Orthogonally reactive SAMs as a general platform for bifunctional silica surfaces

Article abstract of DOI:10.1021/la1041647

We report the synthesis and self-assembly of azide and amine trimethoxysilanes that result in mixed monolayers on silica. The amine and azide functional groups can be independently reacted with acid chlorides and terminal alkynes, respectively. Consequent

Full text of DOI:10.1021/la1041647

pubs.acs.org/Langmuir
© 2010 American Chemical Society
Orthogonally Reactive SAMs as a General Platform for Bifunctional
Silica Surfaces
Md. Shafiul Azam, Sean L. Fenwick,† and Julianne M. Gibbs-Davis*
Department of Chemistry, University of Alberta, Edmonton, Alberta, T6G 2G2, Canada.
† Present address: Department of Chemistry, Simon Fraser University, Burnaby,
British Columbia, V5A 1S6, Canada
Received July 19, 2010. Revised Manuscript Received November 10, 2010
We report the synthesis and self-assembly of azide and amine trimethoxysilanes that result in mixed monolayers on
silica. The amine and azide functional groups can be independently reacted with acid chlorides and terminal alkynes, respec-
tively. Consequently, these orthogonally reactive monolayers represent a general starting point for making bifunctional
surfaces. Using X-ray photoelectron spectroscopy, we determined the azide/amine surface ratio as well as the reactivity
of the azide and amine functional groups in the mixed self-assembled monolayer (SAM). Significantly, the surface azide/
amine ratio was much lower than the azide/amine ratio in the self-assembly mixture. After determining the self-assembly
mixture composition that would afford 1:1 azide-amine mixed monolayers, we demonstrated their subsequent function-
alization. The resulting bifunctional surface has a similar functional group ratio to the azide/amine precursor SAM
demonstrating the generality of this approach.
1. Introduction
immobilized groups must be on molecular length scales.10,13
Unlike nanometer-scale surface patterns that can be fabricated
from lithographic techniques, to immobilize pairs of functional
groups within subnanometer distances, the groups must be mixed
within the monolayer.11,16,17
With advances in materials and polymer chemistry, multifunc-
tional materials are becoming increasingly accessible.1 Such
materials that combine targeting, imaging, and therapeutic func-
tions onto one scaffold have the potential to revolutionize the
treatment of disease.2 In addition to materials,3 controlling the
combinationof specific functional groups in self-assembled mono-
layers (SAMs) is desirable for research requiring well-defined sur-
faces4 such as biodiagnostics,5-7 cell biology,8,9 and mechanistic
studies of immobilized catalysts.10,11 With multifunctional mono-
layers, synergistic and cooperative effects can beexploredbetween
neighboring immobilized functional groups.12-14 For example,
mixed adlayers containing two types of catalysts have been syn-
thesized on silica where each immobilized catalyst activates a dif-
ferent reactant in the catalytic cycle.10 Such cooperative catalysts
have garnered much attention because they can increase the scope
and efficiency of catalytic transformations.15 In this and other
cooperative interfacial systems, the distance between the two
Such mixed monolayers are typically synthesized by introduc-
ing the surface to a mixture of two surface-reactive molecules.16 In
many examples, one molecule with a functional group of
interest is diluted in the mixed monolayer by another molecule
that does not have any specific function.4,11,18 For example,
receptor molecules in a monolayer are often spaced out from one
another with an excess of ethyleneglycol-terminated surface reac-
tive groups.16,18 This “dilution” molecule can minimize nonspe-
cific interactions with the monolayer. Alkyl terminated surface
reactive groups are also often used to vary the surface density in
mixed monolayers.16 To quantify the surface density in such mixed
monolayers, Collman, Chidsey and coworkers used cyclic voltam-
metry to measure the absolute surface coverage of an electroche-
mically active group diluted in an alkyl monolayer.19 The ratio of
the electrochemically active molecule and the dilution molecule was
inferred from IR measurements but could not be directly measured
because the dilution molecule lacked any electrochemical or IR
signature. Indeed, it is often difficult to find unique spectroscopic or
electrochemical signatures for each component in a mixed mono-
layer. As a result, the solution ratio of the two components is often
used as a guide for the surface ratio.9
*To whom correspondence should be addressed. E-mail: gibbsdavis@
ualberta.ca.
(1) Iha, R. K.; Wooley, K. L.; Nystrm, A. M.; Burke, D. J.; Kade, M. J.;
Hawker, C. J. Chem. Rev. 2009, 109, 5620–5686.
(2) Ferrari, M. Nat. Rev. Cancer 2005, 5, 161–171.
(3) Kessler, D.; Nilles, K.; Theato, P. J. Mater. Chem. 2009, 19, 8184–8189.
(4) Onclin, S.; Ravoo, B. J.; Reinhoudt, D. N. Angew. Chem., Int. Ed. 2005, 44,
6282–6304.
(5) Mrksich, M.; Whitesides, G. M. Trends in Biotechnol. 1995, 13, 228–235.
(6) Patrie, S. M.; Mrksich, M. Anal. Chem. 2007, 79, 5878–5887.
(7) Frisk, M. L.; Tepp, W. H.; Johnson, E. A.; Beebe, D. J. Anal. Chem. 2009, 81,
2760–2767.
(8) Min, D.-H.; Su, J.; Mrksich, M. Angew. Chem., Int. Ed. 2004, 43, 5973–5977.
(9) Derda, R.; Musah, S.; Orner, B. P.; Klim, J. R.; Li, L.; Kiessling, L. L. J. Am.
Chem. Soc. 2010, 132, 1289–1295.
Other groups have characterized the composition of bifunc-
tional monolayers where the two head groups have distinguish-
able signatures in IR absorbance or X-ray photoelectron spec-
troscopy (XPS). These experiments explored the influence of the
headgroup, the alkyl chain length, and the solvent polarity on the
ratio of the two components in the resulting monolayer.20,21
(10) Huh, S.; Chen, H.-T.; Wiench, J. W.; Pruski, M.; Lin, V. S.-Y. Angew.
Chem., Int. Ed. 2005, 44, 1826–1830.
(11) Banet, P.; Marcotte, N.; Lerner, D. A.; Brunel, D. Langmuir 2008, 24, 9030–
9037.
(12) Huskens, J. Curr. Opin. Chem. Biol. 2006, 10, 537–543.
(13) Mouawia, R.; Mehdi, A.; Reye, C.; Corriu, R. J. P. J. Mater. Chem. 2008,
18, 4193–4203.
(16) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M.
Chem. Rev. 2005, 105, 1103–1169.
(17) Bass, J. D.; Katz, A. Chem. Mater. 2006, 18, 1611–1620.
(18) Hudalla, G. A.; Murphy, W. L. Langmuir 2009, 25, 5737–5746.
(19) Collman, J. P.; Devaraj, N. K.; Eberspacher, T. P. A.; Chidsey, C. E. D.
Langmuir 2006, 22, 2457–2464.
(14) Hudalla, G. A.; Murphy, W. L. Langmuir 2010, 26, 6449–6456.
(15) Jacoby, M. Chem. Eng. News 2007, 40–42.
Langmuir 2011, 27(2), 741–750
Published on Web 12/17/2010
DOI: 10.1021/la1041647 741

Article
Scheme 1. Generating Bifunctional Surfaces from Orthogonally
Reactive SAMs
Azam et al.
was 4-(trifluoromethyl)phenyl acetylene, which was synthesized from
the commercially available trimethylsilyl-protected analog.23
All other reagents or solvents were purchased from Aldrich,
VWR, Fisher, or Gelest and used without further purification.
Silica gel (70-230 mesh, 100 A) was purchased from Sigma-
Aldrich, and silicon (100) wafers (prime grade, 0.5 mm thickness)
were purchased from Fluoroware.
˚
General Procedure for Preparing Azide-Substituted Tri-
methoxysilanes. The following procedure was modified from a
procedure noted in a patent application for synthesizing propyl
azido trimethoxysilane.24 NaN3 (1.34 g, 20.7 mmol) was weighed
in a 50 mL Schlenk flask containing a magnetic stirbar. The flask
was placed under nitrogen, followed by the addition of dimethyl-
formamide (DMF) (6 mL). After stirring for 15 min, chloride-
substituted benzyl or propyl trimethoxysilane (10.2 mmol) was
added to the suspension via syringe. The reaction flask was then
equipped with a water condenser and allowed to stir for 48 h at
70 °C under N2. After the reaction was cooled to room temperature,
dry hexane (10 mL) was cannula-transferred to the reaction
mixture. This biphasic mixture was stirred for 2 h, then allowed
to settle for 1 h. At this point, the upper hexane layer was decanted
via cannula to another 25 mL Schlenk flask under N2, and the
hexane was removed under reduced pressure on a Schlenk line to
yield the desired product as a colorless liquid. Because of the
possible explosive nature of the hydrolyzed silyl azides, dry
toluene (3-5 mL) was immediately transferred to the product,
which was thereafter kept as a solution under nitrogen. The con-
centration and yield were calculated from the 1H NMR data for a
sample made of 50 μL of the reaction mixture using 10 μL of ethyl
acetate in acetone-d6 as an internal standard. No chloride starting
material was visible in the 1H NMR.
Significantly, it was found that the surface ratio could vary dras-
tically from the solution ratio depending on the experimental
conditions. Consequently, this early work in the field of self-
assembly led by the Whitesides group established that surface
ratio determination in mixed monolayers requires quantitative
surface characterization methods.20
Combining multiple functions into a monolayer will be critical
for understanding cooperative effects in biological and materials
systems.12 The difficulty in characterizing and controlling the
distribution and ratio of various functional groups, however, has
slowed the development of well-defined and structurally complex
bifunctional surfaces. Therefore, the development of general
methods for making controlled multifunctional monolayers is
needed to accelerate progress in this area.17 In the following, we
outline a general strategy for accessing different combinations of
functions on silica based on a mixed SAM containing two orthog-
onally reactive groups. By characterizing and controlling their
surface ratio, the ratio of virtually any mixture of surface-bound
molecules can be controlled (Scheme 1). Additionally, by combin-
ing self-assembly with orthogonal reactivity, bifunctional mixed
monolayers become available that are not easily accessed through
direct self-assembly methods. This concept was very recently em-
ployed by Hudalla and Murphy in a trifunctional mixed monolayer
on gold consisting of a small percentage of azides and carboxylic-
acid-terminated thiols mixed within an ethylene glycol-terminated
monolayer. The azide and an activated form of the acid could then
be reacted with two peptides containing alkynyl and nucleophilic
groups, respectively. Although the amount of azide and acid was
varied, the absolute ratio of these two orthogonal groups was
not determined.14 Moreover, the acid and azide represented only
a small fraction of the monolayer, which is not ideal for applica-
tions requiring a highdensity of surface reactive sites like coopera-
tive catalysis. Therefore, to develop densely functionalized ortho-
gonally reactive surfaces of controlled ratios, a new system must
be developed. Herein, we demonstrate a quantitative approach to
control the ratio of functional groups in monolayers post-SAM
formation, starting from a well-defined azide-amine mixed
monolayer.
4-(Trimethoxysilyl)benzyl Azide. Final concentration in
1
toluene: 0.54-0.96 M. Yield: 48.2%. H NMR (acetone-d6): δ
3.60 (s, 9H, CH3-OSi), 4.48 (s, 2H, CH2-N3), 7.56 (dd, 4H, Ar-H).
IR cm-1: 2944.4, 2842.1, 2099.7, 1464.6, 1414.2, 1345.0, 1277.9,
1242.9, 1193.1, 1089.0.
3-(Trimethoxysilyl)propyl Azide. Final concentration in
toluene:0.96-1.83M. Yield 73.1%. 1H NMR (acetone-d6): δ 0.69
(t, 2H, -CH2-CH2-CH2-N3), 1.68 (m, 2H, CH2-CH2-N3), 3.28
(t, 2H, -CH2-N3), 3.55 (s, 9H, CH3-OSi). IR cm-1: 2944.4,
2842.5, 2100.9, 1605.5, 1455.3, 1399.7, 1342.2, 1277.4, 1248.6,
1193.0, 1124.4, 1085.9.
Synthesis of 4-Bromobenzoyl Chloride. 4-Bromobenzoic
acid (2.02 g, 10.1 mmol) and thionyl chloride (5.0 mL, 69 mmol)
were combined in a 25 mL round-bottomed flask equipped with a
stirbar and a water condenser, at which point the mixture was
refluxed for 5 h. After cooling to room temperature, the thionyl
chloride was removed under reduced pressure, and the crude pro-
duct was purified by vacuum distillation (50-52 °C, 100 mTorr)
to yield a white solid (1.66 g, 7.06 mmol, 70%). mp 38.8-40.3 °C
(matched with manufacturer (TCI America) data). 1H NMR
(CDCl3): δ 7.67 (d, 2H, Ar-H), 7.98 (d, 2H, Ar-H). 13C NMR
(CDCl3): δ 131.31 (s, ArC-COCl), 132.27 (s, ArC-Br), 132.54 (s,
ArC), 132.76 (s, ArC), 167.79 (s, -COCl). High-resolution EI-MS
(Mþ•): 219.91139.
Synthesis of 4,4,4-Trifluorobutanoyl Chloride. 4,4,4-Tri-
fluorobutanoic acid (6.88 g, 4.85 mmol) was taken in a 25 mL
round-bottomed flask and dissolved in the minimum volume of
dichloromethane (DCM) (∼3.0 mL). A solution of oxalyl chlor-
ide in DCM (2.0 M, 3.0 mL, 6.0 mmol) was added slowly via
syringe; then, the reaction mixture was placed in an ice bath. A
catalytic amount of DMF (1 drop) was added, and the reaction
was stirred for 15 min. It was then allowed to come to room
temperature and stirred for an additional 6 h. DCM and excess
oxalyl chloride were removed slowly under moderately reduced
pressuretoavoid evaporation ofthe product. Wethen purifiedthe
2. Experimental Section
Materials. 3-{2-[2-(2-Methoxy-ethoxy)-ethoxy]-ethoxy}-
propyne22 was synthesized following a literature procedure, as
(20) Bain, C. D.; Whitesides, G. M. J. Am. Chem. Soc. 1988, 110, 6560–6561.
(21) Kang, J. F.; Liao, S.; Jordan, R.; Ulman, A. J. Am. Chem. Soc. 1998, 120,
9662–9667.
(22) Scates, B. A.; Lashbrook, B. L.; Chastain, B. C.; Tominaga, K.; Elliott,
B. T.; Theising, N. J.; Baker, T. A.; Fitch, R. W. Bioorg. Med. Chem. 2008, 16,
10295–10300.
(23) Bruner, E. L.; Span, A. R.; Bernasek, S. L.; Schwartz, J. Langmuir 2001, 17,
5696–5702.
(24) Verkade, J. G.; Lin, V. S.-Y.; Arunkanti, S. Immobilized Iminophospha-
tranes Useful For Transesterification, US 2005/0176978 A1, Aug. 11, 2005.
742 DOI: 10.1021/la1041647
Langmuir 2011, 27(2), 741–750

Azam et al.
Article
crude product by collecting the acid chloride (colorless liquid) in
a liquid-nitrogen-cooled trap under reduced pressure (0.459 g,
2.86 mmol, 59%). H NMR (CDCl3): δ 2.54 (m, 2H, -CH2-
Self-Assembled Monolayer Formation on Silica Surfaces.
Cleaned samples wererinsedwith a stream of methanol(15 mL), a
mixture of methanol and toluene (1:1 v/v, 15 mL), and finally a
stream of toluene (15 mL) and then immersed in 3 mM solution of
the trimethoxysilane in dry toluene with 0.1% water (v/v) for 4 h.
The same protocol was used for SAM formation in hexane, with
hexane substituted for toluene in the above steps. To make the
orthogonally reactive SAMs, we combined different proportions
of amino and azido trimethoxysilanes to reach a total concentra-
tion of 3 mM trimethoxysilane. After the 4 h of reaction time, the
samples were rinsed thoroughly with toluene (15 mL), sonicated
for 1 min each in toluene (15 mL), toluene/methanol (15 mL), and
methanol (15 mL) and then blown dry with a stream of nitrogen
gas. The samples were either used immediately in a subsequent
reaction or stored under vacuum.
1
CF3). 3.19 (t, 2H, -CH2-COCl), 13C NMR (CDCL3): δ 29.37
(q, -CH2-CF3), 39.59 (q, -CH2-COCl), 125.78 (q, -CF3),
171.75 (s, -COCl). IR cm-1: 2917.4, 2849.1, 1829.7, 1446.3,
1424.3, 127.9, 1232.3, 1143.5, 1114.5, 1063.5, 1008.1, 985.1.
Synthesis of 2,2,2-Trifluoroethyl Hept-6-ynoate. Hept-6-
ynoic acid (0.996 g, 7.89 mmol) was weighed in a 25 mL round-
bottomed flask and dissolved in the minimum volume of DCM
(4.0 mL). An oxalyl chloride solution in DCM (2.0 M, 5.0 mL,
10.0 mmol) was added slowly via syringe; then, the reaction mix-
ture was placed in an ice bath. After the reaction mixture was
cooled to 0 °C, DMF (1 drop) was added, and immediate effer-
vescence of gas was observed. After the reaction mixture was
stirred in an ice bath for 15 min, it was allowed to come to room
temperature and stirred for 2 h. DCM and excess oxalyl chloride
were then rotovaped off, and the reddish brown crude product
was used in the next step. 1H NMR (CDCl3): δ 1.55-1.63 (m, 2H,
CHtC-CH2-CH2-), 1.81-1.88 (m, 2H, CHtC-CH2-CH2-
CH2-), 1.97 (t, 1H, HCtCH2C-), 2.21-2.25 (d of t, 2H,
HCtC-CH2-), 2.93 (t, 2H, CH2-CO).
Modification of Powdered Silica for the IR Absorbance
˚
Experiments. SiO2 powder (0.5 g, 70-230 mesh, 100 A) was
added to a 25 mL flask containing a solution of the respective
trimethoxysilaneintoluene(5 mL, 50mM). Thisslurrywasstirred
for 4 h at room temperature. The higher concentration of trime-
thoxysilane was used because of the high surface area of the silica gel
(surface area = 300 m2/g). After reaction, the functionalizedsilica
powder wasfilteredina Buchner funnelto remove the filtrate. The
powder was then washed with toluene (4 ꢀ 15 mL) and methanol
(4 ꢀ 15 mL). Finally, the powder was annealed for 30 min in an
oven at 90 ( 5 °C. The subsequent acylation or [3 þ 2] cycloaddi-
tion reaction was performed on the functionalized powder using
the same procedure as the silica surfaces with 0.1 g powdered silica
samples.
Hept-6-ynoyl chloride was taken in a 25 mL flask and dissolved
in DCM (4 mL). 2,2,2-Trifluoroethanol (0.7 mL, 9.1 mmol) and
triethylamine (TEA, 0.9 mL, 9.0 mmol) were dissolved in DCM
(3 mL) and added to the flask containing the acid chloride. The
resulting mixture was stirred for 4 h at room temperature. The
reaction mixture was then diluted with DCM (5 mL), washed with
water (5 ꢀ 3 mL), dried over Na2SO4, and filtered in a 25 mL
round-bottomed flask. The solvent and any remaining trifluoro-
ethanol starting material were removed from the filtrate by
rotary evaporation. The crude product was purified by vacuum
distillation (65-70 °C, 48 mTorr) to yield a clear oil (1.17 g, 5.55
mmol; overall yield in two steps 70.3%). 1H NMR (CDCl3):
δ 1.58 (m, 2H, CHtC-CH2-CH2-), 1.79 (m, 2H, CHtC-
CH2-CH2-CH2-), 1.95 (t, 1H, CHtCCH2-), 2.23 (d of
t, 2H, CHtC-CH2-), 2.44 (t, 2H, CH2-CO), 4.47 (q, 2H,
O-CH2CF3). 13C NMR (CDCl3): δ 18.06 (s, CHtC-CH2-),
23.69 (s, CHtC-CH2-CH2-), 27.62 (s, CHtC-CH2-
CH2-CH2-), 33.07 (s, CHtC-CH2-CH2-CH2-CH2-),
60.21 (q, CF3-CH2-), 68.78 (s, CHtC-), 83.65 (s, CHtC-),
123.01 (q, CF3-CH2-), 171.72 (s, CdO). IR cm-1: 3311.1,
2928.6, 2869.9, 2119.3, 1760.9, 1414.6, 1283.1, 1171.3, 1141.5,
1083.3, 978.4.
General Method for Acylation of Amino SAMs. The acid
chloride (0.5 mmol) was dissolved in a 20 mL vial in acetonitrile
(ACN) (5 mL), followed by the addition of triethylamine (0.05 mL,
0.36 mmol). Freshly prepared amino SAMs were immersed in
the vial containing the reaction mixture, which was then capped
and allowed to react overnight. The acylated amino SAMs were
then taken out of the vial and rinsed with ACN (10 mL) and
toluene (10 mL), followed by 1 min sonication in toluene (10 mL)
and methanol (10 mL) each and then blown dry with a stream of
nitrogen gas.
General Method for [3 þ 2] Cycloaddition Reaction of
Alkynes with Azido SAMs. The following procedure was used
for all of the alkynes except for the reaction of benzyl azido SAM
with propargyl-substituted triethyleneglycol monomethyl ether
(TEG-alkyne). Copper sulfate (6.2 mg, 0.025 mmol) and ascorbic
acid (8.8 mg, 0.050 mmol) were transferred to a 20 mL vial and
dissolved in DMF (5 mL) by sonication. The alkyne (0.25 mmol)
was dissolved in that solution at which point freshly prepared
azido SAMs were immediately immersed in the vial, which was
then capped, wrapped in foil, and allowed to react overnight. The
reacted azido SAMs were then taken out of the vial and rinsed
with DMF (10 mL) and toluene (10 mL), followed by sonication
for 1 min each in toluene (10 mL) and methanol (10 mL). Finally,
the samples were blown dry with a stream of nitrogen and stored
under vacuum.
Substrate Cleaning Procedure. Silicon wafers containing a
native oxide layer were used as the substrates for the XPS and
AFM measurements. Glass slides (Fisher, Microscope Slides)
were used for the contact angle measurements. Prior to surface
functionalization, both types of samples were cleaned with
methanol (∼15 mL), followed by drying in an oven for 10 min
at 90 ( 5 °C. The cleaned wafers were reacted with “piranha
solution” (3:1 mixture of sulfuric acid and 30% hydrogen
peroxide) in a glass beaker to make the surface rich with hydroxyl
groups.25 (Caution: Piranha solution reacts explosively with or-
ganic materials, so it should not be combined with any organic
compounds or solvents. Use caution.) The beaker containing the
samples was rinsed copiously with deionized water (5 ꢀ 100 mL);
the individual samples were then each rinsed with a stream of
Millipore water (100 mL), sonicated in Millipore water (100 mL)
for 10 min, and blown dry with a stream of nitrogen gas.
The substrates thus obtained were placed in an oven for 30 min
(90 ( 5 °C). In some cases, the silicon wafers were next cleaned
using a reactive ion etcher (μEtch RIE, Plasma Lab) with oxygen
as reactive gas in a radio frequency (RF)-induced plasma to
remove residual fluorine from the packing material. Process
pressure was 500 mTorr, and process time was 1 min for plasma
cleaning.
[3 þ 2] Cycloaddition of the Benzyl Azido SAM with
TEG-Alkyne. AFM analysis suggested that different cycloaddi-
tion conditions were required for the benzyl azido SAM with
TEG-alkyne. Therefore, copper(I) iodide (9.5 mg, 0.050 mmol)
and triethylamine (0.1 mL, 0.7 mmol) were transferred to a 20 mL
vial containing water (2.5 mL) and methanol (2.5 mL). TEG-
alkyne (0.05 g, 0.25 mmol) was dissolved in that solution, and
freshly prepared benzyl azido SAMs were immersed in the vial,
which was then capped. The SAM sample was taken out of the
vial after overnight reaction and rinsed with water (10 mL),
methanol (10 mL), toluene (10 mL), and methanol (10 mL), fol-
lowed by sonication in methanol (10 mL) for 1 min. The samples
were blown dry with nitrogen and stored under vacuum.
Bifunctional Modification of Orthogonally Reactive SAM.
The orthogonally reactive SAMs were reacted via acylation and
(25) Zhang, F.; Srinivasan, M. P. Langmuir 2004, 20, 2309–2314.
Langmuir 2011, 27(2), 741–750
DOI: 10.1021/la1041647 743

Article
Azam et al.
Scheme 2. Synthesis of Azido Trimethoxysilanes
cycloaddition, sequentially, following the same modification pro-
cedures detailed above.
XPS Analysis. XPS measurements were performed on sam-
ples that had been prepared within 4 days using the AXIS
ULTRA spectrometer (Kratos Analytical). The base pressure in
the analytical chamber was <3 ꢀ 10-8 Pa. The monochromatic
Al KR source(hν = 1486.6 eV) was used ata power of 210 W. The
photoelectron exit angle was 90°, and the incident angle was 35.3°
from the plane of the surface. The analysis spot was 400 ꢀ 700 μm.
The resolution of the instrument was 0.55 eV for Ag 3d and
0.70 eV for Au 4f peaks. Survey scans were collected for binding
energies from 1100 to 0 eV with an analyzer pass energy of 160 eV
and a step of 0.35 eV. The high-resolution spectra were run with a
pass-energy of 20 eV and a step of 0.1 eV. Relative sensitivity
factors (RSFs) for different elements were as follows: 1 for F (1s),
0.477 for N (1s), 0.955 for Br (3d). Only one set of XPS scans was
performed on a given sample; therefore, XPS analysis before and
after surface reactions was performed on different samples.
AFM Measurements. AFM experiments were performed on
samples no more than 2 days old using a MFP-3D (Asylum)
operated in tapping mode with commercially available Si canti-
levers (Olympus, 300 kHz frequency). The force constant of the
cantilevers was 42 N/m. The oscillation frequency used for
tapping mode was 310 ( 5 kHz.
Contact Angle. Contact angles of water were measured on
samples that had been prepared within 2 days using a FTA200
(First Ten Angstroms) system for contact angle and drop shape
analysis fitted with manufacturer software. Contact angles were
determined from sessile water drops of 8.0 μL volume. The
average ofatleast six contact angle measurementswasdetermined
from, at minimum, two separately prepared slide samples.
Calculation of Azide/Amine Ratio from N-Peak Fitting.
The high-resolution N (1s) scan for the monofunctional azide
and azide-amine mixed monolayers was analyzed using the
multipeak fitting program in Igor Pro 6.05. Two Gaussian
peaks, one centered at 401 eV and another at 405 eV, resulted
after fitting the N (1s) region of the high-resolution XPS
spectrum. The peak areas were determined from the following
equation
the elemental peak areas in the XPS spectra of the amine-azide
mixed monolayers adjusted with their corresponding RSFs.
Specifically, the measured Br/N and F/Br ratios were substituted
into the following equations to determine the azide/amine and
triazole/amide surface ratios.
azide
1 - ðBr=NÞ=0:87
From Br=N ratio ðassuming 87% conversionÞ :
¼
amine 3 ꢀ ðBr=NÞ=0:87
triazole
F=Br
From F=Br ratio :
¼
amide
3
3. Results and Discussion
Our strategy requires the selection of two reactive groups that
exhibit orthogonal reactivity; in other words, they should not
react with one another or the other’s reactive partner (Scheme 1).
One obvious choice of reaction is the Cu-mediated cycloaddition
of surface-bound azides with terminal alkynes owing to the func-
tional group tolerance, specificity, and high reactivity associated
with this click reaction.1,19,26-29 As a complementary reaction,
amine-terminated monolayers have been used to introduce sub-
stituted electrophiles onto surfaces and other materials.4 Com-
mercially available p-aminophenyl trimethoxysilane was selected
because it is known to make uniform reactive SAMs.25 For the
azide monolayer, most azido SAMs on silica are synthesized post-
self-assembly by reaction of an alkylhalide SAM with sodium
azide.26,30 Because alkyl halides and amines react with one an-
other, we had to develop a method for the direct assembly of azido
SAMs to generate mixed monolayers with amino silanes. There-
fore, we synthesized two azido trimethoxysilanes from the reac-
tion of sodium azide with p-(trimethoxysilyl)benzylchloride and
n-(trimethoxysilyl)propylchloride (Scheme 2). With the corre-
sponding benzyl and propyl azido silanes in hand, we were able
to explore the influence of the azide structure on monolayer self-
assembly. Additionally, by directly incorporating the azide into
the monolayer, we prevented the incorporation of unreacted alkyl
halides, which is a potential problem in post-self-assembly
modification strategies.30
The synthesis and reactivity of the monofunctional SAMs were
first explored by exposing silica surfaces to 3.0 mM solutions of
the corresponding trimethoxysilane in toluene (0.1% water v/v)
for 4 h. The presenceofthe amine and azide functionalgroups was
confirmed by AFM and water contact angle measurements (vide
infra). To quantify the extent of reaction between the amino or
azido SAM with their solution-phase reactive partners, we elected
to use XPS and CF3-labeled acid chlorides and alkynes, respec-
tively. From the XPS spectra, the percent completion can easily be
determined, where 100% surface conversion of the amines leads
to a F/N ratio of three, and conversion of all azides leads to a F/N
ratio of one. Quantification of the XPS F1s and N1s signal
pffiffiffi
peak area ¼ πðamplitude ꢀ widthÞ
The calculated ratio of the 401/405 eV peak areas for the benzyl
azido SAM was 3.07(7), whereas the corresponding peak ratio for
the propyl azido SAM was found to be 4.1(1). Substituting these
values and the peak areas calculated for the mixed monolayers, the
azide/amine ratio was determined from the corresponding equation
benzyl azide
amine
ð1 þ 3:07Þpeak area405eV
¼
3
1
ꢀ
peak area401eV - ðpeak area405eV ꢀ 3:07Þ
propyl azide
amine
ð1 þ 4:1Þpeak area405eV
¼
3
1
ꢀ
peak area401eV - ðpeak area405eV ꢀ 4:1Þ
For peak fitting, the peak center, width, and amplitude values
were allowed to vary with the exception of the 1:1 propyl azide-
amine sample. For this spectrum, the small peak at 405 eV was
difficult to fit without holding any parameters. Therefore, a peak
width of 0.74 eV was held for the smaller peak at 405 eV, because
this value was generated for the peak at 405 eV from fitting both
the monofunctional propyl azido SAM and the 5:1 propyl azide-
amine SAM spectra.
(26) Lummerstorfer, T.; Hoffmann, H. J. Phys. Chem. B 2004, 108, 3963–3966.
(27) Collman, J. P.; Devaraj, N. K.; Chidsey, C. E. D. Langmuir 2004, 20, 1051–
1053.
(28) Fleming, D. A.; Thode, C. J.; Williams, M. E. Chem. Mater. 2006, 18, 2327–
2334.
(29) Devaraj, N. K.; Collman, J. P. QSAR Comb. Sci. 2007, 26, 1253–1260.
(30) Heise, A.; Stamm, M.; Rauscherc, M.; Duschnerc, H.; Menzel, H. Thin
Solid Films 1998, 327-329, 199–203.
Calculation of the Azide/Amine and Triazole/Amide Ra-
tios. Atomic compositions of the surfaces were calculated from
744 DOI: 10.1021/la1041647
Langmuir 2011, 27(2), 741–750

Azam et al.
Article
intensities indicated that exposing the aminophenyl SAM over-
night to CF3-substituted benzoyl chloride led to quantitative for-
mation of the amide (Table 1, entry 1; Figure 1A). To determine
the amino SAM reactivity with aliphatic acid chlorides, we per-
formed the same experiment with CF3-butanoyl chloride (Table 1,
entry 3). High yields were also observed for this reactant, which
supports previous reports indicating that different substituted
acid chlorides can be attached to these aminophenyl mono-
layers.31 Next, we examined the reactivity of the propyl azido
SAM and benzyl azido SAM in the overnight reaction between
the azido monolayers and p-(trifluoromethyl)phenyl acetylene. We
were pleased to note that quantitative conversion to the triazole
occurred with azido SAMs directly assembled on silica (Table 1,
entries 4 and 6; Figure 1B,C). High conversions were also
observed with an aliphatic alkyne, indicating that our new azido
SAMs have broad reactivity (Table 1, entries 5 and 7). The ability
of all of our SAMs to react with aromatic and aliphatic reactants
is important because it indicates that a wide variety of substituted
alkynes and acid chlorides can be attached using these mono-
layers, which supports the generality of our strategy. Finally, con-
trol experiments were performed that indicated that neither the
azido SAMs nor the amino SAM exhibited cross reactivity with
acid chlorides and alkynes, respectively, under appropriate reac-
tion conditions. (See the Supporting Information.)
Table 1. XPS Data and Percent Conversions for the Amine and
Azide Monofunctional SAMs before and after Reaction with
Various Reactantsa
To confirm the structural changes of the azides and amines
after undergoing their respective reactions, infrared absorbance
experiments were performed on powdered silica using the reactants
shown in Table 1 in a similar manner as the silica-coated wafers
used in the XPS experiments. In all IR absorbance spectra, a broad
absorption was observed from 2800 to 3700 cm-1 attributed to
water and H-bonded silanol.32,33 For the benzyl azide- and propyl
azide-modified powdered silica, the intense absorption peak at
2107 cm-1, characteristic of the azide stretch, confirmed the attach-
ment of the azides (Figure 2A, blue traces). In general, this peak
disappeared after reaction with alkynes, indicating quantitative
conversion to the triazole product. The one exception was the
reaction of the aliphatic alkyne, 2,2,2-trifluoroethyl heptynoate with
propyl azide-modified silica (Figure 2A (ii), black trace), where
the azide peak was diminished but still observable. However, for
both azide samples, a new peak appeared at 1740 cm-1 after reac-
tion with this ester substituted alkyne, confirming the presence of
the carbonyl (CdO stretch, Figure 2A, black traces). Similarly new
a Data represent the average of at least two samples made from
separate batches. The error indicates the range of measured values.
Figure 1. N1s and F1s XPS data for the monofunctional (A) amino, (B) propyl azido, and (C) benzyl azido SAMs before and after reaction
with their corresponding CF3-labeled reactive partner. The intensity of the N1s XPS signal has been multiplied by 3.5 to allow comparison
with the stronger fluorine signal.
Langmuir 2011, 27(2), 741–750
DOI: 10.1021/la1041647 745

Article
Azam et al.
Figure 2. IR absorption spectra of powdered silica modified with: (A) (i) Propyl azido silane and (ii) benzyl azido silane, before (blue) and
after reaction with an aromatic alkyne (red) or an aliphatic alkyne (black); (B) Phenylamino silane before (blue) and after reaction with an
aromatic acid chloride (red) and an aliphatic acid chloride (black).
Figure 3. Tapping-mode atomic force microscopy images of SAMs before and after reaction with alkyne-substituted triethyleneglycol
(TEG). (A) Benzyl azido SAM. (B) Amino SAM. (C) Benzyl azido SAM reacted with TEG. (D) Mixed monolayer synthesized from a 5:1
molar mixture of benzyl azido and amino trimethoxysilanes after reaction with TEG. (E) Mixed monolayer synthesized from a 1:1 molar
mixture ofbenzyl azido and amino trimethoxysilanes after reactionwith TEG. Table: Contact angles ofwater droplets on SAMs. The contact
angle decreases with increasing hydrophilicity of the monolayer.
peaks were visible in the IR after reaction with the aromatic alkyne,
p-trifluoromethylphenyl acetylene. Specifically, a new peak ap-
peared at 1328 cm-1 attributed to the CF3 antisymmetric stretching
vibration mode of the aromatic 1,4-isomer (Figure 2A, red traces).34
The reactivity of amine-modified silica with acid chlorides was
similarly explored using changes in the IR absorbance spectra.
After reaction of the amino silica and either the aliphatic or aro-
matic acid chloride, a new band at 1653 cm-1 appeared (corre-
sponding to the CdO), confirming the formation of amide groups
(Figure 2B, red trace). This band overlapped with the band at
∼1630 cm-1 present for both the amine- and amide-modified
silica that is attributed to the vibration of the adsorbed water
on silica.35 The peak at 1329 cm-1 corresponding to the aromatic
CF3 on the substituted benzoyl chloride was also visible after
amide formation.34 In addition to confirming the structural
changes associated with forming and reacting these SAMS, the
high reactivity observed for the modified silica powder suggested
that our strategy is amenable to functionalizing porous materials,
with potential uses in multicomponent catalyst10 or therapeutic2
applications requiring high surface area materials.
We next synthesized mixed monolayers from mixtures of azido
and amino trimethoxysilanes with solution mole ratios of 1:1 and
5:1, respectively. For the benzyl azide system, the resulting mixed
monolayers exhibited contact angles with water of 55 ( 1° for the
1:1 mixture and 58 ( 1° for the 5:1 mixture, in between the values
for the corresponding monofunctional SAMs (Figure 3, table; see
Figure S-1 of the Supporting Information for characterization of
propyl azide-amine SAMs). To verify that the surface azide/
amine ratio increased with increasing amounts of azide trimethoxy-
silane in solution, we reacted the azide in these mixed monolayers
with alkyne-substituted triethyleneglycol (TEG). The result-
ing AFM images are shown in Figure 3. The surface roughness
exhibited by the root-mean-square (rms) height values for the
(31) Stokes, G. Y.; Buchbinder, A. M.; Gibbs-Davis, J. M.; Scheidt, K. A.;
Geiger, F. M. J. Phys. Chem. A 2008, 112, 11688–11698.
(32) Mondragon, M. A.; Castano, V. M.; Garcia, M. J.; Tellez, S. C. A. Vib.
~
ꢀ
Spectrosc. 1995, 9, 293–304.
(33) Costa, C. A. R.; Leite, C. A. P.; Galembeck, F. J. Phys. Chem. B 2003, 107,
4747–4755.
(34) Nomarua, K.; Gorelika, S. R.; Kurodaa, H.; Nakaia, K. Nucl. Instrum.
Methods Phys. Res., Sect. A 2004, 528, 619–622.
(35) Anderson, M. W.; Klinowski, J. J. Chem. Soc., Faraday Trans. 1 1986, 82,
1449–1469.
746 DOI: 10.1021/la1041647
Langmuir 2011, 27(2), 741–750

Azam et al.
Article
mixed monolayers after reaction with TEG increased from 1.07
to 1.53 nm with increasing amounts of azide in solution
(Figure 3A-D). This increase in roughness can be explained
by the reaction of the long-chain alkyne with the surface-bound
azides. Additionally, the correlation between surface roughness
and the solution fraction of azido trimethoxysilane is consistent
with an increase in the amount of azide in the monolayer.
Differences in hydrophilicity of the SAMs were also used to
characterize the mixed monolayers. The water contact angle of
the benzyl azide SAM decreased markedly from 60.7( 0.9 to 36 (
2° upon incorporating the hydrophilic TEG chain. For the mixed
monolayer, the trend in contact angle also confirmed that
increasing the azide content led to greater incorporation of the
hydrophilic TEG group (Figure 3, table).
Figure 4. N1s high-resolution scans of (A) benzyl azide-amine and
(B) propyl azide-amine mixed monolayers synthesized from solu-
tions with varying mol % azide (total [trimethoxysilane] = 3.0 mM).
By fitting each spectrum to two Gaussian peaks (red dashed lines)
and comparing the peak ratios of the mixed and monofunctional
azido SAMs, the mole percent of azide in each mixed monolayer was
determined.
After establishing that we had successfully synthesized mixed
monolayers with varying azide ratios, one question that remained
was how closely the azide/amine surface ratio matched the ratio of
azide and amine trimethoxysilanes in solution. To correlate the
solution and surface composition, we took advantage of the
signature double peak in the N1s region of the XPS spectrum that
corresponds to the azide (Figure 1B,C).36 The distinctive two peaks
and the unique position of the smaller peak at 405 eV provide a
handle for characterizing the amount of azide in our mixed
monolayers. Fitting the high-resolution N1s scan to two Gaussian
peaks resulted in a large peak centered at 401 eV and a smaller peak
centered at 405 eV for both azido SAMs. Interestingly, the peak
area ratios differed for the two azido SAMs; the benzyl azido SAM
exhibited a peak ratio of 3.07 ( 0.07, and the propyl azido SAM
exhibited a ratio of 4.1 ( 1. The different ratios for the two sub-
stituted azides suggest that the oxidation state of the azide nitrogen
atoms depends on azide substitution, which is consistent with
different resonant contributions for the propyl and benzyl azides.
Next, we measured the ratio of the peak areas from the high-
resolution XPS data of monolayers synthesized from 1:1 molar
mixtures of the azido and amino trimethoxysilanes. As expected,
the 401 to 405 eV peak ratios were much greater than that of the
monofunctional azido SAMs, owing to overlap of the amine and
azide signals at 401 eV. Specifically, peak area ratios of 13 ( 2 and
12 ( 1 were observed for the benzyl azide-amine and the propyl
azide-amine SAMs, respectively. Using these values, the area of
the peak at 405 eV, and the peak ratio for the corresponding
monofunctional azido SAM, the relative amount of azide was
determined for each mixed monolayer. To illustrate, the following
equation was used to determine the azide/amine ratio for the
benzyl azide-amine mixed monolayers
surface compared with the solution composition (18 ( 2 vs
50 mol %, respectively, Figure 4B). From these results, we con-
clude that the solution ratio differs substantially from the
surface ratio. To achieve 52 mol % of azide in the benzyl
azide-amine mixed monolayer, benzyl azido trimethoxysilane
must be five times more concentrated than aminophenyl tri-
methoxysilane in solution (i.e., 83 mol % azide). For the propyl
azide-amine mixed SAM, a solution azide/amine ratio of 8 was
needed to achieve 53 mol % of the propyl azide in the mixed
monolayer. (See the Supporting Information.) These results
reveal that the composition of azide-amine mixed monolayers
varies significantly from the ratio of silanes in solution. Although
it has long been recognized that surface reactivity and solubility37
can influence surface ratiosinmixed monolayer systems20 because
of the difficulty in quantifying surface ratios, the solution ratio is
often used to interpret the structure and behavior of the mixed
monolayer.7,14 In our mixed monolayer system, such an assump-
tion would introduce significant error to the interpretation be-
cause of the disproportionate amount of amine groups versus
azides.
One explanation for the larger amount of aminophenyl groups
in our mixed monolayers is the ability of aminophenyl silanes to
interact with the silica surface through hydrogen bonding and
self-associate via pi stacking.25 The former should enhance their
presence at the interface and has been observed in flexible amino-
silanes as the first step prior to surface reaction.38 The latter has
been inferred from the well-ordered monolayers that result from
phenylamino silanes.25 Additionally, the ability of amines to act
as base catalysts capable of cross-linking trimethoxysilanes or
condensing them with surface silanol groups could also increase
the incorporation of aminosilanes into the monolayer, particu-
larly if they are self-associating.38,39 In our mixed monolayers, the
azide/amine surface ratio was similar regardless of the azide struc-
ture, which is consistent with the proposed surface concentration
enhancement of the amines due to hydrogen bonding. Moreover,
the similarity suggested that neither azide interacted strongly with
the phenylamine, which was surprising for the benzylazido silane.
benzyl azide
amine
ð1 þ 3:07Þpeak area405eV
¼
3
1
ꢀ
peak area401eV - ðpeak area405eV ꢀ 3:07Þ
We also attempted a three-peak fit of the XPS data to resolve the
amine and azide contributions at 401 eV, but the similarities in the
peak positions (400.8, 400.9, and 401.2 eV for the amine, propyl
azide, and benzyl azide, respectively) led to azide/amine values
that were inconsistent with the peak area at 405 eV.
The peak ratios reveal that only 12 ( 3 mol % azide has been
incorporated into the monolayer for a 1:1 molar mixture of benzyl
azido and amino trimethoxysilane (50 mol % benzyl azide in solu-
tion) (Figure 4A, Table 2). Similarly, the propyl azide-amine mixed
monolayer exhibited a much smaller amount of azide on the
(37) Oyerokun, F. T.; Vaia, R. A.; Maguire, J. F.; Farmer, B. L. Langmuir 2010,
26, 11991–11997.
(38) Vrancken, K. C.; Possemiers, K.; Voort, P. V. D.; Vansant, E. F. Colloids
Surf., A 1995, 98, 235–241.
(39) Vrancken, K. C.; Van Der Voort, P.; Gillis-D’Hamers, I.; Vansant, E. F.;
Grobet, P. J. Chem. Soc., Faraday Trans. 1992, 88, 3197–3200.
(36) Wollman, E. W.; Kang, D.; Frisbie, C. D.; Lorkovic, I. M.; Wrighton, M. S.
J. Am. Chem. Soc. 1994, 116, 4395–4404.
Langmuir 2011, 27(2), 741–750
DOI: 10.1021/la1041647 747

Article
Azam et al.
Table 2. Azide/Amine Surface Ratio for Mixed Monolayers Assembled in Different Solvents
toluene hexane
solution azide mol %
N1s XPS ratioa
surface azide mol %
N1s XPS ratioa
surface azide mol %
83% BnAz
50% BnAz
83% PrAz
50% PrAz
4.3 ( 0.2
13 ( 2
6.3 ( 0.2
12 ( 1
52 ( 6
12 ( 3
44 ( 2
18 ( 2
4.5 ( 0.3
23 ( 11
5.1 ( 0.1
10.0 ( 0.9
48 ( 4
9 ( 5
63 ( 2
23 ( 3
a Ratio represents the XPS peak areas for signals centered at 401 and 405 eV, respectively.
mixed monolayer, despite the solution ratio. Yet the difference
between the surface and solution ratios systematically decreased
with decreasing solvent polarity.21 The resulting decrease in polar
content of the propyl azide-amine monolayer was thus attributed to
a change in monolayer stability with solvent polarity. For our
benzyl azide-amine mixed monolayer, the polarity of the benzyl
azides should be closer to that of the phenylamines, minimizing
the effect of solvent polarity on surface ratio. Future work will
address how the solvent influences the spatial distribution of the
azides and amines,4 which might provide more clues as to the
origin of monolayer stability for these mixed SAMs.
A key aspect of the orthogonally reactive-SAM strategy is that
the ratio of azides to amines will determine the ratio of functional
groups that are subsequently attached. To characterize the ratio
of the triazole and amide groups that result from reacting the azides
and amines, respectively, we once again turned to XPS. Figure 5A
illustrates the functionalization of benzyl azide-amine SAMs with
p-bromobenzoyl chloride, followed by the fluorine-labeled alkyne
used in previous experiments. As a reference, the Br/N ratio
for the monofunctional amino SAM after reaction with
p-bromobenzoyl chloride yielded a ratio of 0.87 ( 0.02, indicating
that the maximum conversion was only 87 ( 2% with this
reactant (Table 1, entry 2). Taking this slightly lower yield into
consideration, the calculated azide/amine surface ratios from
XPS analysis of the Br/N ratio after functionalization are very
similar to those calculated from the peak-fitting of the high-
resolution N1s scan before functionalization (Figure 5B). More
importantly, the ratio of the triazole and amide products deter-
mined from the F/Br ratio was very consistent with the azide/
amine surface ratios, indicating that the ratio of orthogonally
reactive groups dictated the amount of surface products. In
contrast, efforts to synthesize 1:1 CF3 and Br mixed monolayers
from monofunctional SAMs led to CF3/Br surface ratios much
greater than one. Specifically, when a 1:1 mixture of CF3- and
Br-labeled acid chlorides was reacted with an amino SAM, the
resulting CF3/Br surface ratiowas 3.2 (Figure 6A). Similarly, a 1:1
mixture of CF3- and Br-labeled alkynes reacted with a benzyl
azido SAM led to CF3/Br ratios of 2.3 (Figure 6B). The sub-
stantial difference between the observed ratios and the desired
ratio of one emphasizes the difficulty in controlling surface ratios
with post-SAM modification strategies using monofunctional
SAMs. Once again, these experiments highlight the utility of
orthogonally reactive SAMs in controlling surface ratios.
Figure 5. (A) Functionalization of orthogonally reactive SAMs
with (1) Br-labeled acid chloride, followed by (2) CF3-labeled
alkyne. (B) Surface ratios determined from the XPS data of mixed
monolayers prepared from 1:1 or 5:1 molar mixtures of azido
and amino trimethoxysilane (1 Az:Am and 5 Az:Am, respectively).
The azide/amine ratio was determined from the Br/N ratio (Br/N
ratio) and the N peak ratio prior to functionalization (N Peak
Fit). The triazole/amide product ratios were determined from the
F/Br ratio (F/Br ratio). *Denotes reversal of reaction sequence,
that is, the mixed monolayer was reacted first with the alkyne and
then the acid chloride.
We reasoned that toluene might minimize any pi-stacking inter-
actions between the benzyl azides and phenylamines, thereby
impacting their surface ratio. Therefore, we repeated the mixed
monolayer experiments in hexane. Interestingly, changing the
solvent from toluene to hexane had no effect on the benzylazide/
amine surface ratio, suggesting that pi stacking of the solvent did
not play a substantial role. To see whether the same held true for
the propylazide-amine mixed monolayer, we also measured the
role of solvent on its azide/amine surface ratio. For this system, a
greater amount of the propyl azide was observed for monolayers
prepared in hexane rather than toluene (63 ( 2 vs 44 ( 2 mol %
azide, respectively). We expected hydrogen bonding between the
aminosilane and the silica to be enhanced by the hexane; instead
the decrease in amine content indicated either enhanced reactivity
of propylazide in hexane or a change in monolayer stability.
Kang and colleagues observed that the more polar of two surface-
reactive species tended to make up a larger component in the
We did, however, observe some sensitivity in our orthogonally
reactive-SAM strategy to the sequence of functionalization steps.
When we performed the cycloaddition reaction first and the acyl-
ation second, the amount of surface bromine, quantifying the effi-
ciency of amide formation, decreased (Br/N of 0.17 ( 0.01 vs 0.23
( 0.02 for samples acylated second and first, respectively). The
smaller Br/N ratio led to a larger calculated azide/amine ratio
(Figure 5B, Br/N ratio: 5 Az:Am*). Instead, the lack of bromine is
most likely due to lower conversion of the amines to the bromine-
labeled amides. This lower conversion suggests that amide formation
748 DOI: 10.1021/la1041647
Langmuir 2011, 27(2), 741–750

Azam et al.
Article
Figure 6. (A) Monofunctional amino SAM reacted with a 1:1 molar mixture of 4-trifluoromethyl- and 4-bromo-benzoyl chloride. The
resulting CF3/Br surface ratio was more than three times higher thanthe solution ratio. (B) Monofunctional benzyl azido SAM reacted with a
1:1 molar mixture of 4-trifluoromethyl- and 4-bromo-phenyl acetylene. The resulting CF3/Br surface ratio was more than twice the solution
ratio. These experiments indicate that generating mixed monolayers post SAM formation does not improve the correlation between the
solution and surface ratio.
is sensitive to the local environment around the amine. The for-
mation of the triazoles might sterically block the amine groups,
preventing them from reacting with the p-bromobenzoyl chloride.
In contrast, the amount of fluorine-labeled alkyne was similar
whether the click reaction happened first or second (F/N of 0.79(
0.04 vs 0.78 ( 0.05, respectively). These results suggest that the
alkyne-azide click reaction is not affected by changes in the local
environment around the azide. This similarity in conversion indi-
cates that the high reactivity of the azido monolayer with this
alkyne can compensate for difficulty in accessing the surface sites.
Future work will explore other reactions orthogonal to the click
reaction that are more tolerant of reaction sequence than acyla-
tion by acid chlorides. For example, the reaction of thiols with
maleimides or other olefinic electrophiles should be orthogonally
reactive to the cycloaddition reaction and perhaps less sensitive to
the order of reaction.1
We have synthesized two azide-substituted trimethoxysilanes
that can be directly assembled with aminophenyl trimethoxysi-
lane to yield azide-amine mixed monolayers. Using XPS, we
have demonstrated how to quantify the amount of orthogonally
reactive azides and amines in our monolayers and tune their sur-
face ratio. Through subsequent cycloaddition and acylation reac-
tions, this azide-amine mixed monolayer can be used to anchor
combinations of functional groups to silica that are not necessar-
ily amenable to silane chemistry. Moreover, by controlling the
amine/azide ratio in the SAM, the ratio of subsequent groups
reacted with the azides and amines can be controlled independently
of their surface reactivities. The many examples of surface func-
tionalization reactions based on click chemistry or amine mod-
ification indicate that there are many combinations of molecules
available for forming mixed monolayers with this strategy.16
The high reactivity observed for the azide and amine groups with
different substituted reactive partners suggests that this platform
should be general for attaching a wide variety of molecules.
Furthermore, this strategy of making mixed monolayers on silica
containing orthogonally reactive groups should work with other
reactive pairs besides azides and amines; for example, azides or
alkynes should be orthogonally reactive with other nucleophiles
besides amines like thiols or other electrophilic groups besides
acid chlorides like maleimides, epoxides and N-hydroxysuccini-
midyl esters.1 Finally, our strategy forcharacterizingorthogonally
reactive mixed monolayers containing azides should be accessible
to other substrates besides silica as well as other reactive group
pairings besides azides and amines.14 Current work is aimed at
determining the surface distribution of the two reactive groups
in the orthogonally reactive SAM.
4. Conclusions
Mixed monolayers represent an important avenue for combin-
ing functions in materials and in model systems.4,16 In the field of
self-assembly, many early reports on SAMs demonstrated that
solvent and interactions among assembling molecules could
influence the resulting mixed monolayer composition.20,37 As a
result, the surface ratios do not often reflect the composition of
the assembly solution. Despite this long-held wisdom in SAM
formation, the difference in surface versus solution composition is
often overlooked in bifunctional monolayer systems because of
the difficulty in characterizing the functional group ratio. The
challenge in identifying surface ratios underscores the need for
general strategies, like orthogonally reactive SAMs, that provide
a well-defined starting point for making complex mixed mono-
layers of controlled composition.
Acknowledgment. We gratefully acknowledge the Canada
Foundation for Innovation and the Province of Alberta for
Langmuir 2011, 27(2), 741–750
DOI: 10.1021/la1041647 749

Article
Azam et al.
funding. We also thank the Alberta Centre for Surface Engineer-
ing and Science facility (ACSES) at the University of Alberta
for use of the AFM and water contact angle instruments and
the technicians at ACSES for performing the XPS experiments.
This work was partially inspired by the work of Professor Victor
S.-Y. Lin, who recently passed away. We regret that we never got
the opportunity to meet with him, but we are grateful for his many
contributions to silica chemistry.
Supporting Information Available: Additional IR absor-
bance, XPS, and AFM data. This material is available free of
charge via the Internet at http://pubs.acs.org.
750 DOI: 10.1021/la1041647
Langmuir 2011, 27(2), 741–750