The role of substrate identity in determining monolayer motional relaxation dynamics

Article abstract of DOI:10.1021/ja972574i

We report on the lifetime and motional dynamics of Zirconium Phosphonate (ZP) monolayers containing oligothiophene chromophores in a range of concentrations. Monolayers were formed on fused silica substrates and on a 15 ? oxide layer formed on crystalline Si(100) substrates. For both interfaces, the fluorescence lifetime behavior of the chromophores is identical and does not depend on chromophore concentration within the monolayer. Transient anisotropy measurements reveal that, for both substrates, the chromophores are oriented at ~35°with respect to the surface normal. For monolayers formed on silica, there is no evidence for chromophore motion, while motion is seen for monolayers formed on silicon. Despite the substantial similarity between the two families of monolayers, the surface roughness of the primed silicon substrate allows for greater motional freedom of the chromophores in the monolayers. We discuss these findings in the context of the differences in substrate surface roughness and domain sizes as measured by atomic force microscopy (AFM).

Full text of DOI:10.1021/ja972574i

6336
J. Am. Chem. Soc. 1998, 120, 6336-6344
The Role of Substrate Identity in Determining Monolayer Motional
Relaxation Dynamics
J. C. Horne and G. J. Blanchard*,1
Contribution from the Michigan State UniVersity, Department of Chemistry,
East Lansing, Michigan 48824-1322
ReceiVed July 28, 1997
Abstract: We report on the lifetime and motional dynamics of Zirconium Phosphonate (ZP) monolayers
containing oligothiophene chromophores in a range of concentrations. Monolayers were formed on fused
silica substrates and on a 15 Å oxide layer formed on crystalline Si(100) substrates. For both interfaces, the
fluorescence lifetime behavior of the chromophores is identical and does not depend on chromophore
concentration within the monolayer. Transient anisotropy measurements reveal that, for both substrates, the
chromophores are oriented at ∼35° with respect to the surface normal. For monolayers formed on silica,
there is no evidence for chromophore motion, while motion is seen for monolayers formed on silicon. Despite
the substantial similarity between the two families of monolayers, the surface roughness of the primed silicon
substrate allows for greater motional freedom of the chromophores in the monolayers. We discuss these findings
in the context of the differences in substrate surface roughness and domain sizes as measured by atomic force
microscopy (AFM).
Introduction
Metal-phosphonate (MP) chemistry has proven to be a remark-
ably effective and robust route to the formation of multilayer
Interfacial molecular assemblies have achieved wide use for
the modification of surfaces. Surface-modified materials have
potential application in many areas including tribology, nonlinear
1
8-34
assemblies.
Like other systems that exhibit mesoscopic
organization, metal phosphonate structures have been used
successfully in many studies, including optical second harmonic
generation, artificial photosynthesis, and light harvesting.7
2
-9
optics, and device patterning.
The most widely examined
,28,29,35-37
self-assembled monolayers (SAMs) are formed from alkanethi-
Layered metal phosphonates are attractive materials for
several reasons. Synthesis of the layered assemblies is simple,
involving immersion of a primed substrate into alternating metal
ion and R,ω-bisphosphonate solutions. The low solubility of
the complex formed between phosphonates and several metal
ions makes the structures especially robust, in contrast to
Langmuir-Blodgett layers, which are characterized by weak
interlayer associations. Zirconium phosphonate (ZP) layered
1
0,11
ols on gold.
While this family of monolayers has been
investigated extensively, the opportunity to form complex
interfacial structures is limited unless specially functionalized
molecules are used in the formation of the initial layer.
For alkanethiol monolayers, the terminal methyl groups ef-
fectively preclude chemical addition beyond the first layer, and
alternative chemistries that allow for the facile formation of
multiple layers have been developed to circumvent this problem.
1
2-17
(
18) Hong, H.-G.; Sackett, D. D.; Mallouk, T. E. Chem. Mater. 1991, 3,
521.
(19) Thompson, M. E. Chem. Mater. 1994, 6, 1168.
(20) Katz, H. E.; Wilson, W. L.; Scheller, G. J. Am. Chem. Soc. 1994,
116, 6636.
(
(
(
1) To whom correspondence should be addressed.
2) Bain, C. D.; Whitesides, G. M. Science 1988, 240, 62.
3) Wilbur, J. L.; Biebuyck, H. A.; MacDonald, J. C.; Whitesides, G.
M. Langmuir 1995, 11, 825.
(
(
4) Drawhorn, R. A.; Abbott, N. L. J. Phys. Chem. 1995, 99, 16511.
5) Xia, Y.; Zhao, X.-M.; Kim, E.; Whitesides, G. M. Chem. Mater. 1995,
(21) Yonemoto, E. H.; Saupe, G. B.; Schmehl, R. H.; Hubig, S. M.;
Riley, R. L.; Iverson, B. L.; Mallouk, T. E. J. Am. Chem. Soc. 1994, 116,
4786.
7
, 2332.
(
6) Ford, J. F.; Vickers, T. M.; Mann, C. K.; Schlenoff, J. B. Langmuir
(22) Katz, H. E.; Bent, S. F.; Wilson, W. L.; Schilling, M. L.; Ungashe,
S. B. J. Am. Chem. Soc. 1994, 116, 6631.
1
992, 12, 1944.
(
7) Kim, E.; Whitesides, G. M.; Lee, L. K.; Smith, S. P.; Prentiss, M.
AdV. Mater. 1996, 8, 139.
8) Katz, H. E.; Wilson, W. L.; Scheller, G. J. Am. Chem. Soc. 1994,
16, 6636.
9) Neff, G. A.; Zeppenfeld, A. C.; Klopfenstein, B.; Page, C. J. Mater.
Res. Soc. Symp. Proc. 1994, 351, 269.
(23) Frey, B. L.; Hanken, D. G.; Corn, R. M. Langmuir 1993, 9, 1815.
(24) Yang, H. C.; Aoki, K.; Hong, H.-G.; Sackett, D. D.; Arendt, M. F.;
Yau, S.-L.; Bell, C. M.; Mallouk, T. E. J. Am. Chem. Soc. 1993, 115, 11855.
(25) Vermeulen, L.; Thompson, M. E. Nature 1992, 358, 656.
(26) Ungashe, S. B.; Wilson, W. L.; Katz, H. E.; Scheller, G. R.;
Putvinski, T. M. J. Am. Chem. Soc. 1992, 114, 8717.
(27) Cao, G.; Rabenberg, L. K.; Nunn, C. M.; Mallouk, T. E. Chem.
Mater. 1991, 3, 149.
(
1
(
(
37.
10) Dubois, L. H.; Nuzzo, R. G. Annu. ReV. Phys. Chem. 1992, 43,
4
(
(
11) Ulman, A. Chem. ReV. 1996, 96, 1533.
12) Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990,
(28) Katz, H. E.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.;
Hutton, R. S. Chem. Mater. 1991, 3, 699.
1
12, 558.
(29) Katz, H. E.; Scheller, G.; Putvinski, T. M.; Schilling, M. L.; Wilson,
W. L.; Chidsey, C. E. D. Science 1991, 254, 1485.
(
(
(
(
13) Whitesides, G. M.; Laibinis, P. E. Langmuir 1990, 6, 87.
14) Frey, B. L.; Hanken, D. G.; Corn, R. M. Langmuir 1993, 9, 1815.
15) Sun, L.; Kepley, L. J.; Crooks, R. M. Langmuir 1992, 8, 2101.
16) Kepley, L. J.; Crooks, R. M.; Ricco, A. J. Anal. Chem. 1992, 64,
(30) Putvinski, T. M.; Schilling, M. L.; Katz, H. E.; Chidsey, C. E. D.;
Mujsce, A. M.; Emerson, A. B. Langmuir 1990, 6, 1567.
(31) Rong, D.; Hong, H.-G.; Kim, Y.-I.; Krueger, J. S.; Mayer, J. E.;
Mallouk, T. E. Coord. Chem. ReV. 1990, 97, 237.
(32) Lee, H.; Kepley, L. J.; Hong, H.-G.; Akhter, S.; Mallouk, T. E. J.
Phys. Chem. 1988, 92, 2597.
3
191.
(17) Green, J.-B. D.; McDermott, M. T.; Porter, M. D. J. Phys. Chem.
1
996, 100, 13342.
S0002-7863(97)02574-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 06/05/1998

Motional Dynamics of Zirconium Phosphonate Monolayers
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6337
Scheme 1. Structures of Oligobisphosphonates Used in This
Work: (a) 2,2′-Bithiophene-5,5′-diylbis(phosphonic acid)
(BDP); (b) 2,2′:5′,2′′:5′′,2′′′-Quaterthiophene-5,5′′′-diylbis-
(phosphonic acid) (QDP); (c) 1,12-Dodecanediylbis-
(phosphonic acid) (DDBPA); (d) 1,6-Hexanediylbis-
(phosphonic acid) (HBPA)
complexes are most common, but other metal ions have been
used as well.2
3,38,39
The structures are built up one layer at a
time with control over the specific molecular identity of each
layer, affording exquisite control over the chemical, electrical,
or optical properties of each layer and, consequently, of the
entire film.
Figure 1. Absorption (___) and emission (- - -) spectra for (a) BDP
-
5
and (b) QDP solutions (10 M) in DMSO.
consider the fundamental issue that, regardless of the chemical
structure of the information-carrying molecule, any change in
logic state will be mediated by the excited state of the molecule.
It is therefore important to place fundamental limits on the
physical and chemical consequences of effecting an optical
excitation in a multilayer assembly. Our focus is to understand
the energy migration and motional dynamics that occur within
an optically active layer. By using simple chromophores, we
can observe the presence or absence of excitation transport and
motional dynamics without complications associated with large
scale chromophore isomerization. We use oligothiophenes as
chromophores because they are relatively well understood and
form ordered layers due to their rigid structure. We chose 2,2′-
bithiophene-5,5′-diylbis-(phosphonic acid) (BDP) and 2,2′:5′,2′′:
One application for which three-dimensional, layer-by-layer
control of multilayer structure could potentially be useful is in
optical information storage. Present-day storage devices use a
single layer medium, and the addition of a useful third dimension
could enhance the storage density of a given volume of the
surface. The use of MP multilayers with chemically and
optically distinct layers could be the foundation for such a three-
dimensional structure, but there are aspects of this scheme that
could serve to limit the utility of these materials for information
storage. The lateral dimensions of each physical resolution unit
in a MP assembly will be determined by the diffraction limit
of the optical system used to read and write the information.
The stability and layer-specificity of the information written to
the material is key to the implementation of such an approach.
The chemical change associated with writing or erasing, while
still under investigation, will likely be a cis-trans isomerization
but, in any event, will involve the excited electronic state of
the molecule. The conformational change must be an activated
process with a sufficiently high barrier to ensure that the
operation is irreversible under ambient dark conditions. The
layer constituents must also be resistant to oxidative- or
photodegradation over long periods of time and use, and there
is the possibility of excitation transport from one region of the
assembly to another, within a single layer or between constitu-
ents in neighboring layers.
5
′′,2′′′-quaterthiophene-5,5′′′-diylbis(phosphonic acid) (QDP)
(Scheme 1) because the BDP emission band overlaps the QDP
absorption band, as shown in Figure 1. The spectral overlap of
the two chromophores is useful for excitation transport studies.
These chromophores can photoisomerize via ring rotation40 and
this property is of value in understanding their dynamical
response in layered assemblies (Vide infra).
In these initial studies, we concentrate on the intralayer
motional relaxation properties of each chromophore by varying
the chromophore concentration in a series of BDP and QDP
monolayers. The chromophore concentration is controlled by
incorporating optically inactive, alkanebisphosphonate diluents
into the monolayer. We chose 1,6-hexanediylbis(phosphonic
acid) and 1,12-dodecanediylbis(phosphonic acid) as diluents
because, in their all-trans conformations, they are approximately
the same lengths as BDP and QDP, respectively, enabling us
to make the monolayer thickness as uniform as possible, at least
on molecular length scales.
The purpose of this work is not to identify candidate
molecules for the actual storage of information but rather to
(33) Lee, H.; Kepley, L. J.; Hong, H.-G.; Mallouk, T. E. J. Am. Chem.
Soc. 1988, 110, 618.
(
34) Katz, H. E. Chem. Mater. 1994, 6, 2227.
(35) Vermeulen, L. A.; Snover, J. L.; Sapochak, L. S.; Thompson, M.
E. J. Am. Chem. Soc. 1993, 115, 11767.
36) Snover, J. L.; Byrd, H.; Suponeva, E. P.; Vicenzi, E.; Thompson,
M. E. Chem. Mater. 1996, 8, 1490.
(
We have characterized the chromophore dynamics of a single
layer of BDP on fused silica substrates previously.41 In this
work, we compare the behavior of QDP to BDP on fused silica
and oxidized Si(100) substrates. For BDP on silica, we do not
detect any chromophore rotational motion. The fluorescence
(
(
37) Kaschak, D. M.; Mallouk, T. E. J. Am. Chem. Soc. 1996, 118, 4222.
38) Neff, G. A.; Mahon, T. M.; Abshere, T. A.; Page, C. J. Mater. Res.
Symp. Proc. 1996, 435, 661.
39) Feldheim, D. L.; Mallouk, T. E. Chem. Commun. 1996, 2591.
(

6338 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Horne and Blanchard
Scheme 2. Conformations of QDP with Thiophene Rings
cooling to room temperature, bright yellow QDP precipitated from
(
a) All Anti and (b) All Syn and Conformations of BDP
with Thiophene Rings (c) Anti and (d) Syn
solution and was collected by vacuum filtration in 55% yield. Overall
1
yield was 21%. H NMR (DMSO-d
6
): 7.4 ppm (m). The phospho-
rylation of bithiophene to synthesize 2,2′-bithiophene-5,5′-diylbis-
(
phosphonic acid) (BDP) was very similar and has been reported
40
elsewhere.
Synthesis of 1,12-Dodecanediylbis(phosphonic acid) (DDBPA)
and 1,6-Hexanediylbis(phosphonic acid) (HBPA). All chemicals
used in the synthesis were purchased from Aldrich Chemical Co.
Triethyl phosphite was treated with sodium and vacuum distilled.
4
2
DDBPA was prepared by the Michaelis-Arbuzov reaction of 1,12-
dibromododecane with triethyl phosphite. The mixture was refluxed
for 5 h to allow evolution of ethyl bromide. Excess triethyl phosphite
was removed by vacuum distillation. Concentrated HCl was added to
the solution, which was then refluxed for 12 h, and the resulting white
precipitate was collected by vacuum filtration and washed with
acetonitrile. The yield was 68%. 1H NMR (DMSO-d
m). The synthesis of HBPA was identical, using 1,6-dibromohexane
as starting material.
6
): 1.2-1.5 ppm
(
Metal-Phosphonate Multilayer Synthesis. Fused silica and pol-
ished Si(100) (MultiScanning Plus) substrates were cleaned by piranha
solution etch (Caution! Piranha solution is extremely corrosiVe and
is a potent oxidizer) for 10 min and rinsed with flowing distilled water,
followed by hydrolysis in 2 M HCl for 5 min, and a final distilled
dynamics of QDP on silica are essentially the same, as we
describe below in more detail.
2
water rinse. Samples were dried with a N stream and used shortly
The identity of the substrate affects the dynamical behavior
of the chromophores. The 15 Å oxide layer on Si(100)
possesses the same active sites (-OH) as silica, but there are
fundamental differences in site density and surface roughness
on oxidized silicon that affect the primer density and organiza-
tion and thus influence the adlayers. The steady-state optical
response and the fluorescence lifetime behavior were identical
on both substrates, for a given chromophore. We detect
chromophore motion on silicon, and this finding suggests
substantially different organization and structural freedom in
the two systems. The active sites on silica appear to allow a
higher chromophore density on the initial phosphonate primer
layer than is seen for silicon, giving rise to an improvement in
the ordering of the first layers. We present AFM images of
the SiOx substrates that complement the dynamic spectroscopic
measurements and demonstrate the differences in surface
roughness of the substrates.
after for deposition. The substrates were primed by reflux in a 1%
v/v solution of 3-(aminopropyl)triethoxysilane in anhydrous octane for
1
0 min, followed by rinses with hexane and water. After drying with
, the amine surface was derivatized to the phosphonate in a solution
of 0.1 M POCl and 0.1 M collidine in anhydrous CH CN for 1 h,
CN and water. The resulting surface was
N
2
3
3
followed by rinses with CH
functionalized with Zr by immersion in a 5 mM solution of ZrOCl
3
4+
2
in 60% aqueous ethanol solution for 10 min. Monolayers were
deposited from bisphosphonate solutions with a total 1 mM concentra-
tion ([chromophore] + [alkane] ) 1 mM). Alkanebisphosphonate,
BDP, and mixed alkanebisphosphonate/BDP solutions were made in
90% EtOH (aq); QDP and mixed alkanebisphosphonate/QDP were
made in 80% DMSO, 20% (90% EtOH (aq)). The substrates were
immersed in the appropriate bisphosphonate solution for 4 h at 55 °C.
Steady-State Optical Spectroscopy. The absorbance spectra of
oligothiophene chromophores in solution and on surfaces were measured
using a Hitachi U-4001 UV-visible spectrophotometer. Monolayer
samples were held vertically in place at a 45° angle with respect to the
incident beam. Spectra were collected with 5 nm resolution. Fluo-
rescence spectra of solutions were measured on a Hitachi F-4500
fluorescence spectrophotometer. Excitation and emission slits were
adjusted according to spectral intensity. Monolayer emission spectra
were collected on a SPEX Fluorolog 2 spectrophotometer. Samples
were mounted vertically and at a 45° angle with respect to excitation,
to minimize reflection of excitation light into the emission collection
optics. Slits were adjusted according to spectral intensity.
Experimental Section
Synthesis of 2,2′:5′,2′′:5′′,2′′′-Quaterthiophene-5,5′′′-diylbis(pho-
sphonic acid) (QDP) and 2,2′-Bithiophene-5,5′diylbis(phosphonic
acid) (BDP). QDP was synthesized with minimal modification of a
procedure published previously.27 Tetrahydrofuran (THF) was distilled
over CaH and sodium before use. Quaterthiophene (QT) was prepared
2
by the reaction of 20 mmol bithiophene (Aldrich Chemical Co.) in
THF with 1 equiv of n-butyllithium (2.5M in hexanes, Aldrich) at -78
Time Correlated Single Photon Counting Spectroscopy (TCSPC).
TCSPC was used to measure excited-state fluorescence lifetimes and
rotational dynamics of the chromophores. This system has been
described in detail previously,43 and we review it briefly here. The
light pulses used to excite the sample are generated with a cavity
dumped, synchronously pumped dye laser (Coherent 702-2) pumped
by the second harmonic output of a mode-locked CW Nd:YAG laser
°
C for 20 min. The mixture was transferred by cannula to a second
2
flask containing a suspension of anhydrous CuCl (excess) in THF.
The clear yellow solution was allowed to warm slowly to room
temperature and was stirred overnight. The addition of 50 mL of 1 M
HCl resulted in the formation of a yellow precipitate in a brown-green
solution, which was vacuum filtered and dried (64% yield). A solution
of QT (3 mmol) in THF was cooled to -40 °C, and an excess (12.5
mmol) of n-BuLi was added. After stirring for 2 h, a second solution
of 12.5 mmol bis(dimethylamino)phosphochloridate (Aldrich) and 0.625
mmol n-BuLi in THF was added to the first solution via cannula, and
the resulting solution was allowed to warm to room temperature. The
reaction was stopped after 3 days, the product was extracted from water
(Quantronix 416). The excitation beam was frequency doubled (LiIO
3
Type I SHG) to excite BDP samples at 320 nm (640 nm fundamental,
Kiton Red, Exciton) and QDP samples at 390 nm (780 nm fundamental,
LDS 821, Exciton). Monolayer samples were held approximately
horizontally, with 5° tilts away from the horizontal in two directions:
toward the excitation beam and toward the detector. Fluorescence from
the sample was imaged through a reflecting microscope objective.
Lifetimes were collected across the emission bands at 54.7° with respect
to the excitation polarization for solutions and without polarization bias
for monolayers. While such collection can, in principle, lead to small
contributions to the measured relaxation from motional dynamics, for
with ether, and the aqueous layer was extracted twice with CH
The product was purified chromatograpically on silica plates (Fisher)
using a mobile phase of 1:1 CH Cl and THF. The most polar
2 2
Cl .
2
2
component was collected in 46% yield. A portion of this tetraamide
was dissolved in dioxane at 80 °C. The solution was acidified with
2
50 mL each of H O and concentrated HCl and refluxed overnight. Upon

Motional Dynamics of Zirconium Phosphonate Monolayers
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6339
Figure 3. Dependence of monolayer absorbance on concentration of
QDP in deposition solution and regression of experimental data.
Figure 2. Representative instrument response function (0), excited-
state population decay for QDP in DMSO (10 M) (O) and excited-
state population decay for QDP in a monolayer (1%) (b).
-
5
Results and Discussion
our experimental conditions this is not a problem.44,45 Fluorescence
was collected at 390 and 465 nm (5 nm fwhm bandwidth for solution,
In this section we compare the steady-state and transient
optical responses of four optically active chemical systems using
two chromophores and two substrates. A basic understanding
of the optical response and the intralayer relaxation effects of
each chromophore is necessary for future investigation of
interlayer energy transport between the two chromophores in
complex multilayer structures. In addition to the determination
of the optical properties of these layers, the effect of substrate
must also be considered. Several measurements contribute to
our understanding of the optical response, including UV-visible
absorption, fluorescence emission, excited-state population
decays, and induced orientational anisotropy. The results from
the anisotropy measurements are critical for comparing the
systems. We detail the subtle differences between the systems
below.
Monolayers on Fused Silica. Steady-State Spectroscopy.
BDP monolayers on silica were characterized previously.40 For
BDP and HBPA mixed monolayers, we found a linear relation-
ship between the proportion of chromophore in the deposition
solution and the proportion of chromophore in the resulting
monolayer. The calibration did not regress through the origin,
however. This is a simple partitioning effect and can be
understood in terms of preferential solvation of HBPA over BDP
in ethanol. For QDP and DDBPA mixed monolayers, we also
observe a calibration that is linear in QDP concentration, shown
in Figure 3. In this case, the solvent system used for the
monolayer deposition solution gives rise to a calibration curve
that passes through the origin. The implication of these data is
that solution concentration ratios will result in a monolayer with
the same constituent ratio, if the solvent system is chosen
correctly. The proportions of constituents that are incorporated
into a mixed monolayer depend mostly on their solubilities in
the solvent used.
30 nm for monolayers) for BDP and QDP, respectively, with polariza-
tions of 0° and 90° for orientational anisotropy measurements on all
samples. A representative lifetime decay and instrument response
function (∼35 ps fwhm) are shown in Figure 2.
Data Analysis. The lifetimes we report here were fit to sums of
exponentials using commercial software (Microcal Origin) and are
reported as the averages and standard deviations of 4-6 individual
decays. Rotational dynamics parameters were determined in two ways.
For monolayers on silica, eight pairs of alternating parallel and
perpendicular scans were used to produce an average anisotropy
function. Each scan pair was recorded for a different physical location
on the substrate. R(0) was determined to (0.01 by regression of data
at times after the instrumental response, and several data sets were
averaged for each sample. For monolayers on oxidized Si(100), the
rotational diffusion information is reported as the averages and standard
deviations of data from several decays. Four pairs of alternating parallel
and perpendicular scans were used to produce an average anisotropy
function, which was fit with Origin to determine the parameters R(0),
τMR, and R(∞).
Calculations. Semiempirical calculations were performed using
Hyperchem Release 4.0 (Hypercube, Inc.). The PM3 parametrization
used for these calculations is a modification of the AM1 parametrization
that treats molecules containing heteroatoms, such as sulfur, more
accurately than previous parametrizations. An initial optimization of
the structure was performed using a molecular mechanics routine
(MM+) followed by geometry optimization at the semiempirical level
using an SCF algorithm. The torsions of the QDP 2-2′ and 5′-5′′ σ
bonds were set at 10° increments, and semiempirical optimization was
performed until the lowest energy conformation for the fixed inter-
0 1
ring bond torsions was reached. The heat of formation and S T S
transition energy were calculated for each geometrically optimized
conformation.
(
40) Horne, J. C.; Blanchard, G. J.; LeGoff, E. J. Am. Chem. Soc. 1995,
1
17, 9551.
The steady-state emission spectra for both BDP and QDP
are similar in the sense that they reveal more than one form of
the chromophore in the monolayer. The BDP spectrum has two
features, a narrow peak at 390 nm that we see for dilute (<35%
BDP) monolayers and a broader, less intense peak at 425 nm
(
(
(
41) Horne, J. C.; Blanchard, G. J. J. Am. Chem. Soc. 1996, 118, 12788.
42) Bhattacharya, A. K.; Thyagarajan, G. Chem. ReV. 1981, 81, 415.
43) DeWitt, L.; Blanchard, G. J.; LeGoff, E.; Benz, M. E.; Liao, J. H.;
Kanatzidis, M. G. J. Am. Chem. Soc. 1993, 115, 12158.
(
(
44) Flory, W. C.; Blanchard, G. J. Appl. Spectrosc. 1998, 52, 82.
45) Spencer, R. D.; Weber, G.; J. Chem. Phys. 1970, 52, 1654.

6
340 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Table 1. Excited State Lifetimes of Monolayersa
2
(ps) τ (ps)
Horne and Blanchard
τ
1
BDP on silica
QDP on silica
BDP on silicon
QDP on silicon
187 ( 11
232 ( 16
213 ( 12
243 ( 23
1140 ( 337
1265 ( 138
1277 ( 96
1115 ( 156
a
The time constants refer to fits of the experimental data to the
function f(t) ) A exp(-t/τ ) + A exp(-t/τ ).
1
1
2
2
that we observe for concentrated monolayers. These bands
correspond to the two peaks present in concentrated solution
spectra. QDP exhibits four unresolved bands in solution and
typically three are present in monolayer spectra (see Figure
4a,b). For the monolayers, the 445 nm band is not observed.
The QDP spectra also depend on concentration in the monolayer.
Blue-shifted peaks are prominent in the dilute monolayer spectra,
and red-shifted features dominate the concentrated monolayer
spectra. Over time, the BDP spectra anneal to spectrum
40
dominated by the blue band, but it is not clear from our data
that QDP monolayers exhibit this same evolution. The com-
plexity of these monolayer spectra is also manifested in the
temporal evolution of emission intensity. We will discuss this
behavior in more detail in a forthcoming paper.46
Transient Spectroscopies. The dilute solution behavior of
both BDP and QDP is that of a simple chromophore. The
absorption and emission bands of the steady-state spectra are
well-defined, and dilute solution lifetimes are fit to a single
exponential decay for both chromophores. The lifetime for
dilute solutions is approximately 200 ps for BDP and 450 ps
for QDP. The population decays of monolayers approximate a
double exponential, as shown in Figure 2. The fast component
of the monolayer lifetimes is ∼ 200 ps for both chromophores
and the longer lifetime is ∼ 1.2 ns, as shown in Table 1. We
interpret the fast monolayer lifetime as that due primarily to an
isolated chromophore species in an alkane-like environment,
owing to its similarity to the solution phase value, and the long
lifetime primarily to a chromophore within an aggregated island.
These assignments are, of course, only approximations, and the
actual contributions to the measured decay are not as clear-cut.
This relaxation behavior is well understood, and we will present
a detailed treatment elsewhere.45 The lifetimes do not vary with
respect to either chromophore density in the monolayer or
emission wavelength collected for either chromophore. For this
reason, the lifetimes reported in Table 1 are averages over all
concentrations and emission wavelengths used for chromophore
monolayers. However, the fractional contribution of each
lifetime component varies with chromophore concentration and
emission wavelength. For BDP, as the chromophore concentra-
tion is increased or as emission is collected on the red edge of
the emission band, the relative contribution of the long lifetime
increases. This trend is not observed with QDP, as the
contributions of the different species to the total signal varied
with no simple pattern. This effect may be due to the greater
complexity of the QDP emission spectrum compared to BDP.
Rotational motion measurements are not complicated by
population decay functionality effects and can provide insight
into how the chromophores move within their environment,
offering an indication of how ordered the layers are. Rotational
diffusion measurements are most typically made on probe
molecules in solution, where the chromophores can relax into
all possible orientations. We have measured the dynamics of
BDP and QDP in monolayers. Because one end of each
chromophore is attached to a substrate, the motion of the
chromophore is confined to a conic volume. The cone dimen-
sions are described by Lipari and Szabo’s hindered rotor
Figure 4. Emission spectra of (a) QDP monolayers (- - -) 1%; (s)
-
7
5%; (‚‚‚) 100% and (b) QDP solution in DMSO (- - -) 10 M; (s)
-
6
-3
1
0
M; (‚‚‚) 10 M. Spectra were normalized for presentation, and
no information about fluorescence quantum yield should be inferred
from these plots.
model,47 where the base of the cone is the point of attachment
of the chromophore to the substrate, and the semiangle of the
cone is designated θ0. A nonrandom orientational distribution
of the chromophore ensemble in the monolayer is selected with
a polarized excitation pulse, and components of the emission
that are polarized parallel and perpendicular to the excitation
polarization are collected. These data (Figure 5a, for example)
are combined to construct the induced orientational anisotropy
function, R(t)
I (t) - I (t)
|
⊥
R(t) )
(1)
I (t) + 2I (t)
|
⊥
The decay of R(t) (see Figure 5b) is fit to
R(t) ) R(∞) + [R(0) - R(∞)] exp(-t/τMR
)
(2)
τMR is the motional relaxation time constant and the quantity
R(0) is related to the angle between the chromophore absorbing
and emitting transition moments, δ
2
5
R(0) ) P (cos δ)
(3)
2
where P2 is the second-order Legendre polynomial. For BDP
in solution, R(0) ) 0.24, which corresponds to an angle of δ )
31°. For QDP in solution, R(0) ) 0.40, indicating parallel
transition moments. We consider that the QDP transition
moments lie along the long molecular axis. We use these values
for δ in the treatment of monolayer data because this quantity
is intrinsic to the chromophore and, for the solution data,
represents an average over all conformers. We note that R(0)
does depend to a certain extent on environment because the
intramolecular rotational freedom available to the chromophore
depends on the limitations imposed by its surroundings (Vide

Motional Dynamics of Zirconium Phosphonate Monolayers
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6341
of the thiol headgroups on the crystalline gold substrate surface.
For ZP layers, it is not as obvious that intermolecular interactions
dominate the tilt angle because, for the substrates we use, the
spacing between primed sites is significantly less well deter-
mined than for a crystalline metal surface. In the case of ZP
layers, it may be the bonding geometry of the metal-phosphonate
connecting layers that determines the range of structural
properties available to the system.
We do not detect any chromophore motion in either BDP or
QDP monolayers on silica, as indicated by the absence of a
4
0,48
decay in R(t).
At times after the instrument response
function, regression of R(t) yields a line of zero slope. We can
evaluate the possibility that the chromophores are reorienting
significantly faster than our detection system time resolution
(
2
∼35 ps). The time constants of reorientation in solution are
99 ( 15 ps for BDP in ethanol and 1.2 ( 0.2 ns for QDP in
DMSO. Based on viscosity considerations, we believe that the
chromophore imbedded in the monolayer will not reorient more
rapidly than in low viscosity solvents such as ethanol or DMSO.
If there was any motion of the chromophore in the monolayer,
it would occur on a time scale we can determine because of the
proximity of neighboring molecules and dipolar and van der
Waals interactions between them.
Monolayers on Silicon. The substrate on which self-
Figure 5. Reorientation data for a 50% QDP monolayer on silica. (a)
Emission intensity data for parallel and perpendicular polarizations.
assembling structures are synthesized affects the ordering of
9
,10,49-53
the adlayers.
This effect is seen for alkanethiol layers
(b) Induced orientational anisotropy function of above data with fit
on various metal surfaces, because the infrared-active symmetric
and asymmetric CH2 stretching modes are sensitive indicators
of aliphatic chain ordering. We compare the optical response
of the chromophore monolayers on silica to that on oxidized
Si(100) substrates to evaluate the role of substrate-induced
chromophore ordering in determining the dynamical response
of the system. Frey and co-workers have shown effects of the
(
- - -).
Table 2. Induced Orientational Anisotropy Data for Monolayersa
R(0) MR (ps) R(∞) θ (deg)
τ
BDP on silica
QDP on silica
0.06 ( 0.01 35 ( 2
0.07 ( 0.05 39 ( 7
BDP on silicon 0.40 ( 0.15 393 ( 119 0.18 ( 0.08 19 ( 11
QDP on silicon 0.32 ( 0.06 479 ( 153 0.13 ( 0.06 32 ( 7
1
3
substrate on the order of ZP layers, but this effect is seen in
the inorganic portion of the IR spectrum (νa (PO32 )) rather
than in the aliphatic CH2 stretching region. We chose silicon
substrates for the comparison because the surface priming
chemistry is identical to that of silica, due to the 15 Å thick
layer of native oxide that is present on silicon under our
experimental conditions.
-
a
For BDP, the angle between the transition moments is δ ) 31°
and for QDP, δ ) 0°. Angles θ were determined from δ and R(∞)
data using eq 4.
infra). Specifically, syn and anti conformers are likely to exhibit
different values of δ.
For chromophores in solution, R(∞) ) 0 because they can
completely re-randomize at infinite time. For chromophores
in monolayers, |R(∞)| > 0 because of the geometric restriction
on their motional freedom. R(∞), the steady-state anisotropy,
is related to θ, the average tilt angle of the chromophore within
It is fair to consider that vibrational spectroscopy could
address the question of monolayer order for ZP layers on these
two substrates. Aside from the different dielectric responses
13
of the two substrates, especially in the IR spectral region,
vibrational spectroscopy will not be as sensitive to order for
our samples as they are for pure alkanebisphosphonate layers.
Contributions from the thiophene rings combined with the
structural perturbations they may induce in adjacent aliphatic
domains will serve to obscure structural information for our
samples. The strong absorbance of SiOx at low energies
obscures information about the inorganic portion of the struc-
tures. We therefore rely on lifetime and anisotropy measure-
ments that are sensitive to the local environment of the
oligothiophene chromophores.
4
6
the cone by
2
5
2
R(∞) ) P (cos θ )P (cos θ )〈P (cos θ)〉
(4)
2
ex
2
em
2
The angles θex and θem are the transition moment angles with
respect to the cone center axis. The calculation of θ is simplified
by setting θex ) 0° and θem ) δ. In our previous work we
reported that BDP on SiOx exhibited a concentration-indepen-
dent steady-state anisotropy of 0.06 ( 0.01, which yields an
average tilt angle of 35° ( 2°.40 We report here that QDP on
silica has a similar, concentration-independent steady-state
anisotropy of 0.07 ( 0.05 and an average tilt angle of 39° ( 7°
The concentration dependence of the lifetime data is important
in understanding excitation transport in these monolayers. If
excitation transport does play a significant role in the transient
response of these systems, then we would expect a highly
(Table 2). Our measurement of the same tilt angles for both
5
4
chromophores implies that the oligothiophene/alkane environ-
ment is not affected by the lengths of the constituent molecules.
We note that the angles we report here for layer tilt are quite
similar to the tilt angle of ∼30° seen for alkanethiol monolayers
concentration-dependent population decay. The fact that we
observe concentration independent lifetime behavior indicates
that either excitation transport in this system does not play an
important role in determining the observed motional dynamics,
excitation transport dynamics are faster than can be resolved
for all of the systems we examined, or the double exponential
9
on gold. In the case of alkanethiol/gold SAMs, the tilt angle
is determined by aliphatic interchain interactions and the registry

6342 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Horne and Blanchard
Figure 6. Emission intensities (a) and induced orientational anisotropy
function (b) for a 26% BDP monolayer on silicon.
Figure 7. Emission intensities (a) and induced orientational anisotropy
function (b) for a 1% QDP monolayer on silicon.
decay is itself characteristic of the excitation transport dynamics.
We consider this issue in detail elsewhere.45 The chromophores
we have chosen possess moderately high fluorescence quantum
which is one possible explanation for the dynamics we detect.
If this is the dominant motion in our systems, a larger or longer
chromophore would experience more hindered movement and
would exhibit a correspondingly longer relaxation time. This
is essentially a viscous drag argument, borrowing from the
underlying principles of the Debye-Stokes-Einstein model for
-
1
yields and large absorption cross sections (ꢀmax ) 27 200 M
-
1
-1
-1
cm for BDP and 27 300 M cm for QDP). Because the
oligothiophene chromophores relax primarily by radiative decay
and they possess a large Stokes shift (Figure 1), the oligoth-
iophene chromophores are ideal candidates for energy transport
studies.45 We are presently characterizing intralayer excitation
transport in mixed monolayers of BDP and QDP. Multilayer
assemblies of these chromophores will be useful in investigating
interlayer excitation transport phenomena.
5
6,57
rotational diffusion in liquids.
If large amplitude motion dominates our experimental data,
then we would also expect to see a concentration dependence,
since the thiophene chromophores occupy a larger “footprint”
in the lattice sites than alkanes do. A chromophore with
chromophore neighbors will move more slowly in its environ-
ment than if it were surrounded by aliphatic constituents because
of geometric constraints and dipolar coupling. The average
motional relaxation time, τMR (Table 2), for BDP and QDP on
silicon are the same to within the experimental uncertainty,
arguing that the relaxation we measure is not large amplitude
precessional motion within a conic volume.
While the lifetime data for the chromophores reveal no
discernible dependence on the identity of the substrate, the
rotational dynamics of BDP and QDP on silicon are fundamen-
tally different than those on silica. As we show in Figures 6
and 7, a measurable decay of the anisotropy is present for BDP
and QDP, with an average reorientation time of ∼435 ps for
both, that was not seen for these chromophores bound to silica.
These data were taken over a broad spectral range, with emission
collected at 465 nm ( 15 nm, and the values reported in Table
Another type of motion that is important in these chro-
mophores is intramolecular relaxation, specifically the rotation
of the thiophene rings about their connecting bonds. This
motion is intrinsic to the molecule, and is not limited to a
2
are averaged over all concentrations because there was no
statistically meaningful concentration dependence. The average
tilt angles, θ, calculated from R(∞) using eq 4, are approximately
the same for both QDP and BDP monolayers, ∼35°, (Table 2)
39
hindered environment. As seen in Scheme 2, a transition from
(46) Horne, J. C.; Huang, Y.; Liu, G.-Y.; Blanchard, G. J. J. Am. Chem.
Soc. Manuscript in preparation.
which correlates well with tilt angles determined for other MP
(
47) Lipari, G.; Szabo, A. Biophys. J. 1980, 30, 489.
monolayers.2
3,55
We note that the average tilt angles of the
(48) Karpovich, D. S.; Blanchard, G. J. Langmuir 1996, 12, 5522.
chromophores on silica are similar, so measurement of θ cannot
explain the existence of motion on the oxidized silicon substrate
and not silica.
(49) Smith, E. L.; Porter, M. D. J. Phys. Chem. 1993, 97, 8032.
50) Fenter, P.; Eisenberger, P.; Li, J.; Camillone, N. III; Bernasek, S.;
Scoles, G.; Ramanarayanan, T. A.; Liang, K. S. Langmuir 1991, 7, 2013.
51) Chau, L.-K.; Porter, M. D. Chem. Phys. Lett. 1990, 167, 198.
(52) Chen, S. H.; Frank, C. W. Langmuir 1989, 5, 978.
(
(
Origins of Chromophore Motion. A necessary first step
is to determine the nature of the molecular motion we sense in
a single layer. In particular, we need to distinguish between
inter-ring isomerization and large amplitude precessional motion.
The hindered rotor model describes the physics of a chro-
mophore imbedded in a monolayer. It is typically assumed that
a decay in R(t) is related to large amplitude motion of a
chromophore in its environment about its tether to the surface,
(
(
53) Sondag, A. H. M.; Raas, M. C. J. Chem. Phys. 1989, 91, 4926.
54) Gochanour, C. R.; Andersen, H. C.; Fayer, M. D. J. Chem. Phys.
1
979, 70, 4254.
(55) O’Brien, J. T.; Zeppenfeld, A. C.; Richmond, G. L.; Page, C. J.
Langmuir 1994, 10, 4657.
56) Debye, P. Polar Molecules; Chemical Catalog Co.: New York,
929.
57) Fleming, G. R. Chemical Applications of Ultrafast Spectroscopy;
Academic Press: New York, 1986.
(
1
(

Motional Dynamics of Zirconium Phosphonate Monolayers
J. Am. Chem. Soc., Vol. 120, No. 25, 1998 6343
an all-anti to an all-syn conformation of QDP represents a
substantial change in the orientation of the transition moments
relative to the substrate. We believe that rotation of the ring(s)
is more reasonable, and it can account for our data (Vide infra).
The optical anisotropy decay is a correlation function for the
transition moment relaxation, so the dynamics we observe can
be accounted for by inter-ring conformational changes in QDP
and large amplitude motion does not need to be invoked. The
anisotropy decay due to isomerization is not as pronounced for
BDP as it is for QDP because BDP possesses fewer confor-
mational degrees of freedom, but there is still a discernible
change in the transition moment angle with respect to the
substrate normal (Scheme 2). The BDP and QDP measured
parameters of R(0), τMR, and R(∞) are likely similar because
they are constrained to the same molecular area, as dictated by
the substrate, primer, and metal phosphonate lattice structure.
Within the constraints of a well-packed, uniform monolayer,
BDP and QDP may actually have the same net change in
transition moment direction, because QDP will not be able to
access all of the degrees of conformational freedom available
to an isolated molecule (Scheme 2b).
Based on semiempirical calculations,5
8-62
rotation of the
thiophene rings of an isolated ground state chromophore is a
low energy activated process and certainly occurs at room
temperature. We have reported previously that the BDP
chromophore possesses low barriers to rotation between con-
formations in the ground state (<1 kcal/mol).40 This finding
implies that there will be a statistical distribution of conformers
that are excited, with any bias in their conformational distribu-
tion being induced by environmental, not structural constraints.
In Figure 8, we show results from semiempirical calculations
on QDP for twists between two end rings and between the two
center rings. These plots are representative of most energetically
favorable conformations, with the exclusion of contributions
from the twist between the other two end rings, which were
held at 0° in the anti conformation. In the first excited singlet
state, energetic minima are present at approximately 0° and 180°
ring torsion (anti and syn), while a maximum of ∼10 kcal/mol
is calculated for an inter-ring angle of 90°. This maximum is
the result of the tendency of thiophenes to form a planar quinoid
structure in the excited state. The syn and anti conformations
correspond to the lowest energy difference between S0 and S1,
so transitions associated with these conformations are expected
to produce the red shifted features we see. From these
calculations we can see that QDP exhibits facile interconversion
between all conformations in the ground state and the excited-
state possesses definite minima. These calculations are, of
course, qualitative and are limited by the utility of the PM3
parametrization in the calculation of S1 surfaces. Despite these
limitations, they do demonstrate the likelihood for the chro-
mophores to exhibit ring rotation in the S1, consistent with
intramolecular relaxation being the dominant phenomenon we
detect in the dynamical measurements.
Figure 8. Calculated energy barriers for QDP inter-ring rotation for
the first two inter-ring bonds: (a) S and (b) S . Twist angles indicated
0 1
are for angles with respect to their adjacent ring.
based on the data for the Si(100) substrate and comparison to
bulk liquid data that motion of the thiophene rings on silica
would occur within a time window accessible to our measure-
ments. We believe that thiophene ring rotation on the silica
substrate must be hindered in some way. A simple difference
in the dielectric responses of the two substrates cannot account
for the differences in motional behavior, because the population
dynamics, which depend sensitively on the dielectric response
of the substrate, do not vary.
The surface morphology of the substrates is more likely to
provide an explanation for our observations. As shown in
atomic force microscopy (AFM) images (Figure 9), the two
substrates are characterized by very different surface features.
The silica has a granular, irregular surface, and the oxidized
silicon exhibits striations, probably a result of polishing. The
grooves in the oxidized Si surface are 200-500 Å apart and
10-30 Å deep. The roughness of the two surfaces was
determined from the fwhm of a histogram of the imaged area,
which is slightly narrower for the silica substrate. On a length
scale more relevant to our spectroscopic measurements (50 Å),63
the variations in height across the surfaces reveal much flatter
domains on silica ((2 Å) than on silicon ((10 Å). Because of
the greater variation in surface relief, chromophores on silicon
should have more motional freedom than on a corresponding
monolayer formed on silica. An additional piece of information
The semiempirical calculations serve to limit the possible
explanations for why we observe motional relaxation for
monolayers on oxidized Si(100) and not on fused silica. The
observation of motion on the silicon substrate could be taken
to imply that we are observing motion in this system because
the rotation of the chromophore rings is not hindered. It is clear
(
58) Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J.
Am. Chem. Soc. 1985, 107, 3902.
(63) We estimate the F o¨ rster critical transfer radius for these chro-
mophores to be 25 Å. For rotational diffusion measurements, the relevant
distance is probably shorter, on the order of the nearest neighbor spacing.
Due to the amorphous nature of these surfaces, we can only estimate this
distance to be e 10 Å.
(
(
(
(
59) Dewar, M. J. S.; Dieter, K. M. J. Am. Chem. Soc. 1986, 108, 8075.
60) Stewart, J. J. P. Comput.-Aided Mol. Des. 1990, 4, 1.
61) Dewar, M. J. S.; Thiel, W. J. Am. Chem. Soc. 1977, 99, 4899.
62) Dewar, M. J. S.; Thiel, W. J. Am. Chem. Soc. 1977, 99, 4907.

6344 J. Am. Chem. Soc., Vol. 120, No. 25, 1998
Horne and Blanchard
Figure 9. Atomic force microscopy images for 300 nm × 300 nm areas of oxidized silicon (top) and fused silica (bottom) substrates. Cursor plots
show height profiles along the line indicated in the 3-D images. Image histograms indicate surface relief (∆z) distribution of area imaged.
that corroborates this explanation is that for SiOx, xsilicon < xsilica,
which could result in a less dense layer on silicon. Haller has
shown that, for this priming chemistry on silicon, under all
except the most stringent conditions, the triethoxysilane primer
polymerizes easily, resulting in large islands on the surface
instead of a single layer.64 It is possible that the higher density
of hydroxyl active sites on the silica surface precludes much of
the polymerization because the tighter packing does not allow
as much multilayer buildup. For silicon, with fewer active sites
on which to grow the bisphosphonate monolayer, the chro-
mophores will be spaced further apart, and therefore ring rotation
is less hindered. The relative active site density for the two
substrates does not explain, however, why we observe a larger
fluorescence signal from the monolayers on silicon than on
silica. One factor is that the polished silicon substrate is more
reflective, so that a larger fraction of emitted light is collected.
A more important factor is that, in the more dense monolayers
on silica, aggregation of the chromophores is likely. Aggregates
in the monolayers formed on different substrates. For both silica
and oxidized Si(100) substrates, the fluorescence lifetime
behavior of each chromophore is identical and is concentration
independent, pointing to the limited contribution of intralayer
excitation transport in these systems. Transient anisotropy
measurements reveal that, for both substrates, the chromophores
are oriented at ∼35° with respect to the surface normal. For
monolayers formed on silica, there is no evidence for chro-
mophore motion, while motion is seen for monolayers formed
on the silicon substrate. Despite the substantial similarity
between the two families of monolayers, the surface roughness
of the polished silicon substrate allows for greater motional
freedom of the chromophores in the monolayers. The com-
paratively lower site density for active hydroxyl groups on the
oxidized Si(100) surface likely also plays a role in determining
the observed dynamics. These data underscore the important
role of substrate morphology in determining the properties of
molecular interfaces and provide a foundation for the examina-
tion of interlayer excitation transport studies.
typically exhibit lower emission quantum yields than isolated
chromophores,65 and the less densely packed monolayers on
silicon will, on average, have more nonaggregated chro-
mophores. Regardless of these qualitative observations relating
to emission intensity, our motional relaxation data can be
understood in the context of substrate roughness, as character-
ized by the AFM data.
Acknowledgment. We are grateful to the National Science
Foundation for support of this work through Grant CHE 95-
08763. We are deeply indebted to Prof. G. Y. Liu and Y. Huang
(Wayne State University) for their assistance in AFM imaging
of the oxidized silicon and fused silica substrates.
Conclusion
JA972574I
The lifetime and motional dynamics of oligothiophene
chromophores within ZP monolayers reveal a subtle difference
(65) Song, Q.; Bohn, P. W.; Blanchard, G. J. J. Phys. Chem. B. 1997,
(64) Haller, I. J. Am. Chem. Soc. 1978, 100, 8050.
101, 8865.