For the
other methods of determining the bulk optical constants,
IR experiments the reflectivity RlP for p-polarized light of course, are possible, but we have found this particular
one convenient.
The orientations of the various adsorbate molecular
(18) Greenler, R.G. J. Chem. Phys. 1966,44,31&35. McIntyre, J. D. groups which have IR active modes can be calculated
E. In "OpticalProperties of Solids, New Developments"; Seraphin, B. O., (under specified conditions given below) from reflection
Ed.; American Elsevier: New York, 1976; Chapter 11. McIntyre, J. D.
E. In 'Advances in Electrochemistry and Electrochemical Engineering", spectra because of the anisotropy of the surface electric
Delahay, P., Tobias,C. W., E&.; Wiley New York, 1973; Vol. 9, pp 61-66 field. It follows from the theoretical description given in
and references cited therein. this section that, for the glancing angle reflection of a
(19) Hansen, W. N. In "Progress in Nuclear Energy"; Elion, H. A.,
Steward, D. C., Eds.; Pergamon: New York, 1972; Vol. 11, Chapter 1. infrared beam off most metals, the only effective contri-
Hansen, W. N. In 'Advances in Electrochemistry and Electrochemical
Engineering";Delahay, P., Tobias, C. W., Eds.; Wiley: New York, 1973;
Vol. 9, pp 1-60 and references cited therein. (21) For example, see: Stern, F. In "Solid State Physics, Advances in
(20) Allara, D.L.;Baca, A.; Pyrde, C.A. Macromolecules 1978, 11, Research and Applications";Seitz, F., Turnbull,D., Eds.; Academic Press:
1215-1220. New York, 1963; Vol. 15, pp 299-408.
Spontaneously Organized Molecular Assemblies. 2 Langmuir, Vol. 1, No. 1, 1985 55

Table I. Infrared Spectra and Mode Assignments for CllHSICOzHAdsorbed on Oxidized Aluminum
.
freauencv "
C16H31C0ZH C15H31C02Na C15H31C02H/A1203 mode" ref
2965 -2958 (br) 2966 CH3, asym str ip CCC, r, 24, 25
2955
2935
2918
- 2935
2919
(high-frequency side of 2926 band)
2926
CH,, asym str op CCC, rb-
CH,, sym str,* cryst interactnb
CH2, asym str, d-(?r)
24,
24,
25
25
2871 2874 2880 CH3, sym str, r+ 24, 25
2850 2851 2857 CH2, sym str, d+(r) 24, 25
1737 \

I
1729
1716 C=O str 27, 28
1703
1685
1557 1608 -COP-, asym str 26
1475 -C02-, sym str' this work
1471 1471 } CHz, scissors def 25, 26
1464 1463
1446 -COf, sym str 27
1431 (C-0 str) + (OH de0 28
1423 -CO;, sym str 27
1417 -CO;, sym str' this work
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

(a-CH2,scissors de0 + (other modes?)


- 1412
1150-1300 - 115O-13OO CH2,wag and twist modes
28
29-32
"The following symbols are used str for stretching, def for deformation, ip for in plane, and op for out of plane. The symbols r,, rb-,
d-(?r), and r+ are taken from ref 26. *Sym str (CH,) in Fermi resonance with overtone of def (CH,) and crystalline interactions (see ref 25).
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

'Assuming Cpu symmetry for the assignment. The actual surface species may not have a twofold rotational axis (see text).

bution to surface electric field is normal to the surface18 adsorbed layer undergoes significant changes relative to
( z direction in Figure 1). Since, in general, the intensity the bulk material with regard to force constants, normal
of a given mode is proportional to the square of the scalar modes, and charge distributions. In these cases spectra
product of the E field (eq 2) and the dipole moment de- are expected to differ significantly between bulk and ad-
rivative (or transition dipole) with respect to the normal sorbate with changes in band shapes, vibrational fre-
coordinate, quencies, and intensities. In such cases it is impossible to
make simple interpretations from comparisonsof bulk and
adsorbate spectra. Other less rigorous but useful ap-
proaches for interpreting reflection spectra intensities
specifically in terms of orientations thin film overlayers
it follows that the mode intensity will vary with orientation have been developed recently by Debe.23i24 These ap-
at the surface according to proaches use the basic relationships of eq 11 and 12 to-
gether with approximate relationships between bulk
spectra and surface spectra which circumvent the more
rigorous approach outlined based on eq 1-10. We prefer
where 8 is defined as the angle between M and the surface the present approach since it requires fewer approxima-
normal and M = (dp/dq). The simplest case of an ad- tions and has been shown previously to be quantitatively
sorbed layer is one in which the adsorbate structure is valid. ltZo

homogeneous and isotropic and identical with the structure


of a known bulk material. In this case, for an ideally flat 4. Results
substrate with known optical functions and for a known
overlayer thickness, one, in principle, can calculate the real 4.1. Infrared Measurements and Mode Assign-
infrared spectrum accurately.20 A second more compli- ments. Mode assignments are presented in Tables 1-111.
cated case arises when the adsorbate overlayer is aniso- First we will present results pertaining to the head groups
tropic but otherwise identical with the bulk structure. In followed by the CH groups of the saturated acids and
this caae the theoretical isotropic adsorbate spectrum will finally the CH and terminal groups of the terminally un-
differ from that observed by the intensity of each mode saturated acids.
scaled according to eq 11with the condition that alignment In general agreement with previous IETS studies,68 we
of the transition moment along the z axis gives an intensity observe two major differences between the spectra of bulk
for that mode of 3 times that for the intensity for a random
orientation. This relationship is shown in eq 12,where lobed (22) The major scale of roughness for our samples is on the size of -50
X 200 A rounded low hills as determined from SEM photographs. On
this scale the major point to note is that the measured orientation of the
adsorbate molecules determined by IR is relative to the normal to the
and Iddare the experimentally observed and the isotropic tangent of the surface curvature (the tangential E field will be -0 by
simple electrostatic arguments for good conductors). Other effects have
calculated spectra, respectively, and 0 is as defined above. to do with the absolute values of the intensities of the local E fields in
Smaller differences may arise because of the need to the vicinity of the oscillators. We do not expect these to be significantly
correct for the changes in the surface field due to an an- different from the ideal surface values. Further discussions of these
effects are given elsewhere: Kotler, Z.; Nitzan, A. Surf. Sci. 1983, 130,
isotropic fi (i.e., fi should be formulated as tensor rather 124-154.
than a scalar). However, these changes are sufficiently (23) Debe, M. K. J. Appl. Phys. 1984,55, 3354-3366.
small that they can be neglected. Other errors in inter- (24) Debe, M. K. Appl. Surf. Sci. 1982-1983, 14, 1-40.
(25) Hill, I. R.; Levin, I. W. J. Chem. Phys. 1979, 70,842-851.
pretation can arise because of the effects of surface (26) Snyder, R. G.; Hsu, S. L.; Krimm, S. Spectrochim. Acta, Part A
roughnessz2 A third class of cases are those in which the 1978,34A, 395-406.
56 Langmuir, Vol. 1, No. 1, 1985 Allara and Nuzzo

Table 11. Infrared Spectra and Mode Assignments for H2C=CH(CH2),&02H Adsorbed on Oxidized Aluminum
frequency
H2C=CH-
HZC=CH(CHz) l&OzH (CHZ)leCOzK H,C=CH (CH2)&02H/A1203 modea ref

3078 3078
3086 / =CH2, C-H asym str 27

R,
3001 3000 3000 ,C-H str 27
H
2981 2980 =CH2, C-H sym str 27
2918 2916 2920 -CHz, C-H asym str, d-(n) 25, 26
2850 2851 2851 -CH2, C-H sym str, d+(r) 25, 26
1735 (sh) -C02H, C=O str 27, 28
1705
1644 1644 1644 c=c str 27
-1590 (br) -CO<, asym str 27
1473 1472 -CH2, def 25, 26
1472 -COL, sym str this work
1464 1465 (sh) -CH2, def 25, 26
1437 (C-0, str) + (OH, def) 28
1427
1415
- -COz-, sym str 27
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

1427
1411 (=CH2, C-H def) + (cu-CH2,def) 27, 28
+ (other modes?)
- 1150-1350
995
- 1150-1350
995
-CHz, wag and twist modes
C=C twist, op def
29-32
27
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

912 914 914 =CH2, CH2 op def 27


OFor definition of mode abbreviations see footnote a, Table I.

1470
I

I
N N
0 0
Y ot X
4 KBr a
'557 4703 KBr

1423 1446 1471

Na SALT

. ACID
1000 1200 1400 1600 1800 2000 10 1200 1400 1600 1800 2000
WAVENUMBERS WAVENUMBERS
Figure 2. Adsorbate and KBr-bulk material composite spectra Figure 3. Spectra for n-C1$31C02Hin the mid-frequency region.
for n-C15H3,C02Hin the mid-frequency region. The intensity For details see caption for Figure 2.
units for the adsorbate spectrum are absorbance. Experimental
conditions are given in the text. 1608-cm-' band to the asymmetric carboxylate stretching
mode.27 The latter is known to be quite variable as a
acids (KBr dispersions) and the spectra of the corre-
sponding adsorbates on aluminum oxide as shown in
Figure 2 for the particular case of hexadecanoic acid. Most (27) Bellamy, L. J. "The Infrared Spectra of Complex Molecules";
importantly, the carboxylic acid C=O absorption at 1703 Wiley: New York, 1975; and references cited therin.
(28) Hayashi, S.; Umemura, J. J. Chem. Phys. 1975, 63, 1732-1740.
cm-' is absent in the adsorbate. Second, a new broad peak (29) Holland, R. F.; Nielsen, J. L. J . Mol. Spectrosc. 1962,8,383-405.
at lower frequency, 1608 cm-', has appeared. Other dif- (30) Jones, R. N.; McKay, A. F.; Sinclair, R. G. J. Am. Chem. Sac.
ferences can be noted such as the loss of the low-frequency 1952, 74, 2575-2518.
(31) Meiklejohn, R. A.; Meyer, R. J.; Avonovic, S. M.; Scheutte, H. A.;
series of peaks between 1100 and 1300 cm-' and the Maloch, V. W. Anal. Chem. 1957,29, 329-334.
broadening of the two peaks between 1400 and 1500 cm-'. (32) Snyder, R. G.; Schachtachneider, J. H. Spectrochin. Acta 1963,
These data are quite consistent with the dissociation of 19, 85-116.
(33) Bellamy, L. J. 'The Infrared Spectra of Complex Molecules", 2nd
the acid proton to form an anionic surface carboxylate ed.; Chapman and Hall: New York, 1980; p78.
species if we make the reasonable assignment of the (34) Boobyer, G.J. Spectrochim. Acta, Part A 1967, 23A, 325-333.
Spontaneously Organized Molecular Assemblks. 2 Langmuir, Vol. I , No. 1, 1985 51

Table 111. Infrared Spectra and Mode Assignments for HC=C(CH2),&02H Adsorbed on Oxidized Aluminum
frequency
HC=C(CH2)19C02Na HC=C(CH2)&OzH/AlZ03 mode" ref
3307 (sh) 3325 } =C-H, CH strb 33,34
3290
2916 2920 CH2, asym str, d-(n) 25, 26
2850 2852 CHz, sym str, d+(r) 25, 26
2117 2122 C=C, str 27

1
1615
1579 -COT, asym str 27
1564
1488 -C02-, sym str this work
1471 -CH2, def 25, 26
1435 -CO<, sym str this work
1427
1412 t -coz-l sym str 27

-1100-1400 CH2 wag and twist modes 29-32


=For definition of mode abbreviations see footnote Q Table I. *The splitting in the pure salt band is attributed to Fermi resonance,35
Splitting presumably occurs also in the adsorbate band, but noise limits the resolution of individual features in this asymmetric band (see
Figure 16).
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

function of environment and c ~ u n t e r i o n .In


~ ~the sodium 0.3
salt spectrum in Figure 2, the asymmetric stretch is as-
signed to the peak at 1557 cm-'. The assignment of the
symmetric carboxylate stretch is more difficult since it
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

usually occurs in the same frequency region as the CH2


scissors mode. The latter can be assigned to the doublet
at 1464 (1463) and 1471 cm-' in the acid and sodium salt
spectra. This doublet is due to factor group splitting ar-
ising from interactions between the two types of chains in
the orthorhombic unit cell.36 In order to simplify the
symmetric carboxylate assignment, the spectra of the
deuterated hexadecanoic acid and the corresponding so- N

dium salt were examined. These spectra are shown in 2


x o
Figure 3. The CD2scissor deformation doublet is shifted a
to lower frequencies by 380 cm-' relative to CH2 This shift
allows assignment of the asymmetrically shaped band at
1438 cm-' in the deuterated acid to the symmetric car- 2850 i
boxylate. In the case of the undeuterated salt (Figure 2)
the 1423- and 1446-cm-' absorptions presumably are due
i s
to the symmetric carboxylate mode which suggests mul-
tiple carboxylate species and is consistent with the asym-
metrically shaped absorption band of the asymmetricmode
(1557 cm-l) of the deuterated salt. However, the 1446-cm-'
peak could also contain contributions from the asymmetric
CH3band which usually occurs at 1450 f 10 cm-lS2' With
the above discussion in mind it would appear that in the 27
adsorbate spectrum of Figure 2, the 1475- and 1417-cm-' WAVENUMBERS
peaks should be assigned to the CH2 scissor deformation Figure 4. Spectra for n-CI5H3,CO2Hin the high-frequency region.
and the symmetric carboxylate, respectively. However, For details see caption for Figure 2. The bulk acid/KBr ex-
examination of the deuterated adsorbate spectrum (Figure perimental spectrum is accompanied by a deconvolved spectrum
3) where a strong 1470-cm-' absorption, free of an inter- (dashed curve), which presents enhanced resolution.
fering CH2scissors mode, is evident makes it clear that the
1475-cm-' band is primarily due to the symmetric car- of the CH2 groups along a fully extended trans zigzag
boxylate mode. This assignment is in agreement with that hai in.^^,^ This progression also exists in the acid salt but
of Evans and Weinberg3' and Walmsley et al.3s for ad- with significantly weaker intensities. These features barely
sorbed acetate. Further, the shoulder at 1409 cm-', in the can be seen with the scale given in Figure 2. Within the
deuterated adsorbate spectrum (1417 cm-' for the H ad- noise of our spectra, no evidence exists for absorption in
this spectral region for the adsorbed acids on aluminum
sorbate), may be due to a second carboxylate species.
The lower frequency series of sharp peaks between
1150 and 1350 cm-' in the bulk acid spectrum of Figure
- oxide. However, in view of the variable intensity of these
modes, they may be present in the adsorbate but too weak
2 is assigned to the coupled twisting and wagging modes to be resolved in the spectra. Further, since other ob-
servations (see below) are consistent with the trans zigzag
chain, we presume that this conformation exists for the
(35) Nyquist, R. A.; Potts, W. J. Spectrochim. Acta 1960,16,419-427. adsorbates. It is important to note that we have observed
(36) Snyder, R. G. J. Mol. Spectosc. 1961, 7, 116-144. this band series in the case of evaporated silver sub-
(37) Evans, H. E.; Weinberg, H. W. J. Chem. Phys. 1979, 71, strate~.~~
4789-4798.
(38) Walmsley, D. G.; Nelson, W. J.; Brown, N. M. D.; DeCheveignC,
S.; Gauther, S.; Klein, A.; LCger, A. Spectrochim. Acta, Part A 1981,37A, (39) Snyder, R. G. J. Mol. Spectrosc. 1960,4, 411-434.
1015-1019. (40) Holland, R. F.; Nielsen, J. R. J.Mol. Spectrosc. 1962,9,436-460.
58 Langmuir, Vol. 1, No. 1, 1985 Allara and Nuzzo
0 ; I I I I I

c,~D~,co~H/AI~o~ 2222

I
2195
2098
2075

N
z
x (
a
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

WAVENUMBERS
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

Figure 7. Spectra for n-CH2CH(CH2)19C02Hin the high-fre-


2000 2100 2200 2300
WAVENU MBE R S
quency region. For details see caption for Figure 2.
Figure 5. Spectra for n-C1&C02H in the high-frequencyregion. I(
For details see caption for Figure 2. HC sC(CH2)qgC02H/A1203

0.7: 2122 x1 5 3325


I

1472

1427
A 1644
I 1
2920
2850 1

N
z(
a
N

O (
a 1705 2117
I I 2s

A
912
I
1473

i;, , c
3283

' J
I I I I I I I

I 1 I000
1
1mo
1
1400
WAVENUMBERS
At600
15)L)7' IS00 2000
21 1 2200 2400 2600 2800
WAVENUMBERS
Figure 8. Spectra for n-CHC(CH.J&O2H in the high-frequency
region. For details see caption for Figure 2.
3330 3200 3400

Figure 6. Spectra for ~ L - C H ~ C H ( C H ~ ) ~in~ Cthe


O ~mid-fre-
H
dicular to the surface as will be discussed more fully below.
quency region. For details see caption for Figure 2.
Figure 5 shows the deuterated hexadecanoic acid spectra
The C-H stretching spectra are shown in Figure 4. for the C-D stretching region. Because of the yet unclear
Since the spectra of the bulk acid and the corresponding complications of Fermi resonance, factor group splitting,
sodium salt are so similar, only that of the acid is shown. and other effects in the spectra of deuterated alkanoic acid
The absorbate and bulk spectra are obviously very dif- chain^,^^*^ these spectra were not assigned. It is important
ferent and this can be interpreted in terms of the adsorbate
alkyl chains oriented with their chain axis nearly perpen- (42) Hsi, S. C.; Tullock, A. P.; Mantsch, H. H.; Cameron, D. G. Chem.
P h y ~Lipids
. 1982,31,97-103.
(43) Evans, H. E.; Weinberg, W. H. J. Chem. Phys. 1979, 71,
(41) Allara, D. L.; Nuzzo, R. G., unpublished results. 1537-1542.
Spontaneously Organized Molecular Assemblies. 2 Langmuir, Vol. 1, No. 1, 1985 59

to note that no CH modes were observed in the CD Table IV. Estimaied Directions of Selected Transition
spectra. This excludes the possibility of solvent incorpo- DiDole Moments"
ration in the films. modeb (in order of
Spectra also were obtained for the vinyl-terminated acid decreasing e) direction of M re1 to molec coordsc
CH2=CH(CH2)19C02H and the propargyl-terminated acid =CH, C-H str 11 C-H bond
CH=C(CH2)&O2H. Mode assignments are given in =CHp, C-H asym str I C=C axis, ip HCH
Tables 11 and 111. The spectra of the adsorbates on A1203 =CHR, CZo-H str // C-H bond
and the pure acids are given in Figures 6-8. The 1000- =CH2, C-H sym str 11 C=C bond
2000-cm-' region of the propargyl-terminated acid is not ra- I C-CH:, bond, ip CCC backbone
shown as the features observed are virtually identical with rb- I CCC backbone plane
-CH3, sym str, cryst I/ C-CH:, bond
those of the saturated acids given in Figures 2 and 3. From interactn
the general similarities of the spectra of saturated and I CCC backbone plane
unsaturated samples in the lOoQ-2000-~m-~ region, one can r+ 11 C-CH, bond
conclude that the carboxylate head-group species in all d+(d I chain axis, ip CCC backbone
cases are very similar. However, the lower frequency region C=C str /I C=C bond
of the vinyl-terminated acid does differ in that features C=C str I/ C=C bond
--COT asym str I C-CO; bond, ip OCO
assigned to out-of-plane (op) deformation modes of the --COT sym str 11 C-C02- bond
vinyl CH2group and to the C=C stretch appear (see Table -CHz scissors def I chain axis, ip CCC backbone
11) in both the bulk acid and salt spectra and the adsorbate -CH2 wag and twist
spectra. The higher frequency region of the vinyl-termi- modes
nated acid (Figure 7) shows the expected absorption of the =CH2, C = C op def I C=CH2 plane
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

CH2 stretching modes and also three higher frequency =CH2, CH2 op def I C=CH2 plane
weak peaks assigned to C-H stretches associated with the a Directions are approximated from the normal mode descrip-

vinyl group. Figure 8 shows absorption assigned to the tions. Exact directions could be as much as several degrees off
C& stretch in the 2120-an-' region, the - C H 2stretching those calculated from the normal mode potential energy distribu-
tions and need to be arrived at by charge distribution calculations
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

modes in the 2850-, 2920-cm-' region and the high-fre- which are outside the scope of this study. *More detail is found in
quency propargyl C-H stretch. The latter is clearly split Tables 1-111. Modes are listed in order of decreasing frequency for
into two components in the bulk acid spectrum and shows the adsorbate spectra. 11, parallel; I ,perpendicular; ip, in plane.
asymmetry in the adsorbate spectrum. This behavior is
I I I I
attributed to Fermi resonance effects.35 A further obser- 1551
vation of interest is the striking 42-cm-'-frequency shift I
between the propargyl C-H stretch absorption maxima of
the observed bulk and adsorbate spectra. This shift will
be discussed further in the section detailing the calculated
surface spectra.
As a last point we comment on the long-term stability
of these films under formation conditions. Our data show
that after some minimum formation period, specific to each
sample, of several hours to several days no significant I\
changes in film spectra occur with longer formation times
of the same magnitude. However, a limited number of
samples were further immersed for 4 months and exam-
ined. Their spectra show no significant changes in the C-H
stretching region. However, in the 1000-2000-cm-' region
three changes are observed: (1)the asymmetric carbox-
ylate mode is broadened a bit to the higher frequency side
and intensified by 50-100%, (2) the symmetric carboxylate
modes (-1475 cm-' with the weak 1417-cm-' shoulder)
N

are intensified by 50-loo%, and (3) a new broad peak


appears at -1125 cm-' as a distinctive shoulder on the
sloping tail of the bulk A1-0 stretching peak. These
changes seem to indicate a long-term restructuring of the
head groups with regard to adsorbate-substrate bonding, WAVENUMBERS
but at present we are unable to make specific structural
assignments. Figure 9. Calculated (see text) and observed adsorbate spectra
for n-C1SD31C02H.The intensity units are absorbance. The large
4.2. Calculations of Molecular Orientations and tail a t low frequencies is the edge of the LO A1203phonon ab-
Peak Shifts. In this section we present theoretical spectra sorption. (- - -) Calculated spectrum; (-) observed spectrum.
calculated from isotropic bulk optical constants. The
theoretical basis for these calculations was presented in the relationship between molecular coordinates and surface
an earlier section. The importance of this approach for coordinates are given in the Appendix.
quantitative analysis and correct interpretation of spectral Figure 9 shows results at lower frequencies for deuter-
shifts has been noted.20 The directions assigned to the ated hexadecanoic acid. The interpretation of the deu-
transition moments for the various modes are given in terated adsorbate spectrum is simplified because the CH
Table IV. Calculations of orientation are made using eq modes are moved away to lower frequencies presumably
11. Details on coordinate transforms needed to determine leaving only the carboxylate modes. The calculated
spectrum is based on the optical constants of the bulk
sodium salt. Although one can see some similarity in
(44) Mertens, F. P. Surf. Sci. 1978, 71,161-173.
(45) Evans, H. E.; Bowser, W. M.; Weinberg, W. H. Appl. Surf. Sci. features, it is clear that the spectra differ in peak positions,
1980,5, 258-274. line widths, and shapes. A full interpretation requires
60 Langmuir, Vol. 1, No. 1, 1985 Allara a n d Nuzzo
4 .o 1 I I I

c,~H,,co~ ti / A I ~ O ~
2922
I
I

Figure 10. Diagram of a surface carboxylate unit of idealized


structure defining the angles between the surface normal and the
n
assigned transition dipoles of the symmetric and asymmetric 0
stretching modes, Bs and BA, respectively. I
a
L
- 100
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

I
300 2900 300C
WAVENUMBERS
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

F i g u r e 12. Calculated (see text) and observed adsorbate high-


frequency spectra for n-C,5H31C02H. The intensity units are
U U U
-
absorbance. (- -) Calculated spectrum (x0.2); (-) observed
spectrum.

Table V. Orientation" of C-H Modes for Adsorbed


F i g u r e 11. Representation of proposed structures of the ad- C15H31C02H
sorbate species having fully extended all trans chains. The mo- angle of M from surf
lecular orientations relative to the surface coordinate system can normal
be arrived at by performing the indicated chain tilting in the X,Y
plane ( 10') and rotating around the selected C-C bond axes
N
calcd
(-55' for the vinyl group). from
mode (frequency) low/(3Zcdcd) spectra from model'
considering both orientational and bonding differences ra- (2966) >0.5d <45' 47' (64')
(force constants) between the bulk sodium salt and the rb- (2955)b co.01 285' 82' (82')
CH3, sym str (2935)b C 44' (28')
surface species. If we make the very crude assumption that d-(?r) (2926) 0.023 81' (kl) 81' (81')
the isotropic optical constants for the asymmetric and r* (2880) >0.2d <63' 44' (28')
symmetric modes are not very different for the bulk and d+(r) (2857) 0.036 79' (k1) 81' (81')
surface carboxylates and if we assume CZupoint group Transition moment directions in terms of molecular coordi-
symmetry, then some qualitative conclusions of surface nates are given in Table IV. Mode assignments are given in Table
carboxylate orientation relative to surface coordinates can I. Calculations are made from eq 11 and 12. bFrequency for the
be drawn using the transition moment directions (relative bulk acid. This band is weak and unresolved in the adsorbate
to molecular coordinates) given in Table IV. Since both spectrum. Band insufficiently resolved in calculated and ob-
the asymmetric and symmetric modes are observed, then served spectrum to allow calculation. Calculated peak insuffi-
the CZuaxis can be neither exclusively parallel nor per- ciently resolved to allow quantitative calculation of intensity; only
limit of intensity given. eValues are given for a chain tilt of 12O in
pendicular to the surface. We can estimate from the ratio the 1: direction and a twist of 45' around the chain axis as shown
of the observed and calculated integrated intensities that in Figure 12. Values in parenthesis are for a tilt in the --x direc-
the line connecting the oxygen atoms is tipped away by tion.
c o d 0.181/2 = 65" from the surface normal. Similarly for
the 1470 cm-' observed symmetric mode species absorption the C-C02 bond. Qualitatively then, these data establish
we can estimate the CZuaxis also to be tipped away from a very skewed conformation of carboxylate group on the
the surface normal by about 0.67l/' = 35". Obviously these surface. In addition, of course, the multiple absorption
calculations can be only suggestive since the peak positions attributed to the symmetric carboxylate mode and the
and frequencies are significantly different for the two sets broad absorption of the asymmetric mode are indicative
of spectra. There is no reason, however, to expect vastly of a distribution of carboxylate surface species with various
different dipole derivative values for bulk and adsorbate tipping and twisting geometries and binding states.
carboxylate modes. We estimate that, for dipole deriva- Information on the orientation of the alkyl tail of the
tives within a factor of 2 of each other, the errors in the hexadecanoic acid can be obtained from the high-frequency
above calculations crudely are within a 15" range. The spectra shown in Figure 12. These spectra provide some
relevant coordinate systems are shown in Figure 10 where difficulty in quantitative interpretation because there are
0, and Os are the angles of the asymmetric and symmetric (at least) six different modes (see Figure 4 and Table I)
transition moments, respectively, from the surface normal. contained in the spectra and not all are clearly resolved.
For a chain standing at a 10" tilt to the surface, as shown For those modes where absorptions are not clearly resolved,
in Figure 11, Os is required to be 29" and OA 2 62", with approximate estimates of intensities were made directly
the latter value dependent on the degree of rotation about by eye from the spectral presentations. From the ratios
Spontaneously Organized Molecular Assemblies. 2 Langmuir, Vol. 1, No. I , 1985 61
I
' r2919 Table VI. Orientation" of C-H Modes for Adsorbed
C19H39C02H
angle of M from surf
normal
calcd
from
mode (frequency) Zom/( 3 4 3 spectra from modele
ra- (2966) >0.5d C45' 48' (62')
rb- (295~5)~ C 83' (83')
CH3, sym str (2938) c 43' (29')
c) d-(r) (-2920) -0.017 82' ( f l ) 83' (83')
eX
r+ (2879) 0.12-0.25d 45-60' 43' (29')
-z d+(r)(2851) 0.011 84' ( f l ) 83" (83')
" See footnote a, Table V. Frequency for bulk acid C15H31C02H
given in Table I. This band is weak and unresolved in the adsor-
bate spectrum. Band insufficiently resolved in calculated and/or
observed spectrum to allow calculation. Calculated peak insuffi-
ciently resolved to allow quantitative calculation of intensity; only
upper limit of intensity given. e Values are given for a chain tilt of
10" in the x direction and a twist of 45' around the chain axis as
shown in Figure 12. Values in parenthesis are for the chain tilt in
the -x direction.
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

50 I 1
Ill I I I I I I
0
2700 2800 2900 3000 I

WAVENUMBERS
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

Figure 13. Calculated (see text) and observed adsorbate high-


frequency spectra for n-CI9H&O2H. The intensity units are 4 0-
absorbance. (- - -1 Calculated (X0.25) spectrum; (-) observed
spectrum.

of experimental to theoretical isotropic spectral intensities,


30-
transition moment orientations relative to the surface
coordinates were calculated for each individual mode, using c)

eq 12. These values are given in Table V together with 0


X

the angular variations based on reasonable estimates of a


the maximum combined errors associated with measuring 20-
the intensities. To calculate the structural arrangement
of the adsorbed molecule one needs to choose a molecular
geometry that interrelates the various transition moment
directions in a manner consistent with their orientations
10-
relative to the surface coordinates. For these calculations
we chose a molecular model consisting of a fully extended,
trans zigzag chain conformation since evidence such as
packing densities and film thicknesses point to this
structure as most reasonable. The average orientation of 0.o
the individual oscillator groups then was determined by 2700 2800 2900 3100
WAVENUMBER
arranging the adsorbate molecule in various reasonable
orientations on the surface until an optimal match between Figure 14. Calculated (see text) and observed adsorbate high-
experimentally based and model based moment directions frequency spectra for CHzCH(CHz)19C02H. The intensity units
are absorbance. (-- -) Calculated spectrum (X0.125); (-) observed
occurred. This result is given pictorially in Figure 11 in spectrum.
terms of the spatial operations on a normal standing chain
needed to achieve a reasonable match to the experimen- the chain in the -x direction, as can be seen in Table V.
tally determined transition moment directions and the Of course, as can be seen from the model, the CH2 mode
model coordinates. For hexadecanoic acid, the actual intensities are not a function of the sign of the tilt angles.
values of the chain tilt angles chosen are 12O (note that Spectra and calculations were also carried out for eico-
the values of loo given in Figure 11are for the longer chain sanoic acid (arachidic acid), Cl9H&O2H. The spectra are
molecules). This operation is, of course, equivalent to a shown in Figure 13. The major difference with the hex-
12O tilt and a 45' rotation around the chain axis. The adecanoic acid results is the smaller relative intensity
skewed nature of the carboxylate group also is indicated contribution of the CH2modes to the spectra of the longer
in the figure (see earlier discussion). Table V shows the acid compared to the case of the shorter acid. The cal-
transition moment directions inherent in the model for 1 2 O culations in Table VI show that the CH2 groups are ac-
chain tilts in the x and y directions. The nearly equal cordingly more parallel to the surface by several degres for
values of the d-(r) and d+(s) CH2 stretching mode inten- eicosanoic than for hexadecanoic acid (Table V), and chain
sities require that the chain be tipped nearly equally in tilt angles of loo were used in the model calculations.
both the x and y directions, as shown in Figure 11. The Unfortunately, reliable data for -CH3 group orientations
actual values of 10' fit quite nicely for the longer acids but are not obtainable because of the complexity of the modes
values of about 1 2 O would be more appropriate for the C16 and the lack of resolution of the absorptions.
acid. From the rather uncertain intensity data of the CH3 In Table VI1 and Figures 14 and 15, data are given for
modes there is a weak suggestion of poorer fits for tilting the orientation of adsorbed terminally unsaturated CZ2
62 Langmuir, Vol. 1, No. 1, 1985 Allara and Nuzzo
0751 I I I 1 I

~ C H 2 =CH (CH2 CO2 H / A l 2 O 3


HC = C ( C H 2 1 , 9 CO2H/AI2O3

3325
913

Q b
Q
X
4
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

\
I
n 1 I 1 1 I J
3200 3300 3400
WAVENUMBERS
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

F i g u r e 15. Calculated (see text) and observed absorbate mid- F i g u r e 16. Calculated (see text) and observed adsorbate high-
frequency spectra. The AZO3 phonon has been removed by digital frequency spectra of CHC(CH2)&02H. Intensity units are ab-
subtraction of a reference spectrum to display the absorption sorbance. (- - -) Calculated spectrum; (-) observed spectrum.
between 800 and 1000 cm-'. Intensity units are in absorbance.
(- - -) Calculated spectrum; (-) observed spectrum.
Table VIII. Orientation' of Selected Modes for Adsorbed
HC=(CH2)&02H
Table VII. Orientationa of Selected Modes for Adsorbed
CH24H(CHZ)lsC02H angle of M from surf
normal
~

angle of M trom surt


normal calcd
from
calcd mode (frequency) ZoM/(31~cd) spectra from modele
Lbsdl from from
mode (freauency) (343 spectra modeld propargyl C-H (3325)* 0.35' 54' (f5) 29' (43')
d-(r) (2920) 0.017 83' ( f l ) 83' (83')
vinyl CH2, asym str (3086)' 0.49 46' (f5) 51' (64') d+(r) (2852) 0.016 83' ( f l ) 83' (83')
vinyl C-H str (3000)' C=C str (2122)d 0.80 27' ( f l 0 ) 29' (43')
vinyl CH2, sym str (2980)'
d-(r) (2920) 0.019 82' (fl) 83' (83') 'See footnote a, Table V. *This absorption peak exhibits a 36-
d+(r) (2851) 0.018 82' ( f l ) 83' (83') cm-' shift from the calculated peak. The band shapes are quite
C=C str (1644) 0.29 58' (f3) 50'' (49') asymmetric (from Fermi resonance effects, see Table 111) and sig-
C=C twist, op def (995)c nificantly differ (see Figure 18). CBecauseof the asymmetry of the
vinyl CH2 op def (914) 0.15 67' ( i 2 ) 63'' (52') bandsb intensity
A,.... .
ratios
,.
were calculated. . .. - " absorD-
.. from integrated
.". - -1,. .
tions. u'i'nisaDSOrptiOn peak exnims a a-cm-' snirt rrom tne cai-
.
See footnote a, Table V. *The calculated absorption maximum culated peak. 'Values are given for a chain tilt of 10' in the I
occurs at 3078 cm-', an 8-cm-I shift, and the line width of the cal- direction and a 45' rotation around the chain axis as shown in
culated peak is ca. twice that of the observed line width. The cal- Figure 12. Values in parenthesis are for a 10' tilt in the -z direc-
culation of the relative intensities accordingly was done on an area tion.
basis. 'This mode is not resolved in the adsorbate spectrum. The
frenlrenrv uiven i s for the hiilk ncirl Inr anltl anertrnm Orientatinn
calculations for this mode were not made. a Values are given for a frequencies and a narrower line width than t h e calculated
chain tilt of 10' in the z direction and a chain twist of 45' as isotropic spectra. However, orientations of the vinyl group
shown in Figure 12. Values in parenthesis represent the best fit to calculated from the asymmetric CH2 stretch are i n rea-
the data for a -10' chain tilt in the z direction. The corresponding sonably good agreement with the results from the lower
angle of the rotation of the vinyl group is -55'. Other rotation frequency C=C stretch and the one out-of-plane defor-
angles give even poorer fits. e For a 68' rotation of the vinyl group m a t i o n m o d e which is well resolved i n t h e adsorbate
(rather than 55') the experimental value of 58" is obtained for the
angle of M in the model geometry. 'For a 40' rotation of the vinyl spectrum (see Figure 15). The rotation of t h e vinyl group
group (rather than 55') the experimental value of 67' is obtained around the adjacent C-C bond represents an extra degree
for the angle of M. of freedom i n t h i s particular molecule compared to the
others in this study. If one fixes the chain tilting at +loo
acid. Figure 16 shows the C-H stretching modes. Calcu- i n the x,y directions, based on the CH2mode fits, then an
lations d e r i v e d f r o m the a l k y l C-H modes a p p e a r optimal rotation angle of the vinyl g r o u p can be chosen
straightforward, supporting the same chain tilt angles of according to matches of calculated and experimental vinyl
loo as determined for the saturated acids (see Figure 11). m o d e intensities. T h i s procedure leads to a consistent
Interpretation of the vinyl C-H stretching modes, however, picture of a vinyl group "flopped" over around the adjacent
has limitations. The symmetric CH2 and the C21 C-H C-C bond b y about 5 5 O (in the rotation direction shown
stretching modes are not resolved i n the experimental i n Figure 12). Our data do not fit well at rotation angles
spectra. The asymmetric vinyl CH2stretching mode, while greater than f 1 5 O from 5 5 O so t h i s is the range assigned
resolved, exhibits an interesting shift of 8 cm-' to higher to the accuracy of this calculation. Rotation angles for a
Langmuir, Vol. 1, No. 1, 1985 63
4.0
H C r C (CH2119C02H/A12 03

2919
I
21 2 2

0
P
X
4
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

0
2000 2too 2
2800 2900 3000 300
WAVENUMBERS
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

WAVE NUM B E RS
Figure 18. Calculated (see text) and observed adsorbate high-
Figure 17. Calculated (see text) and observed adsorbate high- frequency spectra of CHC(CH2)19C02H. Intensity units are ab-
frequency spectra of CHC(CH,),gC02H.Intensity units are ab- sorbance. (- - -) Calculated spectrum; (-) observed spectrum.
sorbance. (- - -) Calculated spectrum (X0.067); (-) observed
spectrum. dissociation of the carboxylic acid group to form carbox-
ylate species, in agreement with extensive IETS studies
loo chain tilt in the -x direction (see Figure 12) give (e.g., see ref 8). Other work suggests this type of binding
considerably poorer fits than those for the proposed model. should be quite strong as evidenced by the thermal sta-
In Table VI11 and Figures 16-18, results are given for
bility of adsorbed acetate at temperatures up to 300 0C?7746
the propargyl-terminated CZ2 acid. Again the C-H
As in all previous studies of chemisorption of carboxylic
stretching modes (Figure 17) form a consistent picture of
acids on oxidized aluminum, of which we are aware, there
a loo tilted chain rotated 4 5 O around the chain axis. The
is no evidence for the specific disposition of the dissociated
C=C stretching mode intensity (Figure 18) shows good proton. Most likely the proton would be located at some
agreement with the model shown in Figure 11 in which the
basic lattice oxygen site with the carboxylate counterion
chain is tilted in the +x direction. However, the propargyl located close by in order to minimize charge pair separa-
C-H stretch intensity (Figure 16) suggests tilting in the tion. The native oxide overlayer in evaporated aluminum
-x direction. The latter mode appears very unreliable for
samples is -20-25 A in thickness44and is usually pre-
orientation calculations, however, since there is a sub- sumed to be an amorphous phase that is pinhole free as
stantial shift of 36 cm-l to higher frequency in the observed evidenced by its well-known ability to act as an electron-
spectrum relative to the calculated spectrum. This large
tunneling barrier in junction structures. There has been
shift suggests significant perturbations of the mode. In discussion in the literature as to the chemical structure of
addition, the observed splitting of this absorption into two the oxide layer and it appears that formation by thermal
(or more) peaks with different relative intensity distribu-
oxidation in the presence of traces of water vapor produces
tions reduces the reliability of quantitative interpretations a bulk A1203 phase with a highly hydroxylated surface.45*46
from this mode. Therefore, the + x chain tilt direction is Carboxylate formation could occur by reaction of the acid
assigned as correct. at surface hydroxyls with release of water and/or reaction
5. Discussion at Al-0 lattice bonds with formation of a surface hydroxyl.
Evidence exists that the IETS intensity of the surface
It is clear from the present IR results and the associated
study1 that solution-adsorbed alkanoic acid layers on ox- hydroxyl stretching mode can be increased47or decreased@
idized aluminum can be stable, closely packed, highly by acetic acid adsorption on plasma-formed oxide so it is
oriented monolayer assemblies bound by chemisorptive not clear in general as to what proportion of acid molecules
bonding. In addition, these monolayers may be ordered, will react at particular types of sites. An estimate of 6.5
X 1014adsorption sites/cm2 has been made for a smooth,
as previous diffraction studies would suggest2+ but clearly
uniform A1203 surface on the basis of a calculation from
the present data do not allow definitive statements to be
made with regard to this aspect of structure. In the fol- known ionic radii.47 From data reported in a companion
lowing discussion we will present our interpretations and study1 we calculate (using a film density of 0.93 g/cm3
conclusions from the IR experiments. First we discuss the taken from the bulk material value and using a thickness
head-group binding, then the alkyl tail structures, and of 27 A) that a C, acid monolayer should have -4.8 X 1014
finally the ambient-film interface structure. molecules/cm2. From this apparent mismatch of adsorp-
The important points to note about the head-group
structures are that the important species are carboxylates, (46) Bowser, W. M.; Weinberg, W. H. Surf. Sci. 1977, 64, 377-392.
these groups are tipped and twisted with respect to the (47) Lewis, B. F.;Mosesman, M.; Weinberg, W. H. Surf. Sci. 1974,41,
142-164.
surface, and that a distribution of orientations and bonding (48) Brown, N. M. D.; Floyd, R. B., Walmsley, D. G . J. Chem. Soc.,
structures exist. The adsorption process involves proton Faraday Tram 2, 1979, 75, 17-31.
64 Langmuir, Vol. 1, No. 1, 1985 Allara and Nuzzo
tion site density and film spacing together with the strong, the case of bulk aluminum acetate (see above), it would
localized adsorbate binding it would appear that the car- appear that the long-chain carboxylate adsorbates main-
boxylate head groups cannot pack in an ordered fashion tain a fair degree of Czusymmetry in the head groups.
with the arrangement preferred for a two-dimensional It should be pointed out that further structural changes
alkanoate crystal. In actual fact, the situation is probably of the head groups and possibly the oxide lattice may occur
worse since an amorphous oxide surface would not present with long formation times as suggested by the spectral
an exactly ordered atomic arrangement because of the changes exhibited by the series of samples immersed for
presence of oxide defects. Existence of the latter is based
on a parallel study’ involving acetic acid and trifluoroacetic
acid adsorption. In this study radioisotopic labeling and
-
4 months. The primary feature is the appearance of the
broad 1125-cm-’ absorption, which unfortunately is
unassigned at present.
X-ray photoelectron spectroscopy suggest that small car- In contrast to the structural complexities of the head
boxylic acids, such as acetic, can penetrate the oxide lattice group, the alkyl chains appear to form uniform, oriented
to form coverages in excess of the geometric monolayer assemblies. Whereas previous workers have proposed in-
coverage and that these “interior” adsorbed carboxylates corporation of hexadecane solvent molecules into long-
are in some manner degraded, perhaps by diffusion to and chain alkanoic acid monolayers on various substrates,*v4
reaction with aluminum metal at the oxide interface. it is clear from the deuterium labeling studies that our
Thus, although these oxide films are pinhole free in terms
monolayer films consist only of alkanonic acid chains
of electron-tunneling junction structures, they may exhibit
within experimental error (- <5% solvent incorporation).
very small-scale surface defect structures which provide
Our calculations from experimental and theoretical band
channels to the interior structure for very small molecules.
intensities of adsorbate chain group modes are entirely
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

These defect structures can contribute further to the


difficulty of providing an oxide surface which can induce
head-group ordering in the long-chain acid structures. An
-
consistent with (but do not directly prove) a fully extended
chain tilted from the surface normal by 10’ with a 45O
twist around the chain axis as depicted in Figure 12. There
interesting contrast to the aluminum-aluminum oxide
substrate system may be silver which exhibits IR spectra appears to be increased tendency for longer alkyl chains
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

of n-alkanoic acid adsorbates having sharper features in to orient more normal to the surface compared to the
the carboxylate stretch region which we interpret in terms shorter chains. Comparing the results for the c16 and Czo
of very uniformly oriented head-group structures with little acids given in Tables V and VI and Figures 12 and 13,
skewing.41 It is likely in this latter case that the head respectively, shows the CH, d-(n) and d+(a) modes tilt
groups are not tightly bound to the substrate and rear- -2-5’ closer to parallel with the surface in the long-chain
rangement is facile. This would be in agreement with acid, which we interpret straightforwardly as an increase
earlier electron diffraction data of long-chain acids on silver in perpendicular chain-surface orientation. Thus it ap-
which has been interpreted in terms of ordered structure^.^ pears that the extra four CH2 groups contribute to the
From IETS s t ~ d i e s ~ofp small
~ ~ * acids,
~ ~ such as acetic, ability to sustain the perpendicular configuration. The Cm
it has been proposed that the major adsorbed carboxylate acid orientation is in good agreement with the two ter-
species on these types of surfaces are symmetrically minally unsaturated CZzchain acids as seen by comparing
structured (meaningmostly C,,, symmetry),bridge-bonded results in Tables VI-VIII. Although this general picture
carboxylate ions, attached directly at two AI sites in the seems entirely reasonable there are two cautions. First,
oxide lattice surface. The primary evidence for this type these are average orientations. The chain group mode
of structure has been the similarity between the IETS peak frequencies should all be generally unaffected by small
positions for the symmetric and asymmetric adsorbed differences in interchain interactions and changes in ori-
carboxylate stretching frequencies and the reported IR entation relative to the surface. Therefore line shapes are
frequencies for aluminum acetate, which is presumed to independent of these factors, and the data give us no in-
have the symmetrical, bridge-bonded structure.49 The formation on the distributions of orientation. However,
latter is reported49to exhibit the symmetric and asym- at these small tilt angles the most severe but uniform
metric carboxylate frequencies at 1465 and 1590 cm-l, distribution would have to range roughly between the
respectively, while the range of values reported for the limits of 0’ and 20° tilt. Second, it is a consideration that
former range within f 5 and *15 cm-l of these values, the data could be rationalized by invoking some fortuitous
respectively, provided that correct mode assignment^^'^^^ combination of gauche-CH, group interactions along oc-
are applied. The values in our studies (Tables 1-111) fall casional twisted chains but with a very high proportion of
close to those above and accordingly suggest that the chains having fully extended trans zigzag conformation and
distribution of carboxylate species we observe have oriented exactly normal to the surface. It is certainly
structures predominantly of Cz, symmetry and are bridge reasonable, withint the limits of the present IR data and
bonded. However, if bridged bonding to the substrate does the coverage data in the companion study,’ that such
occur then it must be asymmetric in order to accommodate factors as local surface oxide irregularities and incomplete
the tipping and twisting of the head groups required both coverage could lead to “grain boundaries” and small
by the necessity to align the hydrocarbon tails nearly patches comprising several percent of the monolayer film.
normal to the surface and by our interpretation of the Such structural aberrations could certainly contribute to
carboxylate stretching mode intensities. This type of some small proportion of chain structures having a dis-
asymmetric bonding, viz., with two different carboxyl ox- tribution of orientations and gauche conformations as
ygen-surface aluminum distances per carboxylate group, delineated above. In fact, it may be that such defects are
is not implicit in the models others have p r e ~ e n t e d . ~ , ~a~necessary
~~~ part of the mechanism for fast molecular ex-
With this asymmetry in mind it might be expected that change in solution between adsorbate and solute as dis-
the C02- group would not be of CZusymmetry. However, cussed e1sewhere.l Finally, there is suggestive evidence
since our carboxylate stretching frequencies are so similar that the shorter cl6 chains, which seem more tilted, may
to those cases where C,, symmetry is proposed, including exhibit some tendency toward conformational disordering
as judged from the small upward shift of the frequencies
(49) Alcock, N. W.; Tracy, V. M.; Waddington, T. C. J . Chem. SOC., of the d-(n) and d+(n)modes in going from the bulk to
Dalton Trans. 1976, 76, 2243-2246. adsorbate phase. This can be seen in Figure 12 where the
Spontaneously Organized Molecular Assemblies. 2 Langmuir, Vol. 1, No. 1, 1985 65

adsorbate frequencies of 2857 and 2926 cm-l are 6 and 4 again appears to optimize hydrogen atom density at the
cm-' larger, respectively, than the calculated "bulklike" ambient interface, as in the above CH, and propargyl cases.
values. For totally conformationally disordered poly- Reminiscent of the propargyl case also is the 8-cm-l shift
methylene (liquid phase) the frequency values of these of the vinyl CH2 (asymmetric) stretching mode frequency
modes are expected to be 6-8 cm-' larger than all-trans to a higher value for the adsorbate compared to the cal-
chains (crystalline phase).50 It should be noted that the culated bulk derived value (Figure 14). This shift is similar
spectra of the longer chains do not show these shifts in magnitude and direction to that observed in gas-con-
(Figures 13, 14, and 17). The small shifts we observe, densed phase comparisons of 01efins.~' The line width of
however, must be viewed with caution as the noise inherent the adsorbate peak (Figure 14) is also about half that of
in the spectra and the ability to clearly resolve modes put the calculated peak and may imply a decrease in local
limits of several reciprocal centimeters on accurate values. interactions for the adsorbate compared to the bulk. The
Perhaps, one of the most interesting features of these other vinyl C-H stretching modes unfortunately are too
films is the ambient-film interface. It is the structure of poorly resolved to allow characterization. The vinyl CH2
this region that makes the major contributions to such out-of-plane deformation frequency, however, is resolved
important material properties as wettability and lubricity. but shows only a negligible 1-cm-l shift.
The results for the CH3-terminated (n-alkanoic acids) The present results, considered together with those of
chain acids are not as quantitative as potentially possible the companion study, show that, under appropriate con-
because of the inherent difficulties in resulving the complex ditions, long-chain alkanoic acids will spontaneouslyadsorb
CH3 stretching modes (see Table I and Figures 4,12, and at oxidized aluminum surfaces to form close-packed,
13). However, the data appear to fit reasonably the model monolayer films with highly oriented alkyl tails but with
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

in which the CH3group has ita C-CH3 rotation axis tipped a broad distribution of carboxylate head-group orientations
closer to parallel to the surface when the alkyl chain axes and surface binding states. The results further show that
are tipped -10" off normal (see Figure 11). This geometry the environment of the chain terminal groups at the am-
increases the contribution of H atom exposure relative to bient interface is much more similar to the gas phase than
C atom exposure at the interface. the bulk material and that the orientation of the terminal
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

Similarly, we interpret our data for the propargyl-ter- groups may be determined, in part, by a driving force to
minated acid in terms of a loo chain axis tip with the maximize surface hydrogen atom density. A number of
propargyl group aligning its axis more closely parallel to details of the monolayer structures and formation prop-
the surface normal, as shown in Figure 11. In both the erties remain unanswered. The existence of structural
methyl and propargyl cases what apparently happens is features such as chain ordering and thermal phase tran-
the presentation of the maximum hydrogen atom density sitions as well as chemical features such as the effects of
at the ambient-film interface as allowed by the other solvent and substrate variation on film formation needs
constraints determining the film structure. Clearly, the to be examined. Appropriate experiments are in progress
environment of the terminal propargyl group at this am- to investigate these areas involving substrate metals such
bient interface must be significantly different than that as silver and platinum and the application of additional
of crystallites of bulk alkali metal salts as evidenced by structural probes such as surface Raman spectroscopyand
the large changes in the shape and position of the par- electron diffraction.
pargyl C-H stretch mode absorption (see Table 111 and
Figure 16). This mode is reported to be sensitive to hy- Acknowledgment. We are grateful to Professor George
drogen bonding and solvent effect^.^^-^^ It seems reason- Whitesides for stimulating discussions and to Yu Tai Tao
able that the adsorbate propargyl C-H bond, pointing for his assistance in supplying a portion of the long-chain
nearly straight out and exposed at the ambient-film in- acids.
terface, should encounter significantly less polar interac- Appendix
tions than the propargyl C-H bond of bulk material does.
This type of environment change is similar to that of gas Relationship between Molecular and Surface Co-
to condensed phase and such a change is normally ac- ordinates. The relationship between a given molecular
companied by large C-H stretching mode peak shifts to group coordinate system (U') of an arbitrarily oriented
lower f r e q ~ e n c i e of
s ~the
~ magnitude that is observed in adsorbate molecule and the surface coordinate system (U)
the present case (36 cm-'). For example, this shift is re- can be related simply by linear transformation matrices:52
ported to be 25 cm-' for n-pentyne gas and The U' = vu
changes in the asymmetrically shaped doublet structure
of the spectra (Figure 16) are much more difficult to in- where V is a matrix operator consisting of the relevant
terpret and complicate the actual measurements of peak direction cosines. The general form of V is given in terms
shifts above. The band splitting has been attributed to of the matrix elements
Fermi resonance interaction%but the mechanism has not uXI = cos 4 cos 0 cos a - sin 0 sin a!
been resolved unambiguously and we offer no explanations uxty= cos 4 sin 0 cos a! + cos 0 sin a
of the line-shape changes at present.
The spectra of the vinyl-terminated acid are interpreted uxjz = -sin 4 cos a!
in terms of a 10" chain axis tilt, but the disposition of the
vinyl orientation is complicated by the additional degree uylx = -cos 4 cos 0 sin a - sin 0 cos a
of freedom offered by rotation of the vinyl group about
the connecting C-C chain bond. The best fit to our data
~ 4 sin 0 sin
u ~ =, -cos a! + cos 0 cos a!

using our model involves chain axis tipping in which the uyrr = sin 4 sin a
vinyl group C=C axis is pushed away from parallel to the
surface normal and rotated -55" (direction as shown in
Figure 12) about the C-C connecting bond. This operation (51) For example, we observe the vinyl CH2 asymmetric Stretching
mode for 1-octene to be 3088 cm-* in the gas phase and 3079 cm'' in the
pure liquid phase.
(50) Snyder, R. G.; Strauss, H. L.; Elliger, C. A. J. Phys. Chem. 1982, (52) For example, see: Wilson, E. B.; Decius, J. C.; Cross, P. C.
86, 5145-5150. "Molecular Vibrations";McGraw Hill: New York, 1950; p 285.
66 Langmuir 1985, 1 , 66-71
u,., = sin 4 cos 0 The direction cosine terms vxtz,vg and vztzindicate the
U,Iy = sin 4 sin B orientation of the molecular coor mates relative to the
surface normal. To relate a series of n sequential trans-
cos 4
u,., = formations of molecular reference frames Vi to the surface
where 4 is the angle between z and z', 8 is the angle be- coordinates the overall matrix V is just given by the
tween the projection of z'in the x y plane, a is the angle product of the individual transforms, V = II;=' Vi.
of rotation of the x 'y' plane around z' measured by the Registry No. A1203, 1344-28-1; n-CI6H3,CO2H,57-10-3;
angle between the intersection of the x y' ' and x y planes CH+2H(CHZ)i,COzH, 53821-23-1; CHzC(CHz)19C02HI
and y', and z is the surface normal direction. 93645-43-3;n-C,$H&O,H, 506-30-9.

Ordered Ionic Layers Formed on Pt( 111) from Aqueous


Solutions
Downloaded by UNIV DE BUENOS AIRES UBA on August 25, 2009 | http://pubs.acs.org

John L. Stickney, Stephen D. Rosasco, Ghaleb N. Salaita,t and


Arthur T. Hubbard*
Department of Chemistry, University of California, Santa Barbara, California 93106
Publication Date: January 1, 1985 | doi: 10.1021/la00061a008

Received May 1, 1984

Ordered layers (adlattices) formed spontaneously when the Pt(ll1) surface was immersed into aqueous
ionic solutions. On the basis of Auger spectroscopy and LEED, the following salts formed adlattices as
indicated: KCN, Pt(ll1) (2d3x2d3)R3O0-KCN; KSCN, Pt(ll1) (2X2)-KSCN; K2S, Pt(ll1) (diffuse
lxl)-K,S; KI, Pt(ll1) (3X3)-I; KBr, Pt(ll1) (3X3)-KBr. KI solution yielded a layer of neutral I atoms
requiring no cationic counterion, in agreement with previous studies of HI vapor and aqueous electrosorption.
All ionic concentrations studied (lo4 to lo-' M) gave similar results. The adsorbed layer of anions functioned
as a cation exchanger: K+ ions were quantitatively replaced when the surface was rinsed with M HCl
or CaC1,. Exchange of cations did not change the LEED pattern at room temperature; however, recon-
struction occurred on heating to about 100 "C in some cases. Auger spectra indicated that hydro-
quinonesulfonate (KHQS) displayed a packing density transition as a function of concentration, as expected
from electrochemicaldata; LEED patterns displayed no fractional-index beams due to the KHQS layer.

Introduction solutions, including constituents commonly found in


While the vital role played by the electrical double layer electroplating baths16 and other practical electrolytes.
a t electrode-solution interfaces' has been recognized for
many years,2 experimental difficulties have precluded
direct structural characterization of the electrode surface. (1)(a) Delahay, P. "Double Layer and Electrode Kinetics"; Intersci-
ence: New York, 1965. (b) Overbeeck,J. T. G. Pure Appl. Chem. 1965,
In particular, the structural arrangement of electrode 10,359. (c) Spamaay, M.J. "The International Encyclopedia of Physical
surface atoms and electrolytic ions is considered to be Chemisty and Chemical Physics"; Everett, D. H., Ed.; Pergamon Press:
important on practical3 and theoretical ground^.^ Work New York, 1972;Vol. 4,Topic 14. (d) Hurwitz, N. D. "Electrosorption";
Gileadi, E., Ed.; Plenum Press: New York, 1967.
leading up to the present study has dealt with hydrogen (2)(a) von Helmholtz, H. L. F. Ann. Phys. (Leipzig) 1853,89(2),211.
electrodep~sition,~ halogen chemisorption,6 solvent vapor von Helmholtz, H. L. F. Ann. Phys.(Leipzig) 1879,7 (3),337. (b) Gouy,
chemi~orption,~ competitive adsorption of solvent and G. J. Phys.Radium 1910,9(4),457. Gouy, G. C. R. Hebd. Seances Acad.
Sci. 1910,149,654. (c) Chapman, D. L. Philos.Mag. 1913,25(6), 475.
electrolyte vapor* on Pt(100) and P t ( l l l ) , oriented ad- . , Stern. 0.Z.Elektrochem. 1924. 30. 508.
(c)
sorbed molecule^,^ oxidationlo and reduction of oriented (3)Hubbard, A. T. Acc. Chem. Res.' 1980,13,1977.Hubbard, A. T.
intermediates,l' the influence of microscopic surface J. Vac. Sci. Technol. 1980,17,49.
roughness on adsorbate orientation and reactivity,12elec- (4)Barlow, C. A.,Jr.; MacDonald,J. R. J. Chem. Phys.1964,40,1535.
(5)Ishikawa, R. M.; Katekaru, J.; Hubbard, A. T. J . Electroanab
trodeposition of Cu13and AgI4 onto well-defined surfaces Chem. 1978,86,27:.
of Pt containing adlattices of iodine atoms, and with (6)(a) Felter, T.E.; Hubbard, A. T. J. Electroanal. Chem. 1979,100,
preparation and identification of well-defined Pt surfaces 473. (b) Garwood, G. A., Jr.; Hubbard, A. T. Surf. Sci. 1980,92,617.(c)
Garwood. G. A., Jr.: Hubbard. A. T. Surf. Sci. 1982,112,281.
a t atmospheric pressure.15 Results of previous work5-15 (7)Garwood, G.'A., Jr.; Hubbard, A,' T. Surf. sci. 1982,118, 223.
illustrate the remarkable degree to which surface structure (8)Katekaru, J. Y.; Hershberger, J.; Garwood, G. A., Jr.; Hubbard, A.
and molecular orientation influence electrochemical and T.Surf. Sci. 1982,121,396.
(9)Soriaga, M.P.; Hubbard, A. T. J. Am. Chem. SOC.1982,104,3937.
catalytic reactivity although much remains to be explored. (10)(a) Soriaga, M.P.; Stickney, J. L.; Hubbard, A. T. J. Electroanal.
Since electrodeposition processes have been found to be Chem. 1983,144,207.(b) Stickney, J. L.; Hubbard, A. T. J. Mol. Catal.
highly sensitive to adsorbed layer s t r ~ c t u r e ,the
~ ~present
'~ 1983,21, 211.
(11)Soriaga, M.P.; Hubbard, A. T. J.EZectroanaL Chem. 1983,159,
work was undertaken to determine the structures of ad- 101.
layers formed by treatment of Pt(ll1) with aqueous ionic (12)White, J. M.;Soriaga, M. P.; Hubbard, A. T. J . Electroanal.
Chem., in press.
(13)Stickney, J. L.;Rosasco, S. D.; Hubbard, A. T. J.Electrochem.
' Fulbright scholar, University of Jordan, Amman, Jordan. SOC.1984,131,260.

0743-7163/85/2401-0066801.50/0 0 1985 American Chemical Society

You might also like