Molecular engineering of supramolecular scaffold coatings that can reduce static platelet adhesion

Article abstract of DOI:10.1021/ja0775927

Novel supramolecular coatings that make use of low-molecular weight ditopic monomers with guanine end groups are studied using fluid tapping AFM. These molecules assemble on highly oriented pyrolytic graphite (HOPG) from aqueous solutions to form nanosized banding structures whose sizes can be systematically tuned at the nanoscale by tailoring the molecular structure of the monomers. The nature of the self-assembly in these systems has been studied through a combination of the self-assembly of structural derivatives and molecular modeling. Furthermore, we introduce the concept of using these molecular assemblies as scaffolds to organize functional groups on the surface. As a first demonstration of this concept, scaffold monomers that contain a monomethyl triethyleneglycol branch were used to organize these functional units on a HOPG surface. These supramolecular grafted assemblies have been shown to be stable at biologically relevant temperatures and even have the ability to significantly reduce static platelet adhesion.

Full text of DOI:10.1021/ja0775927

Published on Web 01/05/2008
Molecular Engineering of Supramolecular Scaffold Coatings
that Can Reduce Static Platelet Adhesion
Aryavarta M. S. Kumar,† Sona Sivakova,‡ Justin D. Fox,‡ Jennifer E. Green,†
Roger E. Marchant,*,†,‡ and Stuart J. Rowan*,‡,†,§
Center for CardioVascular Biomaterials, Department of Biomedical Engineering and Department
of Macromolecular Science and Engineering, and Department of Chemistry, Case Western
ReserVe UniVersity, CleVeland Ohio 44106
Received October 2, 2007; E-mail: stuart.rowan@case.edu; roger.marchant@case.edu
Abstract: Novel supramolecular coatings that make use of low-molecular weight ditopic monomers with
guanine end groups are studied using fluid tapping AFM. These molecules assemble on highly oriented
pyrolytic graphite (HOPG) from aqueous solutions to form nanosized banding structures whose sizes can
be systematically tuned at the nanoscale by tailoring the molecular structure of the monomers. The nature
of the self-assembly in these systems has been studied through a combination of the self-assembly of
structural derivatives and molecular modeling. Furthermore, we introduce the concept of using these
molecular assemblies as scaffolds to organize functional groups on the surface. As a first demonstration
of this concept, scaffold monomers that contain a monomethyl triethyleneglycol branch were used to organize
these “functional” units on a HOPG surface. These supramolecular grafted assemblies have been shown
to be stable at biologically relevant temperatures and even have the ability to significantly reduce static
platelet adhesion.
Introduction
form highly ordered (epitaxial) surface assemblies commensu-
rate with the ordered lattice of graphite.4,5 In addition, tailoring
Supramolecular chemistry at the interface plays a defining
role in the “bottom-up” approach to nanoarchitectures which
have a myriad of potential technological applications in areas
such as nanoelectronics, biological coatings, and catalytic
processes.1,2 In designing new solid-liquid interfacial (surface)
assemblies, both surface-adsorbate and adsorbate-adsorbate
interactions need to be taken into account.3 For instance,
designing the correct interactions between a molecule and a
surface can be critical in creating ordered surface assemblies.
Thus, designing the appropriate surface-adsorbate interactions
can be a powerful tool in controlling the nature of the molecular
surface assembly. For example, long alkyl chains are known to
the supramolecular interactions (such as hydrogen bonding,
alkyl-alkyl interactions, etc.) between the molecules (adsor-
bates) can also be employed to tune the nature of the resulting
stable, ordered surface assemblies. By manipulating both
surface-adsorbate and adsorbate-adsorbate interactions, a
number of groups have studied the potential of physisorbed
compounds to facilitate molecular assembly on ordered sur-
faces.2,6,7
One class of supramolecular motifs that has received attention
in the area of self-assembly over the years is the nucleobases.8,9
Of particular interest are guanine derivatives, which have the
(4) (a) Leunissen, M. E.; Graswinckel, W. S.; van Enckevort, W. J. P.; Vlieg,
E. Cryst. Growth Des. 2003, 4, 361-367. (b) Stepanow, S.; Lin, N.; Barth,
J. V.; Kern, K. Chem. Commun. 2006, 2153-2155.
† Center for Cardiovascular Biomaterials, Department of Biomedical
Engineering.
(5) Groszek, A. J. Proc. R. Soc. London, A 1970, 314, 473-498.
(6) (a) Barth, J. V.; Weckesser, J.; Cai, C.; Gu¨nter, P.; Bu¨rgi, L.; Jeandupeux,
O.; Kern, K. Angew. Chem., Int. Ed. 2000, 39, 1230-1234. (b) Tao, F.;
Bernasek, S. L. J. Am. Chem. Soc. 2005, 127, 12750-12751. (c) Griessl,
S. J. H.; Lackinger, M.; Jamitzky, F.; Markert, T.; Hietschold, M.; Heckl,
W. M. Langmuir 2004, 20, 9403-9407. (d) Kim, K.; Matzger, A. J. J.
Am. Chem. Soc. 2002, 124, 8772-8773. (e) Qian, P.; Nanjo, H.; Yokoyama,
T.; Suzuki, T. M. Chem. Commun. 1999, 1197-1198. (f) Nath, K. G.;
Ivasenko, O.; Miwa, J. A.; Dang, H.; Wuest, J. D.; Nanci, A.; Perepichka,
S. F.; Rosei, F. J. Am. Chem. Soc. 2006, 128, 4212-4213. (g) Hipps, K.
W.; Scudiero, L.; Barlow, D. E.; Cooke, M. P. Jr. J. Am. Chem. Soc. 2002,
124, 2126-2127. (h) Theobald, J. A.; Oxtoby, N. S.; Phillips, M. A.;
Champness, N. R.; Beton, P. H. Nature 2003, 424, 1029-1031. (i) Canas-
Ventura, M. E.; Xiao, W.; Wasserfallen, D.; Mu¨llen, K.; Brune, H.; Barth,
J. V.; Fasel, R. Angew. Chem., Int. Ed. 2007, 46, 1814-1818. (j) Baker,
R. T.; Mougous, J. D.; Brackley, A.; Patrick, D. L. Langmuir 1999, 15,
4884-4891. (k) Stevens, F.; Beebe, T. P., Jr. Langmuir 1999, 15, 6884-
6889. (l) Hecht, S. Angew. Chem., Int. Ed. 2003, 42, 24-26. (m) Yablon,
D. G.; Ertas, D.; Fang, H.; Flynn, G. W. Isr. J. Chem. 2003, 43, 383-392.
(n) Cox, J. K.; Eisenberg, A.; Lennox, R. B. Curr. Opin. Colloid Interface
Sci. 1999, 4, 52-59. (o) Gong, J.-R.; Yan, H.-J.; Yuan, Q.-H.; Xu, L.-P.;
Bo, Z.-S.; Wan, L.-J. J. Am. Chem. Soc. 2006, 128, 12384-12385.
‡ Department of Macromolecular Science and Engineering.
§ Department of Chemistry.
(1) (a) Ryu, D. Y.; Shin, K.; Drockenmuller, E.; Hawker, C. J.; Russell, T. P.
Science 2005, 308, 236-239. (b) Michl, J.; Magnera, T. F. Proc. Natl.
Acad. Sci. U.S.A. 2002, 99, 4788-4792. (c) Scho¨nherr, H.; Paraschiv, V.;
Zapotoczny, S.; Crego-Calama, M.; Timmerman, P.; Frank, C. W.; Vancso,
G. J.; Reinhoudt, D. N. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 5024-
5027. (d) Russell, T. P. Science 2002, 297, 964-967. (e) Lecle`re, P.; Surin,
M.; Jonkheijm, P.; Henze, O.; Schenning, A. P. H. J.; Biscarini, F.;
Grimsdale, A. C.; Feast, W. J.; Meijer, E. W.; Mu¨llen, K.; Bre´das, J.-L.;
Lazzaroni, R. Eur. Polym. J. 2004, 40, 885-892.
(2) (a) Palermo, V.; Liscio, A.; Gentilini, D.; Nolde, F.; Mu¨llen, K.; Samor`ı,
P. Small 2007, 3, 161-167. (b) Puigmart´ı-Luis, J.; Minoia, A.; Uji-i, H.;
Rovira, C.; Cornil, J.; De Feyter, S.; Lazzaroni, R.; Amabilino, D. B. J.
Am. Chem. Soc. 2006, 128, 12602-12603. (c) Xu, H.; Norsten, T. B.; Uzun,
O.; Jeoung, E.; Rotello, V. M., Chem. Commun. 2005, 5157-5159.
(3) (a) Barth, J. V.; Costantini, G.; Kern, K. Nature 2005, 437, 671-679. (b)
Badin, M. G.; Bashir, A.; Krakert, S.; Strunskus, T.; Terfort, A.; Wo¨ll, C.
Angew. Chem., Int. Ed. 2007, 46, 3762-3764. (c) Li, S.-S.; Yan, H.-J.;
Wan, L.-J.; Yang, H.-B.; Northrop, B. H.; Stang, P. J. J. Am. Chem. Soc.
2007, 129, 9268-9269.
9
1466
J. AM. CHEM. SOC. 2008, 130, 1466-1476
10.1021/ja0775927 CCC: $40.75 © 2008 American Chemical Society

Nanoscale Organization of Functional Groups on HOPG
A R T I C L E S
The thromboresistance of biomaterial coatings is determined
by its interaction with plasma proteins and platelets.14 The most
abundant plasma proteins, albumin, immunoglobulin, and
fibrinogen, constitute the main adsorbed components on hy-
drophobic surfaces within microseconds of contact with blood.15
Typically, over a period of time, an exchange process (Vroman
effect) results in an increase in the proportion of thrombogenic
proteins such as fibrinogen, kininogen, and factor XII on the
surface.16 The hydrophobic nature of the surface causes adsorbed
proteins to denature, exposing peptide residues such as RGD
that bind to specific platelet receptors such as GPIIb/IIIa (CD41/
CD61). This sequence of events begins surface-induced throm-
bosis and causes many significant complications in implanted
devices.17 Healthy blood vessels provide a model to understand
the prevention of surface-induced thrombosis. The glycocalyx
on endothelial cells that line the lumen of blood vessels presents
membrane proteins and carbohydrates that actively prevent
thrombosis by creating a hydrated layer18 at the surface which
acts to reduce nonspecific protein adsorption, protein denatur-
ation, and platelet adhesion. Thus, coatings that present groups
that can mimic the hydrated layer most likely will have
improved thromboresistance and blood biocompatibility.
Figure 1. Concept of organized functional groups arranged through surface
supramolecular polymerization. Monomers are initially in solution followed
by adsorption and assembly to form linear band structures that present side
groups in an ordered array on a hydrophobic surface.
ability to interact with themselves or with complementary motifs
through the use of Watson-Crick binding and/or Hoogsteen
binding. For example, when guanine and its derivatives are
adsorbed on a graphite surface, the resulting aggregates can form
ordered tapes or monolayers that can make use of guanine-
guanine interactions,10,11 in addition to surface-adsorbate
interactions.
In recent years, a number of groups have investigated the
potential of using oligonucleotides to self-assemble two-
dimensional (2D) arrays and to use those as scaffolds to display
a range of functionalities.12 Recently, we initiated a program
aimed at investigating the potential of assembling supramo-
lecular polymers, derived from low-molecular weight nucleo-
base-endcapped monomers, on a surface13 as a way to organize
functional groups at the nanoscale (Figure 1) and as such act
as molecular-scale surface scaffolds. In this manuscript, we
report for the first time the realization of this concept. Using
triethylene glycol monomethyl ether (TEG) groups attached to
low-molecular weight monomers, these supramolecular scaffold-
organized TEG surfaces exhibit reduced protein absorption and
platelet adhesion. To further back up the nature and mechanism
of this scaffold assembly a series of model compounds (without
the TEG units) have been prepared and studied, allowing a better
understanding of how these systems assemble on highly oriented
pyrolytic graphite (HOPG).
A well-known strategy for creating protein resistant surfaces
involves utilizing surface-bound polyethylene glycol (PEG)
chains.19-22 While a general theory of how PEG prevents protein
adsorption is still being developed, it is currently thought that
a hydrated PEG chain provides both osmotic (solvation of PEG
chains) and entropic (conformational entropy of PEG chains)
penalties that can be large enough to disfavor the hydrophobic
interaction between proteins and the hydrophobic surface.19,20c,23
Thus, in order for surface-bound PEG to work effectively to
prevent protein adhesion it needs to be solvated with water
molecules. Modeling studies suggest that when surface-bound
PEG is hydrated it preferentially forms a random coil “helical”
structure (gauche-trans-gauche).18,23b,24 Thus, it has been
proposed that PEG needs to be able to form these helical
conformations in order to hydrate effectively and therefore
prevent protein adhesion.19,23b For example, proteins are repelled
when grafted PEG chains form helical conformations on Au
(14) (a) Horbett, T. A. CardioVasc. Pathol. 1993, 2, 137s-148s. (b) Sharma,
C. P. J. Biomater. Appl. 2001, 15, 359-381.
(7) (a) De Feyter, S.; De Schryver, F. C. Chem. Soc. ReV. 2003, 32, 139-150.
(b) De Feyter, S.; Larsson, M.; Schuurmans, N.; Verkuijl, B.; Zoriniants,
G.; Gesquie`re, A.; Abdel-Mottaleb, M. M.; Esch, J. v.; Feringa, B. L.; van
Stam, J.; De Schryver, F. C. Chem. Eur. J. 2003, 9, 1198-1206. (c)
Gesquie`re, A.; Jonkheijm, P.; Hoeben, F. J. M.; Schenning, A. P. H. J.; De
Feyter, S.; De Schryver, F. C.; Meijer, E. W. Nano Lett. 2004, 4, 1175-
1179. (d) Furukawa, S.; Tahara, K.; De Schryver, F. C.; Van der Auweraer,
M.; Tobe, Y.; De Feyter, S. Angew. Chem., Int. Ed. 2007, 46, 2831-2834.
(e) Mamdouh, W.; Uji-i, H.; Ladislaw, J. S.; Dulcey, A. E.; Percec, V.; De
Schryver, F. C.; De Feyter, S. J. Am. Chem. Soc. 2006, 128, 317-325.
(8) (a) Seeman, N. C. Trends Biotechnol. 1999, 17, 437-443. (b) Li, Y.; Tseng,
Y. D.; Kwon, S. Y.; D’Espaux, L.; Bunch, J. S.; McEuen, P. L.; Luo, D.
Nat. Mater. 2004, 3, 38-42. (c) Xu, S.; Dong, M.; Rauls, E.; Otero, R.;
Linderoth, T. R.; Besenbacher, F. Nano Lett. 2006, 6, 1434-1438.
(9) Sivakova, S.; Rowan, S. J. Chem. Soc. ReV. 2005, 34, 9-21.
(10) (a) Sowerby, S. J.; Edelwirth, M.; Heckl, W. M. J. Phys. Chem. B 1998,
102, 5914-5922. (b) Kelly, R. E. A.; Kantorovich, L. N. J. Mater. Chem.
2006, 16, 1894-1905.
(15) Swalen, J. D.; Allara, D. L.; Andrade, J. D.; Chandross, E. A.; Garoff, S.;
Israelachvili, J.; McCarthy, T. J.; Murray, R.; Pease, R. F.; Rabolt, J. F.;
Wynne, K. J.; Yu, H. Langmuir 1987, 3, 932-950.
(16) (a) Vroman, L.; Adams, A. L. Surf. Sci. 1969, 16, 438-446. (b) Hudson,
S. D.; Jung, H.-T.; Percec, V.; Cho, W.-D.; Johansson, G.; Ungar, G.;
Balagurusamy, V. S. K. Science 1997, 278, 449-452.
(17) Ratner, B. D. J. Biomed. Mater. Res. 1993, 27, 283-287.
(18) Vogel, E. A. AdV. Colloid Interface Sci. 1998, 74, 69-117.
(19) Lee, J. H.; Lee, H. B.; Andrade, J. D. Prog. Polym. Sci. 1995, 20, 1043-
1079.
(20) (a) Prime, K. L.; Whitesides, G. M. Science 1991, 252, 1164-1167. (b)
Prime, K. L.; Whitesides, G. M. J. Am. Chem. Soc. 1993, 115, 10714-
10721. (c) Sigal, G. B.; Mrksich, M.; Whitesides, G. M. J. Am. Chem.
Soc. 1998, 120, 3464-3473.
(21) (a) Malmsten, M.; Emoto, K.; Van, Alstine, J. M. J. Colloid Interface Sci.
1998, 202, 507-517. (b) Vadgama, P. Annu. Rep. Prog. Chem., Sect. C:
Phys. Chem. 2005, 101, 14-52. (c) Tokumitsu, S.; Liebich, A.; Herrwerth,
S.; Eck, W.; Himmelhaus, M.; Grunze, M. Langmuir 2002, 18, 8862-
8870. (d) Capadona, J. R.; Collard, D. M.; Garc´ıa, A. J. Langmuir 2003,
19, 1847-1852. (e) Snellings, G. M. B. F.; Vansteenkiste, S. O.; Corneillie,
S. I.; Davies, M. C.; Schacht, E. H. AdV. Mater. 2000, 12, 1959-1962.
(22) Harris, J. M. Poly(ethylene glycol) Chemistry: Biotechnical and Biomedical
Applications; Plenum: New York, 1992.
(11) Gottarelli, G.; Masiero, S.; Mezzina, E.; Pieraccini, S.; Rabe, J. P. Chem.
Eur. J. 2000, 6, 3242-3248.
(12) (a) Winfree, E.; Liu, F.; Wenzler, L. A.; Seeman, N. C. Nature 1998, 394,
539-544. (b) Park, S. H.; Yin, P.; Liu, Y.; Reif, J. H.; LaBean, T. H.;
Yan, H. Nano Lett. 2005, 5, 729-733. (c) Zheng, J.; Constantinou, P. E.;
Micheel, C.; Alivisatos, A. P.; Kiehl, R. A.; Seeman, N. C. Nano Lett.
2006, 6, 1502-1504. (d) Garibotti, A. V.; Knudsen, S. M.; Ellington, A.
D.; Seeman, N. C. Nano Lett. 2006, 6, 1505-1507. (e) Ding, B.; Seeman,
N. C. Science 2006, 314, 1583-1585. (f) Lin, C.; Liu, Y.; Yan, H. Nano
Lett. 2007, 7, 507-512.
(23) (a) Wang, R. Y.; Himmelhaus, M.; Fick, J.; Herrwerth, S.; Eck, W.; Grunze,
M. J. Chem. Phys. 2005, 122, 164702.1-164702.6. (b) Wang, R. L. C.;
Ju¨rgen Kreuzer, H.; Grunze, M. Phys. Chem. Chem. Phys. 2000, 2, 3613-
2622.
(24) Wang, R. L. C.; Kreuzer, H. J.; Grunze, M. J. Phys. Chem. B 1997, 101,
9767-9773.
(13) Kumar, A. M. S.; Sivakova, S.; Marchant, R. E.; Rowan, S. J. Small 2007,
3, 783-787.
9
J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008 1467

A R T I C L E S
Kumar et al.
surfaces at high density.25 However, on Ag surfaces, PEG chains
are much more densely packed, forcing the chains into an
extended all trans conformation, preventing hydration, which
in turn allows protein adsorption. At the other extreme, low
densities of surface-bound PEG cannot create a uniform hydrated
layer and will not prevent protein adsorption. Thus, controlling
both the density and conformation of presented PEG chains is
important in creating a coating that repels proteins. Covalent
grafting of PEG, including the use of PEG-functionalized
SAMs,26 has been extensively utilized by a number of groups
to produce surfaces with different densities of PEG.27 Addition-
ally, a range of processing steps has allowed grafting on surfaces
that do not normally have the proper chemical functionality for
covalent grafting.26b,27a,28,29 An alternative approach of surface
attachment uses physisorbed assemblies, including Langmuir-
Blodgett films,15,30 that present PEG from the surface.31 In
particular, one of us has used surfactant-like polymers to
assemble on a surface and display functional groups.32 While
initial studies were carried out on HOPG, this approach was
found to work on a variety of hydrophobic surfaces, allowing
presentation of biofunctional groups that in turn can influence
biological processes. However, even though high densities of
functional groups have been obtained, precisely controlling the
surface density and lateral placement of functional groups has
remained a challenge. A tunable, supramolecular surface scaffold
that presents PEG chains could more precisely control surface
density and could allow more accurate mechanistic studies of
protein adsorption on PEG surfaces to be conducted.
Scheme 1. Synthesis of G1nG (n ) 8, 10, 12, 18) and
Deprotection to Generate G2nG (n ) 12, 18)a
a These ditopic monomers have a -(CH2)n- hydrocarbon core flanked
by guanine end groups (R).
monomers, consisting of either C12 or C18 linear alkyl chains
with guanine peptide nucleic acid (PNA) end groups could result
in the formation molecular-sized bands on HOPG when ad-
sorbed from a water/DMSO solution. By building on this
preliminary work and with the goal of further understanding
the nature and mechanism of the surface self-assembly pro-
cesses, we designed and synthesized a series of simple model
compounds (G1nG). These model monomers comprise three
components, (1) a hydrocarbon core with n () 8, 10, 12, 18)
methylene groups, to enhance adsorption onto a hydrophobic
surface in the presence of an aqueous medium, (2) the guanine
end groups, to facilitate adsorbate-adsorbate interactions
through hydrogen bonding, and (3) peptide nucleic acid (PNA)33
chains primarily used to link the hydrocarbon cores and the
guanine moieties. These monomers were prepared by reacting
commercially available benzyl carbamate (Cbz)-protected gua-
nine-PNA groups with 1,8-diaminooctane (n ) 8), 1,10-
diaminodecane (n ) 10), 1,12-diaminododecane (n ) 12), or
1,18-diaminooctadecane (n ) 18) using standard peptide
coupling conditions to produce G(Cbz)1nG(Cbz) (Scheme 1).
The guanine moieties were then deprotected using H2 gas with
a Pd/C catalyst to produce G1nG (15-25% overall yield). The
structures for the materials were confirmed by NMR and
MALDI-MS.
Results and Discussion
In order to create self-assembling molecular scaffolds, which
are able to present functional groups from a surface, a better
understanding of what role the different structural components
within these monomers play in the surface self-assembly is
required. In a previous communication,13 we showed that ditopic
(25) (a) Harder, P.; Grunze, M.; Dahint, R.; Whitesides, G. M.; Laibinis, P. E.
J. Phys. Chem. B 1998, 102, 426-436. (b) Feldman, K.; Ha¨hner, G.;
Spencer, N. D.; Harder, P.; Grunze, M. J. Am. Chem. Soc. 1999, 121,
10134-10141.
(26) (a) Credo, G. M.; Boal, A. K.; Das, K.; Galow, T. H.; Rotello, V. M.;
Feldheim, D. L.; Gorman, C. B. J. Am. Chem. Soc. 2002, 124, 9036-
9037. (b) Zhu, X.-Y.; Jun, Y.; Staarup, D. R.; Major, R. C.; Danielson, S.;
Boiadjiev, V.; Gladfelter, W. L.; Bunker, B. C.; Guo, A. Langmuir 2001,
17, 7798-7803. (c) Sanyal, A.; Norsten, T. B.; Uzun, O.; Rotello, V. M.
Langmuir 2004, 20, 5958-5964. (d) Zou, S.; Zhang, Z.; Fo¨rch, R.; Knoll,
W.; Scho¨nherr, H.; Vancso, G. J. Langmuir 2003, 19, 8618-8621. (e)
Garcia-Lopez, J. J.; Zapotoczny, S.; Timmerman, P.; van Veggel, F. C. J.
M.; Vancso, G. J.; Crego-Calama, M.; Reinhoudt, D. N. Chem. Commun.
2003, 352-353.
As the monomers were not water soluble, surface self-
assembly experiments were carried out by introducing a small
amount (ca. 5 µL) of a DMSO monomer solution (0.94-5.4
mM, 1-5 µg/mL) into a water droplet (ca. 20 µL) on HOPG.
A 1 mL fluid cell was then placed on top of the HOPG, and
images of the surface were captured using a fluid tapping AFM
setup at ambient temperature. All the monomers adsorbed on
the surface within seconds, as measured by a change in surface
roughness relative to bare HOPG, but no order was observed.
However, upon addition of enough water to fill the 1 mL cell
(yielding an overall concentration of the monomers of about
4.7-27 nM and diluting the DMSO to ca. 0.5%), two sets of
linear band patterns were observed in all samples: a molecular-
sized set of bands with widths <5 nm and a larger set of bands
(27) (a) Jo, S.; Park, K. Biomaterials 2000, 21, 605-616. (b) Papra, A.;
Gadegaard, N.; Larsen, N. B. Langmuir 2001, 17, 1457-1460. (c) Yang,
Z.; Galloway, J. A.; Yu, H. Langmuir 1999, 15, 8405-8411.
(28) Chen, H.; Zhang, Z.; Chen, Y.; Brook, M. A.; Sheardown, H. Biomaterials
2005, 26, 2391-2399.
(29) Ghosh, P.; Amirpour, M. L.; Lackowski, W. M.; Pishko, M. V.; Crooks,
R. M. Angew. Chem., Int. Ed. 1999, 38, 1592-1595.
(30) (a) Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir-
Blodgett to Self-Assembly; Academic Press: San Diego, 1991. (b) Peng, J.
B.; Barnes, G. T.; Gentle, I. R. AdV. Colloid Interface Sci. 2001, 91, 163-
219.
(31) (a) Vacheethasanee, K.; Wang, S.; Qiu, Y.; Marchant, R. E. J. Biomater.
Sci. Polym. Ed. 2004, 15, 95-110. (b) Tae, G.; Lammertink, R. G. H.;
Kornfield, J. A.; Hubbell, J. A. AdV. Mater. 2003, 15, 66-69. (c) Kim, K.;
Kim, C.; Byun, Y. J. Biomed. Mater. Res. 2000, 52, 836-840.
(32) (a) Holland, N. B.; Qiu, Y.; Ruegsegger, M.; Marchant, R. E. Nature 1998,
392, 799-801. (b) Ruegsegger, M. A.; Marchant, R. E. J. Biomed. Mater.
Res. 2001, 56, 159-167. (c) Sagnella, S.; Anderson, E.; Sanabria, N.;
Marchant, R. E.; Kottke-Marchant, K. Tissue Engin. 2005, 11, 226-236.
(d) Larsen, C. C.; Kligman, F.; Kottke-Marchant, K.; Marchant, R. E.
Biomaterials 2006, 27, 4846-4855. (e) Larsen, C. C.; Kligman, F.; Tang,
C.; Kottke-Marchant, K.; Marchant, R. E. Biomaterials 2007, 28, 3537-
3548.
(33) Nielson, P. E. Acc. Chem. Res. 1999, 32, 624-630.
9
1468 J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008

Nanoscale Organization of Functional Groups on HOPG
A R T I C L E S
Figure 3. (a) AFM phase image of G112G domains offset by 60° (angular
measurement). Domain boundary defects exist where the different domains
meet (arrow). The angular offset of the domains suggests epitaxial assembly
along the (b) symmetric axes of the HOPG lattice. (c) 2D Fourier transform
of AFM image showing epitaxial organization of domains.
small as 50 nm and as large as 2 µm (see Supporting
Information). The linear bands in each domain were offset from
each other by a multiple of 60°, suggesting epitaxial alignment
of the molecules along the three axes of symmetry in the HOPG
lattice (Figure 3). Between the majority of the domains,
amorphous boundary regions (arrow in Figure 3a) are observed
where the linear bands do not connect and the assembly cannot
continue growing because of interference from adjacent do-
mains. Additionally, the presence of multiple domains suggests
that there are multiple sites of nucleation.
Figure 2. AFM phase images of (a) G18G, (b) G110G, (c) G112G, and (d)
G118G molecular-sized bands on HOPG. Band spacing increases with
increasing length of the hydrocarbon core.
The focus of this current research is the creation of tunable
self-assembling molecular scaffolds. Thus, the fact that tailoring
the molecular structure allowed the width of molecular-sized
bands to be tuned suggests that these assemblies, rather than
the larger band assemblies, are potential candidates for molecular
scaffolds. Interestingly, as the length of the hydrocarbon core
in the monomers increased, the prevalence of the larger bands
in the surface assemblies decreased. For instance, G18G and
G110G assemblies were composed mostly of larger bands with
very few domains of molecular-sized bands. However, molec-
ular-sized bands were found to occupy a significantly larger
proportion of the surface in G112G coatings and were even more
ubiquitous in G118G coatings. As a result, the latter two
monomers are better suited to create tunable scaffold coatings,
and thus, these monomers became the major focus of this
research.
On the basis of the molecular design of G1nG, it was expected
that hydrogen bonding, most likely through guanine-guanine
interactions, would play a role in the surface assembly. Thus,
the surface assembly of two controls, a monomer with reduced
hydrogen-bonding capability (G(Cbz)112G(Cbz)) and 1,12-
diaminododecane, were characterized. G(Cbz)112G(Cbz) is a
precursor to G112G and is protected at the guanine C(2) exo-
amino site by a Cbz group, resulting in greatly reduced Watson-
Crick hydrogen-bond interactions. Fluid tapping mode imaging
of HOPG coated with G(Cbz)112G(Cbz) revealed only the large
bands (5-7 nm widths) and amorphous domain boundary
regions, with no sign of the molecular-sized bands. This suggests
that blocking the Watson-Crick face of the guanine greatly
reduces the tendency of G(Cbz)112G(Cbz) to self-assemble into
Table 1. Summary of Observed and Modeled Resultsa from
Single Monomer Studies
M
L (nm)
Wobs (nm)
Wmod (nm)
G18G
G110G
G112G
G118G
G212G
G218G
G3G
3.6
3.9
4.2
4.8
4.2
4.8
3.4
3.2 ( 0.1
3.5 ( 0.1
3.8 ( 0.1
4.8 ( 0.1
3.8 ( 0.2
4.5 ( 0.2
2.5 ( 0.1
3.4 ( 0.1
3.8 ( 0.1
3.4
3.6
3.8
4.9
3.8
4.4
2.5
3.4
3.8
G4G
4.4
a Summary of end-to-end length (L) of monomers (M) in extended
conformation (from molecular modeling), width of observed bands (Wobs
on HOPG from AFM phase images, and width of modeled bands (Wmod
in energy-minimized models.
)
)
with widths between 5 and 7 nm. The widths of the molecular-
sized bands were dependent on the length of the hydrocarbon
core in the assembling monomer. For example, G18G, G110G,
G112G, and G118G had widths of 3.2 ( 0.1, 3.5 ( 0.1, 3.8 (
0.1, and 4.8 ( 0.1 nm, respectively (Figure 2a-d, Table 1) as
measured by power spectrum and cross-sectional analyses. Thus,
the band widths of molecular-sized bands in the G1nG as-
semblies can be simply tuned by changing the hydrocarbon
length (n). On the other hand, the widths of the larger bands,
which showed more variability (multiple widths between 5 and
7 nm are observed) within the same monomer experiment,
showed no correlation with the size of the monomers. These
larger assemblies form distinctly separate domains from as-
semblies composed of molecular-sized bands. To date we have
not been able to identify the exact nature of these “larger” bands,
although it is possible that they consist either of multilayers or
hemimicellar assemblies.34
(34) (a) Wanless, E. J.; Davey, T. W.; Ducker, W. A. Langmuir. 1997, 13, 4223-
4228. (b) Nishimura, S.; Scales, P. J.; Biggs, S. R.; Healy, T. W. Colloid
Surf. A. 1995, 103, 289-298. (c) Manne, S.; Cleveland, J. P.; Gaub, H. E.;
Stucky, G. D.; Hansma, P. K. Langmuir. 1994, 10, 4409-4413. (d) Fuhrhop,
J. H.; Wang, T. Chem. ReV. 2004, 104, 2901-2937.
Images captured at larger scan sizes (>500 nm) showed that
there were many linear band assemblies that spanned areas as
9
J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008 1469

A R T I C L E S
Kumar et al.
stable molecular-sized bands. The second control, 1,12-diami-
nododecane, adsorbed on the HOPG surface; however, it also
did not form the molecular-sized bands, under the conditions
described previously, suggesting the importance of having
guanine end groups for the desired tunable band spacings
observed in the G1nG assemblies.
The above results suggest that the hydrogen-bonding ability
of the guanine is important to the formation of the molecular-
sized bands. The question then becomes how the guanine is
self-assembling on the surface. From the literature it is known
that guanine and its 9-substitued derivatives form linear tapelike
structures, in solution,9,11,35 the solid state,9,11,36 and on sur-
faces.9,10,11,37 Two major guanine motifs have been identified/
proposed. The first is a non-dipole guanine motif (Figure 4a),
whichhasbeenobservedinSTMimagesofguaninemonolayers10a
on graphite and proposed in the crystal structures of some
guanine derivatives.38 This guanine motif forms centrosymmetric
repeat units, centered between the two guanines (G1 and G2 in
Figure 4a), along the hydrogen-bonded chain. As a result of
this symmetry there is no net dipole in this motif. When the R
group on N(9) is bulky or in the crystal structures of some
guanine derivatives,38 the guanine moieties can instead form a
different guanine tape motif (Figure 4b), presumably, on account
of the steric repulsions between adjacent R groups.10b,11 In this
case the motif is no longer centrosymmetric, and this results in
a net dipole along the entire structure (arrow in Figure 4b). These
two known guanine motifs became the starting point for
molecular modeling studies with the goal of developing a better
understanding of how the G1nG molecules may be arranging
within the molecular-sized bands.
Molecular modeling of all the G1nG monomers, using the
consistent valence forcefield (CVFF, class I forcefield) in
DISCOVER, were carried out on a modeled graphene surface
and in a simulated aqueous environment (distance-dependent
dielectric ꢀr ) 80). The first sets of modeling experiments were
carried out on G112G using either of the two guanine motifs
shown in a and b of Figure 4 (see Supporting Information).
However, the resulting modeled assemblies suggested the
formation of bands with widths of ca. 3.4 nm that did not match
observed 3.8 nm bands measured by AFM. Thus, other possible
guanine motifs were investigated. Interestingly, a double-
stranded guanine tape motif (Figure 4c) was found that resulted
in assemblies with modeled band widths that more correctly
predicted the observed 3.8 nm bands (Figure 5a) and was of
lower calculated energy than the models of either of the
previously described motifs. This centrosymmetric double-
stranded motif is composed of guanine dimers formed through
the Watson-Crick faces of two guanine moieties. This dimeric
Figure 4. (a) Non-dipole guanine motif with centrosymmetric interactions
(i.e., center of symmetry between G1 and G2 indicated by dot), (b) dipole
tape motif with net dipole (arrow), and (c) dimer motif with C(2) exo-
amino hydrogens (*) separated by 3 Å.
motif is extended into a tape through additional nucleobase
hydrogen bonding through the exo-NH2 and N(7) on the
Hoogsteen face of adjacent guanine dimer moieties. While two
of the exo-amino hydrogens (* in Figure 4c) are in close
proximity, modeling indicates that these atoms are separated
by 3 Å with no van der Waals overlap.
It is possible that the PNA-Boc groups within the G112G
assemblies sterically hinder the formation of the guanine motif
shown in Figure 4a. There are potentially a number of reasons
why the double-stranded motif appears to be more energetically
favored over the motif in Figure 4b. These include better packing
efficiency on the surface, enhanced surface-absorbate interac-
tions, and the lack of a dipole in the double-stranded guanine
tape.
Models of the G18G and G110G assemblies using this double-
stranded guanine motif also showed lower modeled energies
than the models using the motifs in a and b of Figure 4. In
addition, the models also predicted the banding patterns to be
3.4 and 3.6 nm, respectively, which match much better with
the observed AFM patterns (3.2 and 3.5 ( 0.1 nm, respectively)
than the patterns predicted by the models using the other two
guanine motifs.
(35) (a) Gottarelli, G.; Masiero, S.; Mezzina, E.; Spada, G. P. HelV. Chim. Acta
1998, 81, 2078-2092. (b) Giorgi, T.; Grepioni, F.; Manet, I.; Mariani, P.;
Masiero, S.; Mezzina, E.; Pieraccini, S.; Saturni, L.; Spada, G. P.; Gottarelli,
G. Chem. Eur. J. 2002, 8, 2143-2152. (c) Yoshikawa, I.; Yanagi, S.;
Yamaji, Y.; Araki, K. Tetrahedron 2007, 63, 7474-7481.
(36) (a) Forman, S. L.; Fettinger, J. C.; Pieraccini, S.; Gottarelli, G.; David, J.
T. J. Am. Chem. Soc. 2000, 122, 4060-4067. (b) Thewalt, U.; Bugg, C.
E.; Marsh, R. E. Acta. Crystallogr. 1971, B27, 2358-2363. (c) Calzolari,
A.; Di Felice, R.; Molinari, E.; Garbesi, A. Physica E 2002, 13, 1236-
1239.
(37) (a) Srinivasan, R.; Murphy, J. C.; Fainchtein, R.; Patabiraman, N. J.
Electroanal. Chem. Interfacial Electrochem. 1991, 312, 293-300. (b) Tao,
N. J.; Shi, Z. J. Phys. Chem. 1994, 98, 1464-1471. (c) Otero, R.; Scho¨ck,
M.; Molina, L. M.; Lœgsgaard, E.; Stensgaard, I.; Hammer, B.; Besen-
bacher, F. Angew. Chem., Int. Ed. 2005, 44, 2270-2275.
(38) (a) Thewalt, U.; Bugg, C. E.; Marsh, R. E. Acta Crystallogr. 1970, B26,
1089-1101. (b) Thewalt, U.; Bugg, C. E.; Marsh, R. E. Acta Crystallogr.
1971, B27, 2358-2363.
9
1470 J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008

Nanoscale Organization of Functional Groups on HOPG
A R T I C L E S
Figure 6. (a) Chemical structure of G3G. (b) AFM phase image of multiple
domains of G3G (“brighter” areas) forming molecular-sized bands with
widths of 2.5 nm (arrows) and 3.4 nm (circle). Surface coverage is
incomplete with surrounding “darker” regions having an amorphous phase.
protecting groups had been removed. If, as suggested above,
the Boc groups play a role in aiding the formation of the “open”
conformation in G118G, then it would be expected that removing
its Boc groups should result in a significant change in the band
spacing. However, if G112G is already in the “closed” confor-
mation as modeling suggests, then removal of the Boc groups
from this monomer would not significantly affect its assembly.
Standard Boc deprotection chemistry (trifluoroacetic acid) was
used to generate G212G from G112G and G218G from G118G.
The removal of the two hydrophobic Boc groups, not surpris-
ingly, makes G212G and G218G more water soluble. As such,
this has consequences for the surface assembly from aqueous
environments. A higher monomer concentration (11-17 µM,
in DMSO yielding an overall concentration of 55-85 nM after
addition of water) was needed before ordered assemblies on
the surface could be observed. Additionally, the bands were
less stable compared to G1nG monomer assemblies, making
the G2nG assemblies more prone to disruption by the AFM
probe. AFM experiments of G212G on HOPG showed molec-
ular-sized bands with 3.8 ( 0.2 nm repeats, the same widths
observed in G112G assemblies, indicating the Boc groups do
not significantly alter the spacing of the molecular-sized bands
in this system. Models confirm these findings, as there is no
change in band spacing or monomer arrangement once the Boc
groups are removed (see Supporting Information). On the other
hand, when G218G was assembled on HOPG, it was found to
form band spacings of 4.5 ( 0.2 nm. Interestingly, this
corresponds to a close-packed alkyl arrangement (see Supporting
Information). This suggests that the Boc groups in G118G play
a role in aiding the assembly of the C18 hydrocarbon core into
the open arrangement. Removal of the Boc groups did not seem
to affect the assembly of larger bands (5-7 nm) as they were
still observed.
Figure 5. Energy-minimized molecular models of (a) G112G in a close-
packed arrangement, (b) G118G in a close-packed arrangement, and (c)
G118G in an open hydrocarbon arrangement. Modeled band spacing (nm)
and unit cell (box) for each proposed arrangement is also shown.
In all of the previous models the hydrocarbons are close
packed, presumably to maximize packing efficiencies. Interest-
ingly, molecular modeling of G118G assemblies with close-
packed alkyl chains (Figure 5b) suggested a band spacing of
4.5 nm, which is significantly less than the observed 4.8 ( 0.1
nm by AFM. Further modeling studies found that a more open
hydrocarbon arrangement (Figure 5c) not only yielded a better
prediction (4.9 nm) of the observed band spacing but also was
lower in energy than the close-packed structure. Models of
G18G, G110G, and G112G in the open alkyl arrangement,
however, all had higher modeled energies than the models of
their alkyl close-packed assemblies. Furthermore, the predicted
band spacings from these open arrangements were much larger
than the experimentally observed ones. Thus, there seems to
be a change in the arrangement of hydrocarbons going from
small hydrocarbons (n ) 8, 10, 12) to the larger one (n ) 18).
The exact reason for this change in hydrocarbon arrangement
is not known; although based on the model (Figure 5c), it
appears that the Boc groups play an important role. In the open
arrangement of G118G the Boc group is adsorbed onto the
graphite surface, and there appears to be hydrogen bonding
between the amide N-H and carbamate CdO of adjacent
molecules. In the more close-packed arrangement (Figure 5b)
the Boc groups sit on top of the molecules, and there is no
amide-carbamate hydrogen bonding.
Thus, to further investigate the influence of the PNA-Boc
linker group on the surface self-assembly a number of different
ditopic guanine monomer derivatives were studied. A first set
of experiments were carried out on monomers in which the Boc
To examine the effect that the PNA linker group has on the
assembly of these systems G3G (Figure 6a), which is simply a
9
J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008 1471

A R T I C L E S
Kumar et al.
ditopic guanine endcapped dodecane with no PNA linker groups,
was designed and synthesized. These monomers were synthe-
sized in 23% overall yield by reacting guanine(Cbz) acetic acid39
with 1,12-diaminododecane, using standard peptide-coupling
conditions, followed by hydrogenation. Like G1nG, G3G
assembled into epitaxially aligned molecular-sized bands;
however, unlike G1nG, different molecular-sized banding spac-
ings, namely 2.5 ( 0.1 nm and 3.4 ( 0.1 nm, were observed
within different domains in the same experiment (Figure 6b).
It should be noted that, while some of the domains appear to
have offsets that are at slightly different angles (offset by ∼5°)
from normal epitaxial angles, this is most likely a consequence
of disruption from the AFM tip due to repeated scanning,
possibly suggesting that G3G domains are less stable than the
G1nG domains.
Molecular modeling was again used to help understand the
arrangement of the molecules within these molecular-sized band
assemblies. Assemblies of G3G were modeled using all three
surface guanine motifs outlined in Figure 4 (see Supporting
Information). A model using the guanine motif in Figure 4a
had bands that matched the 2.5 nm band spacing, and a model
using the double-stranded guanine motif (Figure 4c) had
modeled band widths that matched the observed 3.4 nm band
spacing with the latter having a lower energy. Not surprisingly,
it is difficult to model two or more different guanine motifs
within the same domain as the alignment of the molecules
changes because of the nature of the different guanine-guanine
interactions. Thus, it is not surprising that each domain of G3G
has only a single bandwidth and is not composed of both 2.5
and 3.4 nm bands. These experiments suggest that the PNA
linker in G1nG hinders the formation of the guanine tape in
Figure 4a, presumably on account of steric repulsions, thus only
allowing the system to assemble through one guanine motif,
namely the double-stranded assembly. This is an important
consideration if regular repeatable banding structures are targeted
for the surface scaffolds.
One of the strengths of our monomer scaffold design is,
potentially, the ability to tune the assembly by combining
differently sized monomers. For example, since our models
predicted that both G112G and G118G formed the double-
stranded guanine motif, the two different monomers would be
expected to assembly together within the same assembly.
However, modeling studies suggested structures that contained
both G112G and G118G within one band are not favored because
of the length mismatch (six methylene groups) of these
monomers. Thus, the only way to ensure that both of the chain-
end guanines are involved in hydrogen bonding is for the
different monomers (G112G and G118G) to separate into
different bands (Figure 7), resulting in a surface assembly which
contains two differently sized banding patterns within the same
domain. Since there is no hydrogen-bonding specificity between
the different monomers, the order of bands is expected to be
random. The relative surface concentration of two different
monomer species was also investigated, as it is well-known that
longer n-alkanes have a much larger surface-adsorbate interac-
tion with graphite compared to shorter n-alkanes.5 Therefore,
Figure 7. Concept of G118G (light) and G112G (dark) combined assembly
into molecular-sized bands. Molecules will favor phase separation on
account of mismatched guanines when they assemble in the same band.
it was expected that G118G would have a higher surface
concentration relative to G112G when the two sets of monomers
were deposited onto HOPG from a mixture of the two in
solution.
Using procedures similar to those described previously,
premixed monomer solutions with varying molar ratios of
G112G and G118G were used in surface-assembly experiments.
Four representative phase images (a-d of Figure 8) qualitatively
show the change in surface assembly that results as the molar
ratio of the monomers G118G/G112G is changed from 1:12 to
2:1. Nanophase separation of the two monomers is indeed
observed in all these cases, resulting in two different molecular-
sized band widths (3.9 ( 0.1 and 4.5 ( 0.1 nm most likely
correspond to bands formed by G112G and G118G, respectively)
that occur within the same domain. When compared to the
observed characteristic single monomer assemblies (Figure 2c,d),
the bandwidth of G112G in the combined assembly is within
experimental error of the observed 3.9 nm, while the assigned
4.5 nm is 0.3 nm shorter than the open arrangement (Figure
5c) G118G band spacing. However, the 4.5 nm band spacing
does match with the G118G assembly that has a close-packed
hydrocarbon arrangement (Figure 5b). This suggests that, in
combined assemblies of G112G and G118G, not only does the
difference in size of the hydrocarbon cores result in nanophase
separation of the molecules, but the arrangement of G112G also
seems to influence the packing arrangement of the hydrocarbons
in G118G. This then suggests a possible explanation for the 4.5
nm G118G bands. The presence of G112G molecules, which
adopt a close-packed hydrocarbon arrangement, induce the
G118G to adopt a similar close-packing structure that more
closely resembles the structural and geometric requirements of
the smaller assembly.
This leads to an interesting question, namely what happens
to the assemblies when a large excess of one of the monomers
is present? For example, Figure 9 shows a larger-area AFM
image of the assembly in which a small amount of G118G is
assembled with a large excess of G112G (ratio 1:12). In this
case regions of wider banding patterns (4.5 nm) are observed
within a smaller banding pattern (3.9 nm) matrix. This suggests
that, as there is not enough G118G to form a complete single
band, the G118G phase segregates within a G112G band,
presumably to minimize the energy penalties required to fit the
larger monomer into the smaller banding assembly. These are
regions of nanophase-separated G118G spanning up to 50 nm,
which from the modeled distances corresponds to a band
containing ca. 63 molecules.
(39) (a) Dueholm, K. L.; Egholm, M.; Behrens, C.; Christensen, L.; Hansen, H.
F.; Vulpius, T.; Petersen, K. H.; Berg, R. H.; Nielson, P. E.; Buchardt, O.
J. Org. Chem. 1994, 59, 5767-5773. (b) Thomson, S. A.; Josey, J. A.;
Cadilla, R.; Gaul, M. D.; Hassman, C. F.; Luzzio, M. J.; Pipe, A. J.; Reed,
K. L.; Ricca, D. J.; Wiethe, R. W.; Noble, S. A. Tetrahedron 1995, 51,
6179-6194.
At the other extreme, mixtures of the two monomers with
ratios skewed toward G118G, such as the 12:1 experiments
(Figure 8e,f), result in the self-assembly of molecular-sized
bands of 4.8 nm. This is consistent with the lower concentration
9
1472 J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008

Nanoscale Organization of Functional Groups on HOPG
A R T I C L E S
Figure 9. AFM phase image of 1:12 G118G/G112G surface assembly on
HOPG. Nanophase-separated regions of G118G are up to 50 nm. Bands
with both monomers produce band defects (box). There is also a type of
boundary domain defect (oval).
Figure 8. AFM phase images of two-monomer G118G/G112G assemblies
with solution molar ratios (a) 1:12, (b) 1:2, (c) 1:1, (d) 2:1, and (e,f) 12:1.
Surface concentration of monomer bands matches the solution concentration.
(e) G118G bands with minimal G112G produce bands of 4.8 nm. (f) G112G
(box) introduces defects into the G118G assembly. (g) The ratio of surface
mole fraction of G118G/surface mole fraction of G112G (Xsurf) vs the ratio
of solution mole fraction of G118G/solution mole fraction of G112G (Xsol
)
is plotted and shows a Langmuir-like correlation with a slope of 1.0.
Figure 10. AFM phase images of G118G coating captured at 50 °C showing
presence of linear band assemblies. (a) Coating assembled at room
temperature and then heated to 50 °C and (b) assembly created at 50 °C.
of G112G not having as much of an influence on the G118G
assembly. In regions where the G112G appears to be present
(box in Figure 8f), more disorder is observed than in regions
of the assembly that only seem to contain G118G (Figure 8e).
Quantitatively, the surface coverage of each monomer within
the assemblies of different monomer ratios was calculated and
plotted. Interestingly, there is a linear trend with a slope of 1.0
that correlates the ratio of the solution mole fractions to the
ratio of the surface mole fractions of the two monomers (Figure
8g) which suggests a Langmuir-like correlation.40 Even though
each methylene group that is commensurate with the HOPG
lattice generates 6.28 kJ/mol of free energy,5,41 there was no
observed preference for the surface adsorption of G118G over
G112G.13 As both monomers have two guanine PNA moieties,
this suggests that the guanine PNA-HOPG and the guanine-
guanine interactions are playing a significant role in the
assembly.
All studies discussed to date were carried out at 28 °C.
However, in order for these assemblies to be utilized in
bioenvironments, they have to be stable at physiological
temperatures. Thus, to probe the thermal stability of these
assemblies in more detail, imaging studies of the G118G coatings
were carried out at temperatures (50 °C) much higher than would
be required under physiological conditions. Linear band patterns
were observed with coatings that were assembled at room
temperature and imaged at 50 °C, an indication that these
nanoassemblies are stable at this temperature (Figure 10a). In
addition, studies were carried out to see if G118G would
(40) Masel, R. I. Principles of Adsorption and Reaction on Solid Surfaces; John
Wiley & Sons, Inc.: New York 1996; pp 235-254.
(41) Findenegg, G. H.; Liphard, M. Carbon 1987, 25, 119-128.
9
J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008 1473

A R T I C L E S
Kumar et al.
Scheme 2. Synthesis of G4G
Figure 11. Comparison of the AFM phase images of (a) G112G, forming
3.8 nm bands, and (b) G4G, forming 3.8 nm bands. The 3.8 nm widths of
both these assemblies suggests the molecules in both assemblies are
arranging similarly.
water solubility of G4G or perhaps that the tertiary amine, which
is presumably protonated at neutral pH, reduces the stability of
the surface-adsorbate interactions. After a few hours, however,
both epitaxial molecular-sized and larger bands were observed.
The molecular-sized bands had widths of 3.8 ( 0.1 nm and are
similar in width to the molecular-sized bands observed with
G112G (Figure 11a,b). As is observed in the G112G case, the
majority of the bands on the G4G-coated surface are the
molecular-sized ones.
From the studies mentioned above (e.g., Figures 2, 8, and
9), it is believed that the dark bands in the AFM phase images
correspond to the hydrocarbon segments and the lighter bands
correspond to the PNA-Boc-nucleobase segments. Thus, it was
expected that the TEG side group should be presented in the
center of the darker band in the images. However, no sign of
the TEG could be distinguished on AFM scans. This is possibly
a consequence of the TEG being too small and flexible relative
to any other part of the coating, thus hindering detection by the
AFM tip.
Molecular modeling of G4G (Figure 12b) using a double-
stranded motif and a close-packed arrangement similar to G112G
shows modeled bands of 3.9 nm that match the observed band
spacing. However, by adding an attachment point (the tertiary
nitrogen) for the TEG, the number of core atoms between
guanine PNA moieties changes from being even in G112G to
being odd in G4G. As a result, the molecules close pack in a
slightly different arrangement compared to G112G (Figure
12a,b). In this case, a pseudo-centrosymmetric assembly is
predicted for the G4G, in which adjacent double-stranded
guanine motifs run antiparallel with respect to each other,
whereas in G112G they run parallel. Indeed, there have been
many other reports of assemblies of small molecules that
produce different packing arrangements for even and odd core
atom parities.7b,d,43 Modeling also suggests that the TEG groups
are not large enough to completely cover the hydrophobic
scaffold coating. Calculations from the models suggest a density
of 0.32 TEG groups/nm2.
assemble at 50 °C. Thus, the HOPG surface with a water droplet
on it was heated to 50 °C; this was followed by addition of the
DMSO monomer solution to the water droplet before water (at
50 °C) was used to fill the fluid cell. Linear bands were again
observed (Figure 10b), indicating that the intermolecular
interactions that favor linear band assemblies are strong enough
to both nucleate and assemble at 50 °C.
With the goal of using these assemblies to create oligo-
(ethylene glycol)-covered surfaces, G4G, a monomer structur-
ally similar to G112G with a triethylene glycol monomethyl ether
(TEG) side chain was designed. The short-chain TEG was
chosen to demonstrate the concept of the scaffold coating in
part for ease of synthesis. In our proof-of-concept design, TEG
is anchored from a tertiary amine that is located at the center
of the hydrocarbon core that is flanked by two guanine PNA-
Boc groups. The synthesis of G4G is outlined in Scheme 2.
Iodo-TEG (see Supporting Information) was reacted with the
protected bis(hexamethylene) triamine42 (5) to yield 6. After
deprotection the resulting diamine (7) was converted into G4G,
in 10% overall unoptimized yield from 5, using procedures
similar to those used to access the G1nG monomers.
The assembling ability of G4G was investigated using the
procedures previously described; however, on account of the
increased water solubility of G4G, higher concentrations (8-
13 mM in DMSO) of the monomer were used. AFM fluid
tapping mode images showed that G4G adsorbed immediately
on HOPG as measured by a change in surface roughness. After
diluting the DMSO with water, it took several minutes to an
hour before linear bands were observed. Comparatively, G1nG
band assemblies were observed almost immediately after dilution
of DMSO with water. This may have to do with the greater
If G4G has assembled on the surface as indicated by the
models, then the HOPG has TEG groups supramolecularly
grafted on to it. Thus, if the assembly is stable under biologically
relevant conditions, this TEG coating of the HOPG should
impart some effect on protein adhesion to the surface. However,
(43) Fang, H.; Giancarlo, L. C.; Flynn, G. W. J. Phys. Chem. B 1998, 102,
7421-7424.
(42) Curphey, T. J. J. Org. Chem. 1979, 44, 2805-2807.
9
1474 J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008

Nanoscale Organization of Functional Groups on HOPG
A R T I C L E S
Figure 13. Epifluoresence optical microscopy (400 ×) images showing
static platelet adhesion on (a) bare HOPG, (b) G112G-coated HOPG, (c)
G4G-coated HOPG, (d,e) 6-coated HOPG. (f) Platelet surface coverage
(%) vs surfaces (n ) 5) demonstrating reduction in static platelet adhesion
to 20.8 ( 2.6% on G4G coating relative to bare HOPG (p ) 0.014) and
the G112G coating. (n ) 5, p ) 0.007). 6-coated HOPG had a large standard
deviation due to regions that adhere platelets (d) similar to bare HOPG and
(e) similar to G4G-coated HOPG.
Figure 12. Energy-minimized molecular models of (a) G112G, and (b)
G4G with modeled bands that match the observed band spacings. Change
in number of core atoms changes packing arrangement of molecules so
that adjacent guanine motifs run parallel in (a) and antiparallel in (b).
as mentioned before, the models suggest the TEG units do not
completely cover the hydrophobic surface; thus, close to
complete prevention of protein adhesion is not expected.
Nonetheless, if this material shows any effect in hindering
platelet adhesion, then derivatives which present an increased
density of functional units, either by using higher-molecular
weight PEG side chains or dendritic carbohydrate side chains,44
could result in more dramatic effects being observed. Thus,
initial studies to probe the biological effect of the current
assembly were performed using static platelet adhesion. Platelets
act indirectly as a marker for plasma protein adsorption and
are a critical step in surface-induced thrombosis. Thus, four
surfaces, bare HOPG, G112G-coated HOPG, G4G-coated
HOPG, and 6-coated HOPG, were prepared at room temperature
and incubated with platelet-rich plasma (PRP). The samples
were then fixed with paraformaldehyde, tagged with fluorescein
isothiocyanate (FITC)-labeled anti-CD41a, which binds to the
active site of the R subunit (GPIIb) of the GPIIb/IIIa platelet
receptor, and mounted on slides. Between each step, the surfaces
were washed three times with phosphate buffered saline (PBS)
in order to remove any nonadherent platelets. Figure 13a-e
shows representative epifluorescence optical microscopy images
(400×) of platelets adhering from the PRP solution to (a) bare
HOPG, (b) G112G-coated HOPG, (c) G4G-coated HOPG, and
(d,e) 6-coated HOPG. The percent surface coverage of the
platelets was calculated using a threshold analysis. Static platelet
adhesion on both the bare HOPG and the G112G coating were
statistically equivalent with 34.6 ( 5.6%, 37.6 ( 6.7% surface
platelet coverage, respectively (Figure 13f). These coatings are
both hydrophobic, and adhesion of a relatively large quantity
of platelets was expected. However, surface platelet coverage
on the G4G coating was 20.8 ( 2.6% which is statistically
significant compared to either the bare HOPG (p ) 0.014) or
the G112G coating (p ) 0.007). The reduction in static platelet
adhesion on G4G is consistent with the TEG side chain creating
hydrated patches that repel plasma proteins and platelets. On
all three surfaces, different morphologies of adherent platelets
were observed, which likely represent different stages of platelet
activation.
To understand the effect that the supramolecular network has
in creating a uniform coating of TEG, surface studies of 6, a
synthetic precursor of G4G which does not have guanine-PNA
end groups, were compared with G4G. 6 adsorbed onto HOPG,
but no band assemblies were observed, and the adsorbed
material was not stable to AFM fluid tapping mode imaging
(see Supporting Information), suggesting that the extensive
hydrogen-bonding mediated by the guanine-PNA units plays
an important role in stabilizing G4G assemblies on HOPG.
Static platelet adhesion studies on 6-coated HOPG showed an
average platelet coverage over the sample of 28.4 ( 9.2%
(Figure 13f). The large standard deviation comes from the
heterogeneous platelet coverage in this systems, some regions
had platelet coverage similar to bare HOPG (Figure 13a) while
other regions had platelet coverage closer to G4G surfaces
(Figure 13c). These platelet adhesion results suggest that the
guanine-PNA units in the G4G coatings help to present TEG
side chains more uniformly. Thus, while it seems that the
adsorption of TEG monomers on HOPG can influence platelet
adhesion, studies with this control compound suggest that, with
guanine-PNA end groups a supramolecular network is impor-
tant to create a more uniform coating.
Conclusions
In these studies, tunable surface assemblies have been created
using low-molecular weight ditopic monomers on a HOPG
(44) Zhu, J.; Marchant, R. E. Biomacromolecules 2006, 7, 1036-1041.
9
J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008 1475

A R T I C L E S
Kumar et al.
surface. The assemblies are composed of bands with widths that
can be systematically varied by simply changing the length of
the core hydrocarbon unit. Furthermore, this concept has been
extended into using these assemblies as scaffolds to supramo-
lecularly graft groups (in this case TEG) onto HOPG. These
grafted assemblies have been shown to be stable at biologically
relevant temperatures and have even shown the ability to
influence biological processes, namely static platelet adhesion.
This concept of using the surface assemblies as scaffolds is
potentially a very versatile one in which a wide range of
biologically active (or other functionalities) can be envisaged,
opening the door to systematic, facile functionalization of a
surface using a simple dip-coating process.
Acknowledgment.
Funding for this research was
provided by NIH Grants EB-002067 and NIBIB-EB-001466,
Case MSTP Grant T32 GM07250, and NSF Grant MWN DMR-
0602869.
Supporting Information Available: Synthetic procedures,
1
MALDI-MS, H NMR, 13C NMR for G18G, G110G, G212G,
G218G, G3G, and G4G monomers, molecular modeling pro-
tocols and additional molecular models, AFM protocols, and
details of the platelet experiments. This material is available
free of charge via the Internet at http://pubs.acs.org.
JA0775927
9
1476 J. AM. CHEM. SOC. VOL. 130, NO. 4, 2008