Post-polymerisation modification of surface chemical functionality and its effect on protein binding

Article abstract of DOI:10.1039/c2nj00002d

Derivatisation of polystyrene by carbene insertions followed by diazonium coupling permits the introduction of diverse chemical functionality, providing access to materials with similar bulk properties, but in which surface chemical characteristics are systematically varied across a range of surface polarity, hydration and non-bonding interaction behaviour. Protein binding experiments with bovine serum albumin demonstrate that protein adhesion is dependent upon the identity of the surface chemical group, with tert-butyl, hexyl, dimethylamino, amino, and carboxyl modified systems all exhibiting higher levels of binding, while glycol, hydroxyl, and phosphonate give similar or lower levels of binding, relative to the control. This behaviour has been shown to be time dependent, and an approximate trend of protein binding with cheminformatic descriptors %PSA and contact angle was observed. The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012.

Full text of DOI:10.1039/c2nj00002d

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
NJC
Cite this: New J. Chem., 2012, 36, 1187–1200
www.rsc.org/njc
PAPER
Post-polymerisation modification of surface chemical functionality and its
effect on protein binding
a
b
c
b
a
Cleo Choong, J. S. Foord, Jon-Paul Griffiths, Emily M. Parker, Luo Baiwen,
a
b
Meghali Bora and Mark G. Moloney*
Received (in Montpellier, France) 2nd January 2012, Accepted 10th February 2012
DOI: 10.1039/c2nj00002d
Derivatisation of polystyrene by carbene insertions followed by diazonium coupling permits the
introduction of diverse chemical functionality, providing access to materials with similar bulk
properties, but in which surface chemical characteristics are systematically varied across a range
of surface polarity, hydration and non-bonding interaction behaviour. Protein binding
experiments with bovine serum albumin demonstrate that protein adhesion is dependent upon the
identity of the surface chemical group, with tert-butyl, hexyl, dimethylamino, amino, and carboxyl
modified systems all exhibiting higher levels of binding, while glycol, hydroxyl, and phosphonate
give similar or lower levels of binding, relative to the control. This behaviour has been shown to
be time dependent, and an approximate trend of protein binding with cheminformatic descriptors
%
PSA and contact angle was observed.
9
The control of interfacial properties to modify biocompatibility
and cellular adhesion is of immediate relevance in biomaterial
depending on the layer thickness and surface functionality.
A review of the nature of the effects induced by chemical surface
functionality describes differences in the behaviour of alkyl,
1,2
and drug delivery devices. Previous studies of biomolecule
adhesion have shown that proteins bind most strongly to surfaces
10
hydroxyl, amine and carboxyl groups and the use of tertiary
amine oxide modified surfaces has been shown to resist nonspecific
in the following order: Teflon
4 siliconized glass 4
11
polyvinylchloride 4 Nylon-6,6, and that hydrophobic proteins
protein adsorption. Very recently the effect of the identity of
(
e.g. fibrinogen) adsorb to a greater extent than hydrophilic
surface functional groups on gene transfer efficiency has recently
12
been demonstrated. Overall, while these fundamental studies
3
proteins (e.g. bovine serum albumin). A similar conclusion
4
was made using variously modified titania. However,
have demonstrated the influence of surface chemical functionality
on the interaction of a material with biomolecules, a remaining
challenge is the development of methodology suitable for the
modification of a diverse range of material types which exploits
these phenomena. The significance of precise manipulation of
surface properties can be illustrated by the fact that the inability
to control surface chemistry has been identified as a key limiting
factor in the development of polymer nanofibers as scaffolds for
tissue engineering scaffolds, since no single fabrication technique
has allowed collective control over structure, material composition,
poly(ethylene glycol) grafted onto polystyrene has been shown
to prevent adsorption of fibrinogen and IgG, as well as the
5
adhesion of Streptococcus mutans, and a similar effect has been
6,7
shown on stainless steel. A more detailed examination of
fibronectin binding on methyl, hydroxyl, carboxyl and amino-
terminated self-assembled monolayers (SAM) on a gold surface
has demonstrated that binding affinities of monoclonal anti-
bodies raised against the central cell-binding domain responded
to surface chemical functionality in the order: OH 4 COOH 4
1
3
NH
2
3
4 CH , although there were differences in binding
8
and surface functionality. Surface modification of materials, in
preference for aV and a b integrin. Using phosphotidylcholine
such a way that bulk properties are not modified, has been the
14–16
focus of some attention.
5
1
and hydroxy terminated polymer brushes, the binding force of
bovine serum albumin (BSA) to the modified surface was measured
using an AFM cantilever; this was found to vary from 0.1–1.2 nN
Although a variety of techniques for
surface activation have been developed and applications demon-
1
7
strated, we have found that an approach based upon the reaction
of a wide variety of materials with substituted diaryldiazomethanes,
in which a transient carbene is generated thermally or photo-
a
School of Materials Science and Engineering, Nanyang
Technological University, 50 Nanyang Avenue, Singapore 639798
Department of Chemistry, Chemistry Research Laboratory, The
University of Oxford, 12 Mansfield Road, Oxford, OX1 3TA, UK.
E-mail: mark.moloney@chem.ox.ac.uk
Oxford Advanced Surfaces Group Plc, Begbroke Centre for
18
19
20
chemically; it may be used to introduce colour, fluorescence,
b
21
biocompatibility, biocidal and payload delivery properties to a
22
23
24
14
wide range of substrates, at loadings of the order of 3 ꢀ 10
2 25
molecules per cm . Our approach, which is illustrated in Fig. 1,
c
creates a chemically activated polymer which can then be further
functionalised to the desired surface functionality. The need to
Innovation and Enterprise, Oxford University Begbroke Science Park,
Sandy Lane, Yarnton, OX5 1PF, UK
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1187–1200 1187

View Article Online
Fig. 1 Diagrammatic representation of the surface modification strategy.
understand the three-way relationship between physical, chemical
coupling process led to modified polystyrene polymers 5a–c and
5e–j, possessing terminal tert-butyl, hexyl, iodo, phenyl, glycol,
hydroxyl, phosphonate diester, amine, and carboxylate groups
respectively. Further manipulation of amino 5i and phosphonate
5h gave polymers 5d and 5k–m terminating with dimethylamino,
phosphonate monoester, phosphonic acid and calcium phospho-
nate substituents, respectively. Aside from the inherent flexibility
of this approach, an additional advantage is that successful
diazonium coupling to give the extended diazo-containing chromo-
phore on modified polymers 5a–m as shown in Scheme 1 enables
immediate visual indication by the generation usually of intense red
or orange, or less frequently, yellow colours (Table 1). This
modification could be achieved without physical disruption to the
spherical surface as evident from the SEM images of representative
samples (unmodified polystyrenes, 5a, 5d, 5g; (Fig. 2)), although it
also became apparent that modified samples could become brittle
and excessive abrasion led to onion-peel cleavage of the spheres.
Polystyrenes 3 and 4d (R = H), with no peripheral chemical
functionality, provide useful controls, relative to the other samples.
Detailed analysis of modified polymers 5a–m after extensive
washing was able to confirm successful chemical modification
(Table 1). Thus, the presence of the expected functionality was
readily demonstrated by a combination of ATR FTIR and XPS
analysis, and combustion analysis allowed a direct estimation of
the surface loading levels. Unmodified polystyrene has
strong absorbances at 2931, 1657, 1604, 1446, 1220, 990 and
polymer and functional (e.g. antibacterial and biocompatible)
2
6
polymer properties has been recognised, but of interest to us
was the application of the derivatisation process outlined in Fig. 1
to vary surface chemical functionality at controlled loading levels,
across a range of hydrophobic/hydrophilic and basic/neutral/acidic
groups, to examine the effect of these changes in surface properties
on macroscopic properties and protein binding, and where possible
to establish correlation of such property changes with modern
cheminformatic descriptors of surface chemistry (vide infra). We
used polystyrene beads as a template, since the pattern of protein
adsorption to flat sheets of polymers is reported to be identical to
3
that for adsorption to small polymer particles, and since we have
already shown that this approach is equally applicable to other
19,24
substrates,
the data obtained from this study would be expected
to be generally applicable.
Results and discussion
In order to allow rapid access to functionally variable
polystyrenes, we adapted our previously published approach
for surface modification using diarylcarbenes and its subsequent
2
5
modification by diazonium salt coupling; this is a convenient
and efficient method for the introduction of diverse chemical
functionality onto a material substrate. Thus, bromomethyl-
benzophenone 1 was treated with 2-N-ethylanilinoethanol to give
ether 2a, which was converted to the hydrazone 2b (as a mixture of
syn-/anti-diastereomers) and then oxidised to the diazo compound
ꢁ1 27
902 cm
,
and ATR-IR analysis of modified polymers
provided direct evidence for the surface modification with the
expected functional groups; unfortunately, the diazo absorbance
2c; the success of the latter reaction was immediately evident, as the
ꢁ1
product is a dark pink oil (Scheme 1). This material is not stable to
silica chromatography and is light sensitive, but can be easily stored
for extended periods at 0 1C or lower. Dispersion of this material
on polystyrene XAD in an ether solution followed by careful
solvent removal and heating to 120 1C gave decolourisation,
leading to the modified polystyrene 3; decolourisation is consistent
with loss of the diazo function, and conversion to the reactive
carbene intermediate prior to surface reaction. This material was
then treated with diazonium salts 4a–i, themselves derived from the
required substituted aniline by reaction with isopentyl nitrite (the
presence of the desired diazonium salt was confirmed by application
of the H-acid test (Scheme 2), giving an intense purple colour when
the diazonium function was correctly formed). This two-step
in the 1400–1500 cm region tends not to be well-defined and is
often inactive, and was not observed in any of the materials
28
5a–m. However, bands of interest such as 3400 (O–H), 1723
(C(O)OH), 1600–1585 (aromatic C–C stretch), 1276 (PQO),
1218 (PQO), 1275–1220 (aryl C–O–C stretch), 1075–1020 (alkyl
ꢁ1
C–O–C stretch), 1062–1081 cm (P–O–C), all of which are
associated with the expected functional groups introduced at
the polymer surface, were observed (Table 1). Combustion
analysis for the nitrogen content of the modified polymers
5a–m also confirmed the presence of nitrogen and enabled
determination of the loading stoichiometry of the polymer. The
presence of nitrogen is diagnostic for a successful outcome of the
21,25
two steps leading to derivatised polymers 3 and 5a–m.
1
188 New J. Chem., 2012, 36, 1187–1200
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Scheme 1
Scheme 2
Typical surface loading levels were in the range
found to be 53 and 28% respectively; this drop of efficiency
under alkaline conditions may be due to the hydrolysis condi-
tions eroding the surface of the beads. These data may be further
considered in terms of the density of surface coverage of the
ꢁ1
0
.04–0.36 mmol g and assuming the manufacturer’s value for
2
the surface area of the polymer (725 m g ), this corresponded
ꢁ1
13
ꢁ2
to 0.31–2.97 ꢀ 10 molecules cm (Table 1). These values are
comparable to the loading levels in a related polystyrene system
modified with a dimethoxydiarylmethylene system, determined
layer: the limiting area (A ) for monolayer formation of benzene
o
2
is 0.24 nm per molecule and for substituted anthracenes it is
ꢁ1
12
ꢁ2 25
29
2
30
0.45–0.48 nm per molecule, and assuming a limiting area of
to be 0.04–0.18 mmol g or 2–15 ꢀ 10 molecules cm
,
and
2
31
consistent with other modified polystyrene systems. The
presence of nitrogen species in polystyrene samples 3 and 5a–m
was further confirmed from XPS analysis, in which N1s signals
at binding energies (eV) of 397.7, 398.5, 399.2, 401.6 and 402.7 eV
were observed, with the observed N/C ratios comparing well with
the expected values, assuming the formation of a uniform
monolayer of benzhydryl units of the indicated molecular
formula for modified polymers 5a–m (Scheme 1 and Table 1).
Moreover, the presence of phosphorus in samples 5h, k–l was
confirmed by combustion analysis (Table 2), and comparison of
the P versus N content values permitted estimation of the
efficiency of the final diazonium step in Scheme 1 to be typically
in the range of 72–76%. When blank polystyrene is treated
directly with the diazonium phosphonate 4g followed by hydro-
lysis under acid or alkaline conditions, the coupling efficiency is
1.71 nm per molecule as calculated using Chemicalize for the
cross-sectional area of a typical surface modifying molecule, it is
possible to estimate the surface area coverage as being in the
range 6–25% of the total surface area (Table 1). Expressed in a
different way, the ratio of unmodified to modified sites (N) on the
polymer surface has been calculated to be in the range of 50–200;
this loading density is better than the values in a related system
prepared by dimethoxydiarylmethylene insertion into
22,25
polystyrene.
To confirm that the chemical modification resulting from
the application of the process in Scheme 1 only results in
surface modification, without degradation of the bulk of the
material, DSC and TGA analysis were performed (see Table 1);
the measurable inflection temperature was found to be unaffected
in all the samples relative to unmodified polystyrene XAD, with
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1187–1200 1189

View Article Online
Table 1 Characterisation data for polymers 5 modified according to Scheme 1
e
Combustion analysis %N
XPS N/C found
ꢁ1
mmol g
molecules cm
Area covered
ꢁ
2a
d
(N/C expected )
b
%
c
ꢁ1
c
N
Material and appearance
N
ATR IR/cm
DSC T/1C
f
Polystyrene XAD-4
—
2931, 1657, 1604, 1446, 1220, 990, 902
—
104.4
3
5
(Carbene modified)
0.25
0.056
(0.040)
0
1
2
.179
1
3
.48 ꢀ 10
1601, 1511, 1252, 1169, 1090
104.6
103.8
2
5.4
5
0
a R = t-Bu
0.26
0.086
(0.048)
0
5
8
.062
.. 81 4 ꢀ 10
12
1603, 1513, 1354, 1252, 1178, 1085
4
1
50
5
5
5
b R = C
6
H
13
0.28
0
.020
0
5
9
1
.067
1
2
(0.081)
.. 55 4 ꢀ 10
38
1326, 1602, 1080
104.4
118.8
103.9
1
7
c R = I
1.5
0.041
(0.097)
5
0
2
2
5
.357
13
.97 ꢀ 10
1601, 1377, 1286, 1075, 547
0.4
1
d R = NMe
2
0.21
0
.12
0
3
5
2
.038
12
(0.042)
9
.. 31 2 ꢀ 10
51
1601, 1511, 1351, 1170, 1086
5
5
e R = H
0.7
0
.074
0
1
2
5
.167
1
3
(0.097)
2
.38 ꢀ 10
1602, 1511, 1316, 1195, 1092
104.6
104.6
3.7
2
f R = O(CH
2
CH
2
O) CH
3
3
0.52
0
.044
0
1
1
.124
13
(0.079)
.03 ꢀ 10
1603, 1511, 1362, 1252, 1085
1
9
7.6
7
1
5g R = OH
0.24
0.031
0
4
8
1
.057
12
(0.097)
.. 17 5 ꢀ 10
61
3400, 1603, 1512, 1252, 1169, 1086
103.2
1
1
1
190 New J. Chem., 2012, 36, 1187–1200
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Table 1 (continued )
e
Combustion analysis %N
XPS N/C found
ꢁ1
mmol g
molecules cm
Area covered
ꢁ
2a
d
(N/C expected )
b
%
c
ꢁ1
c
N
Material and appearance
N
ATR IR/cm
DSC T/1C
5
5
5
5
5
h R = CH
2
P(O)(OEt)
2
0.34
0
.038
0
1
1
.121
.01 ꢀ 10
13
12
12
13
13
12
(0.083)
8
1602, 1276, 1217, 1191, 1062
103.6
7.2
7
2
i R = NH
2
0.26
0
.026
0
3
6
.046
.. 68 6 ꢀ 10
(0.129)
3404, 3280, 1602, 1328, 1217, 1062
3020, 1793, 1723, 1602, 1329, 1085
2606, 1602, 1298, 1081, 1071
104.4
104.3
105.0
3
3
2
02
j R = CH COOH
2
0.22
0
.041
0
4
7
.052
.. 43 5 ꢀ 10
(0.091)
7
1
78
k R=CH
2
P(O)(OEt)(OH)
0.63
0
.033
0
1
2
.150
(0.088)
8
.25 ꢀ 10
1.3
5
8
l R = CH P(O)(OH)
2
2
0.61
0
.050
0
1
2
.145
(0.094)
3
.21 ꢀ 10
2601, 1602, 1352, 1276, 1251, 1170, 1083, 1018
104.9
103.9
0.6
6
0
5m R = CH
2
P(O)(O)
2
Ca
0.32
0
.039
0
6
.076
.33 ꢀ 10
2
625, 1601, 1511, 1252, 1170, 1084
(0.094)
6
1
1
0.8
20
a
2
Calculated on the basis of 725 m g
ꢁ1
b
2
A limiting area of 1.71 nm per molecule is assumed as calculated using Chemicalize (www.chemicalize.
.
com). N is the ratio of unmodified sites to modified sites calculated using the combustion analysis or XPS data. Expected values calculated from
c
d
e
molecular formulae assuming a surface monolayer of carbene insertion product. Ratio taken from nitrogen and carbon counts in a survey
f
spectrum. For a thorough list of surface properties of unmodified polystyrene, see http://www.accudynetest.com/polymer_charts.html.
values of 103–105 1C being observed. This transition probably
corresponds to the T of low levels of residual non-cross-linked
With this library of diversely functionalised polystyrenes 3
and 5a–m in hand, an assessment of the changes to macro-
scopic behaviour was made. In order to permit a more
quantitative analysis of the outcomes of this work, an initial
calculation of the value for several molecular characterisation
parameters which are increasingly used to guide the drug
g
material (AMBERLITEt XADt4 is a polymeric adsorbent
based on highly cross-linked, macroreticular polystyrene
3
2
(
http://www.dow.com/products)), since polystyrene without
any additives (M Q374 000, M /MnQ2.5) has been reported
to have T
-values in the range 107 1C ꢂ 2 K. The influence of
polystyrene blending on T
linking of polystyrene has been reported to increase the T
w
w
33
36
g
discovery process, including polarisability, calculated molecular
34
g
has been investigated and cross-
value
refractivity (CMR), log P, polar surface area (PSA), molecular
g
surface area (MSA) and molecular volume, was made using
31
Marvin (www.chemaxon.org, accessed via chemicalize.org),
35
from 98 1C to 125 1C. It was found that the TGA degradation
temperature (all samples have degradation temperatures between
and the relevant data are included in Table 3 (columns 2–7).
Polarisability is a measure of induced charge resulting from polar
470–476 1C) was unaffected after surface modification.
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1187–1200 1191

View Article Online
Table 2 Combustion analysis for P and N for selected substrates
Measured
%
P/ %N/ Expected % Efficiency
1
ꢁ
mmol mmol mmol g for diazonium
ꢁ
1
ꢁ1
a
b
Material Functionality
g
g
I or P
coupling
5k
R =
CH P(O)(OH)(OEt)
0.11 0.63 0.050
72
2
(
(
0.036) (0.150)
0.17 0.61 0.048
5
l
R =
CH P(O)(OH)
75
76
2
2
0.055) (0.145)
5m
R = CH
2
2
P(O)O Ca 0.06 0.49 0.025
(
0.019) (0.076)
a
b
Expected %P = measured %N/3. Efficiency = actual P/expected P.
interactions, and is calculated by the method of Miller and
3
7
Savchik. CMR, calculated using the method of Ghose and
38,39
Crippen,
correlates molecular refractivity with the molecular
volume and therefore London dispersive forces, which are
important in molecular interactions. clog P is the octanol/water
39
partition coefficient, calculated using the method of Ghose et al.,
and is a measure of the lipophilicity of a compound. The polar
surface area parameter (PSA), which correlates the presence of
polar atoms with membrane permeability and therefore gives an
4
0
indication of drug transport properties, has been reported to
2
˚
have an optimal value of 70 o PSA o 120 A for a non-CNS
orally absorbable drug. MSA is the van der Waals molecular
41
42
surface area, and is calculated using the method of Ferrara et al.
Although strictly applicable to molecules in solution, given the
length of the linking tether which permits a high degree of freedom
in the modified polymers 5a–m, this analysis nonetheless provides a
useful measure of the types of surface property likely to impact on
macroscopic behaviour of the modified polystyrene beads. In this
analysis, the steadily decreasing lipophilic and increasing hydro-
philic character of modified materials 5a–m was confirmed
from their decreasing clog P values and increasing %PSA values
(
Table 3). The calculated van der Waals molecular surface area of
2
the modifying surface functionality is in the range 870–1170 A ,
42
˚
the difference arising from the various lengths of the functional
group R of the common structural skeleton in 5a–m (Table 3 and
Scheme 1). This is also reflected in the calculated molecular volume,
3
˚
which is in the range 540–700 A (Table 3).
With this data in hand, an analysis of the macroscopic
properties of modified polystyrenes was made; thus, water
contact angle and Zeta potential (x) were measured. Contact
4
3–46
angle behaviour can be complex,
and polystyrene has been
surface functionalisation of polystyrene
with polar groups and poly-L-lysine has been shown to lead to
4
7,48
thoroughly studied;
5,48,49
the significant reduction of contact angle.
In the case of
the modified polystyrene beads prepared herein, the curved
nature of the beads required special analysis for contact angle,
and this was achieved by using ‘‘floating’’ contact angle measure-
5
0
ments, using the literature method. For these measurements,
the polystyrene microspheres were randomly chosen from each
sample group and carefully placed on the surface of a (deionized)
water droplet contained on a clean glass slide. The wetting
behaviour (i.e. contact angle at the microparticle/water interface)
was then measured in side elevation, and the data are given in
Table 3. This investigation indicated that most materials exhibited
1
192 New J. Chem., 2012, 36, 1187–1200
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
5
1
surfaces proposed by Hoffman 25 years ago. Surface tension
values were derived from contact angle by use of methodology
44,52
reported by Boury et al.
Measurement of the Zeta potential
(x) of modified particles was also made, but was complicated by
difficulties with the dispersion of the particles. However, by using
an acetone-deionised water mixture (11% acetone-deionised water
with pH adjusted to B7.2), suitable particle dispersion could be
achieved. The data obtained (column 7, Table 3 and Fig. 3(b)
and(c)) indicated that surface modification led to significant
alteration of surface charge in most cases relative to unmodified
polystyrene XAD (ꢁ27.8 mV); a broad correlation of Zeta
potential with %PSA and clog P was observed, since increasing
%PSA led to decreasing Zeta potential, while increasing clog P
gave increasing Zeta potential. In sum, more polar chemical
functionality at the surface is reflected in more hydrophilic
macroscopic behaviour at the surface interface, as might be
expected, and since the impact of changes of Zeta potential on
protein (BSA) adsorption on hydroxyapatite has been demon-
strated, similar effects might be anticipated with these modified
53
polystyrenes.
In order to assess the effect of these polymer modifications
on protein adsorption, bovine serum albumin (BSA) adsorption,
chosen as a readily available protein, after 3 hours of incubation
was measured; adsorption was assessed by isolating and washing
the incubated beads, followed by a spectrophotometric assay of
the protein after desorption using sodium dodecyl sulfate solution.
The raw protein binding values are listed in Table 3 (column 11),
but correction of these values by allowance for the differing
functional group loading levels permitted calculation of adjusted
ꢁ
1
protein adsorption (mg protein/mg polymer per mmol g of
polymer chemical functionality loading, Table 3, column 13).
No significant correlation of protein adsorption with several
parameters could be observed, although increasing values of
contact angle, surface tension, and %PSA, and decreasing Zeta
potential were generally reflected in higher levels of adjusted
protein adsorption (Fig. 4a–d). Neither was correlation of mole-
cular volume with adjusted protein adsorption observed (Fig. 4e).
This cheminformatic analysis, based as it is on various measures of
polarity or hydrophobicity/philicity, does not allow for the known
binding of specific chemical functional groups to BSA. In fact,
54,55
plasma protein binding is well studied
and it has been reported
that small molecules, in particular carboxylate, halogens and alkyl
groups, strongly bind to HSA-3A while anilines do so only
56–59
weakly.
Similarly, BSA is known to exhibit strong binding
behaviour with carboxylic acids, with binding activity increasing
6
0
a
with increasing pK . A study of the adsorption of bovine serum
albumin (BSA) at the free water interface using surface tension
measurements has been reported, parameterised using only bulk
61
Fig. 2 SEM images of modified beads: (i) unmodified polystyrene
XAD; (ii) 5a; (iii) 5d; (iv) 5g.
diffusion and that chemical functionality can reduce the binding
of BSA and fibrinogen has been demonstrated by the incorpo-
ration of poly(L-lysine)-g-poly(ethylene glycol) layers onto metal
6
2
altered contact angle relative to unmodified polymer (521), with
increasing contact angle broadly correlating with increasing clog
P (columns 2 and 8, Table 3 and Fig. 3(a)). However, although no
meaningful correlation of polarisability, CMR or %PSA with
contact angle could be observed, more polar groups generally
gave a material which was more hydrophilic, as might be expected
oxide surfaces. Therefore, of interest are the high binding
values for the hexyl, amine, carboxyl, phosphonate mono ester
and calcium phosphonate salt terminated systems 5b,i,j,k,m;
conversely, the weak binding of iodo 5c, dimethylamino 5d, glycol
5f, diethyl phosphate 5h and phosphonate 5l is also noteworthy.
The weak binding of modified system 5f is consistent with the
known low protein binding behaviour of PEG coatings, as a result
(e.g. polymers 5l,m); this outcome is consistent with a qualitative
6
classification scheme for hydrophilic and hydrophobic biomaterial
of the formation of a tightly bound water layer. Thus, the
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1187–1200 1193

View Article Online
Fig. 3 Plots of (a) contact angle versus c log P; (b) Zeta potential versus %PSA; and (c) Zeta potential versus c log P for modified polymers 5a–m
(data from Table 3).
2
preserved, albeit more weakly, at t = 12 h (R = 0.5168), but lost
behaviour of the modified polymers 5a–m is broadly consistent
with the known functional group binding behaviour of
serum albumins, but noteworthy differences are amine 5i and
phosphonate 5l.
2
at t = 24 h (R = 0.3641), consistent with surface saturation of
hydration and/or protein binding over this time period. Of interest
is the increase in the slope of the lines of best fit for correlations of
contact angle and protein binding (Fig. 6(a), (c) and (e)) as the
system approaches saturation, and that in Fig. 6(a), (c) and (e),
there are two groups of chemical function, the more polar group-
ing at the top, for which protein binding is preserved over the time
course of the experiment, and the non-polar at the bottom, for
which protein binding decreases over time. Within the time frame
of the experiment, the aniline 3, glycol 5f and hydroxy 5g modified
surfaces exhibit similar behaviour for protein binding, and this
may be due to the presence of a significant hydration layer, which
is not present in the hydrophobic modifications containing
tert-butyl and hexyl groups 5a,b. By 24 hours, two distinctive
groups have formed; the low-binding group which includes the
aniline, glycol, hydroxy and phosphonate diester modified
polymers, and the high-binding group which is made up of the
more hydrophobic alkyl terminated polymers hexyl and tert-butyl
as well as dimethyl amine, amine and acid terminated groups. The
alkyl terminated polymers will have low (+) DG dehydration
values and would be expected to bind proteins using van der
Waals London forces in a separate binding site to that which
supports the more polar surfaces. tert-Butyl and dimethylamine
modified surfaces have very similar adsorption levels, and this
may reflect similar steric interactions with the protein binding site.
The acid and amine are the two that do not fit in with this group
More detailed analysis in which time course measurements
of the levels of protein adsorption were made at 1, 3, 6, 12 and
2
4 h incubation times was performed; in this case, adsorption
was assessed by the spectrophotometric assay of aliquots of
the supernatant protein solution taken at various time points,
and calculation of adsorbed protein by difference. The resulting
data are summarised in Table 4 and Fig. 5(a) and 5(b). The
former gives raw data and the latter, protein adsorption adjusted
with surface loading density (data taken from Tables 1 and 3). It
is immediately apparent that functional group and time variant
behaviour is observed; thus, tert-butyl 5a, hexyl 5b, dimethyl-
amine 5d, amine 5i, and carboxyl 5j all exhibited levels of protein
adsorption which were 2–3 times higher relative to the aniline-
modified polymer 3 control, but that glycol 5f, hydroxyl 5g, and
phosphonate diethyl ester 5h were similar to or gave only slightly
increased binding relative to the control aniline 3; this outcome is
similar to that observed in the single point time measurements, the
data for which are given in Table 3. However, most samples
exhibited complex protein binding behaviour in the time domain,
and some systems displayed either increasing or decreasing protein
binding over time (e.g. hexyl 5b and hydroxyl 5g respectively).
Initial binding broadly linearly correlated with contact angle at
2
t = 3 h (Fig. 6(a)) with an R value of 0.7425, and this trend was
1
194 New J. Chem., 2012, 36, 1187–1200
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Fig. 4 Plots of (a) contact angle versus adjusted protein adsorption; (b) adjusted protein adsorption versus surface tension; (c) %PSA versus
adjusted protein adsorption; and (d) molecular volume versus adjusted protein adsorption; and (e) adjusted protein adsorption versus zeta potential
for modified polymers 5a–m (data from Table 3).
Table 4 Protein binding data for selected modified polystyrenes
in terms of shape and hydrophobicity; this is likely to reflect
their known binding to alternative binding sites on BSA
63,64
Normalized %protein adsorption
ꢁ1
via electrostatic interactions. Noh and Vogler et al.
break
(
%/mmol g
1/h 3/h
42.3 32.3
)
the Gibbs energy of protein adsorption down into three different
components: (i) DG hydrophobic effect, which can be described as
the energy gained as the protein comes out of solution and the
water forms hydrogen bonds with itself, and this can be assumed
to be constant; (ii) DG dehydration, which is the energy required
to displace the water adsorbed onto the polymer surface (as the
hydration of the surface increases protein adsorption decreases
due to the protein having to dispel the hydrated layer in order to
adsorb onto the surface); (iii) DG interaction, which considers the
affinity of the protein for the surface and energy gained in forming
Samples
6/h
26.6
82.0 136.7
92.1 113.8
169.8 207.0
12/h
24/h
3
5
5
5
5
5
5
5
5
Carbene modified
a R = t-Bu
22.7
126.8
137.7
157.4
38.5
43.74
91.19
117.5
168.9
33.7
119.2
144.0
115.6
38.8
44.6
53.5
132.6
165.2
121.1
88.1
113.6
60.47
104.5
72.6
b R = C
d R = NMe
6 13
H
2
f R = O(CH
g (OH)
2
CH
2
O) CH
3
3
61.9
33.6
27.8
54.37
98.97
86.1
h R = CH
i R = NH
j R = CH
2
P(O)(OEt)
2
2
106.8
117.4
82.1 169.0
148.7 194.0
2
COOH
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1187–1200 1195

View Article Online
Fig. 5 Bar chart illustrating normalised %protein adsorption by functional group.
the protein surface interactions by electrostatic, hydrogen bonding
or van der Waals forces. The latter may be more important over
time and may explain why the amine, acid and hydroxyl materials
bind more protein initially than other functional groups.
included in Table 3, are best accounted with two trends,
enclosed within the indicated circles, one for non-polar and
the other for polar groups defined by their %PSA, with the
less polar surfaces between 1.7–4.2% and the more polar
between 6.4–7.8%, and this may be indicative of binding at
On the other hand, the data in Fig. 6(b) and (d), which
correlate protein binding with cheminformatic parameters
5
4–60
the two known binding pockets of BSA.
By 24 hours
1
196 New J. Chem., 2012, 36, 1187–1200
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Fig. 6 Protein adsorption with various parameters at different time points. (a) Protein binding with contact angle at 1 h; (b) protein binding with
%PSA at 1 h; (c) protein binding with contact angle at 12 h; (d) protein binding with %PSA at 12 h; (e) protein binding with contact angle at 24 h;
(f) protein binding with %PSA at 24 h.
(
Fig. 6(f)), increased segregation into three groups is observed:
Significantly, we have shown that systematic controlled
modification of chemical functionality at polymer surfaces is
possible, and that this results in changes in macroscopic polymer
behaviour such as wettability and protein binding; trends in this
behaviour with cheminformatic descriptors of surface function-
ality can be made. Thus, contact angle and Zeta potential vary
with clog P while an approximate correlation of protein binding
with %PSA and contact angle was observed. This surface
functionality may be directly responsible for protein binding, or
may modify surface hydration leading to a hydrophilic surface
layer. These results are of significance, since for example the
control of fouling by surface modification has recently been
phosphonate diester, hydroxyl and glycol comprise one, most
likely as a result of water binding to the surface preferentially;
hexyl, tert-butyl and dimethyl amine form the second as a
result of their common van der Waals binding mechanism; and
finally the acid and amine due to their greater affinity for the
protein or stronger binding ability via electrostatic inter-
actions. The data presented here suggest that protein binding
may be strongly mediated by surface chemical functionality,
with significantly enhanced binding for tert-butyl groups 5a,
hexyl 5c, dimethylamino 5d and acid 5j groups relative to the
glycol function 5f, which appears to broadly correlate
with %PSA.
65
highlighted as a critical element in the design of new materials.
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1187–1200 1197

View Article Online
Experimental
(4.95 g, 17.8 mmol) in THF (20 ml) was added, the mixture warmed
to room temperature and then left to stir for 72 hours. The THF
was removed in vacuo, and the reaction quenched with sat. sodium
hydrogen carbonate (20 ml). The resulting mixture was washed
with DCM (15 ml), and the organic phase was washed with water
Instrumentation and analytical methods
SEM protocol. A scanning electron microscope (JSM6360A,
JEOL) was used to observe the surface morphology of micro-
particles. Dry microparticles were coated with gold and observed
microscopically under different magnifications.
(
2 ꢀ 20 ml) then dried over MgSO , filtered and concentrated to a
4
dark green oil (8.32 g). Purification via column chromatography
(
SiO
2
, eluent 9 : 1 petrol (40–60):EtOAc) gave product 2a as a
Contact angle protocol. For contact angle measurements, the
polystyrene microspheres were randomly chosen from each
sample group and placed on the surface of a water droplet
contained on a clean glass slide. They were imaged using a
yellow oil (4.59 g, 72%). R 0.45 in 6 : 1 petrol (40–60):EtOAc;
f
ꢁ1
vmax(film)/cm 2970 (N–H), 1820, 1650, 1625; dH (400.2 MHz;
CDCl ) 1.21 (3H, t, J 7.1, NCH CH ), 3.48 (2H, q, J 7.1,
3
2
3
NCH CH ); 3.61 (2H, t, J 6.2, OCH CH N), 3.73 (2H, t, J 6.2,
2
3
2
2
50
contact angle analyzer (FTA32 Video 2.0, First Ten Angstroms).
Deionized water was used from a Milli-Q system.
OCH CH N), 4.65 (2H, s, ArCH O), 6.73 (3H, m, ArCH 2 ꢀ ortho
2
2
2
and 1 ꢀ para to NEt), 7.26 (2H, t, J 8.0, ArCH meta to NEt),
.46–7.53 (4H, m, 4 ꢀ ArCH meta to CQO), 7.62 (1H, dd, J 7.4,
7
Zeta potential protocol. The electrophoretic mobilities of
7
.4, ArCH para to CQO), 7.83 (4H, d, J 7.3, 4 ꢀ ArCH ortho to
microparticles were measured using a Zeta potential analyzer
CQO);
NCH CH
ArCH
d
C
(100.6 MHz; CDCl ) 12.2 (NCH CH ), 45.5
3
2
3
(Nano-ZS, Malvern Instruments). The particles were suspended
(
2
3
), 50.1 (OCH
2
CH N), 68.5 (OCH
3
CH N), 72.24
2 3
in 11% acetone in deionized water at pH 7.18 adjusted using 1 N
NaCl and HCl. The electrophoretic mobility was measured at
room temperature with 3 runs (12 times in each run) for each
sample.
(
2
O), 111.8 (2 ꢀ ArCH ortho to NEt), 115.8 (ArCH para
to NEt), 127.1, 128.3, 129.4, 130.1 and 130.3 (2 ꢀ ArCH meta to
NEt and 2 ꢀ ArCH meta to CQO), 136.8 and 137.7 (2 ꢀ
2
ArC(CQO)), 143.2 (ArCCH O), 147.7 (ArCNEt), 196.4 (CQO);
+ +
m/z (ESI ) 360.2 ([M + H] , 100%).
XPS protocol. The surface functionalization of microparticles
was analyzed using X-ray photoelectron spectroscopy (XPS) mea-
surements. All XPS spectra were measured using a Kratos Axis
Ultra spectrometer with a monochromatic Al–Ka (1486.71 eV)
X-ray source. The surface wide scans were recorded with a pass
energy of 160 eV and elemental narrow scans were obtained with
N-Ethyl-N-(2-((4-(hydrazono(phenyl)methyl)benzyl)oxy)ethyl)-
1
9
aniline 2b . Hydrazine monohydrate (2 ml, 4.2 mmol) was
added to a solution of (4-((2-(ethyl(phenyl)amino)ethoxy)methyl)-
phenyl)(phenyl)methanone 2a (3.1 g, 8.6 mmol) in EtOH (15 ml)
and the mixture was set to reflux. Analysis by mass spectrometry
after 16 hours showed complete loss of the benzophenone, so the
reaction mixture was cooled to room temperature. Water (5 ml) was
added and the EtOH removed in vacuo. DCM (10 ml) was added to
the residue and the two phases separated. The organic phase was
4
0 eV. Casa XPS software was used for curve-fitting analysis
of C1s, N1s and O1s with the calibration of binding energy of
84.5 eV for C1s.
2
TGA analysis. Thermogravimetric analysis (TA Instruments
Q2950 and Q500) was used to measure the thermal stability
washed with water (3 ꢀ 10 ml), dried over MgSO
4
and concen-
trated in vacuo, to yield product 2b as a 1 : 1 mixture of two
d
(decomposition temperature (T )) of the polystyrene micro-
ꢁ1
diastereoisomers as a colourless oil (2.80 g, 87%). vmax(film)/cm
particles. Approximately 10 mg of particles were weighed in a
platinum TGA pan, which is placed in the high precision balance
and heated at a ramp rate of 10 1C per minute over a temperature
range of 600 1C in an atmosphere of nitrogen—40% and
air—60%.
3
405 br, 2865 br, 1595; d (400.2 M Hz; CDCl ) 1.14–1.22 (3H, m,
H
3
NCH CH ), 3.40–3.55, 3.59–3.67 and 3.73–3.76 (6H, 3 ꢀ m,
2
3
OCH
5.45 (2H, br s, NH
ArCH), 7.46–7.57 (4H, m, ArCH); d
and 12.2 (NCH CH ), 45.4 and 45.5 (NCH
(OCH CH N), 68.0 and 68.4 (OCH CH N), 73.0 and 73.1
2
CH
2
N and NCH
), 6.65–6.75 (3H, m, ArCH), 7.20–7.32 (7H, m,
(100.6 MHz; CDCl ) 12.2
CH ), 50.1
2
CH
3
), 4.53 and 4.62 (2H, s, 2 ꢀ ArCH
2
),
2
C
3
DSC analysis. DSC thermograms of polystyrene microparticles
2
3
2
3
were obtained using a differential scanning calorimeter (DSC Q10,
2
2
2
2
TA Instruments) to determine the glass transition temperature (T
g
).
(ArCH O), 111.7 and 111.8 (2 ꢀ ArCH ortho to NEt), 115.7 and
2
Approximately 5 mg of particles were weighed in an aluminium
pan and heated at a ramp rate of 10 or 20 1C per minute over a
temperature range of 25 1C to 260 1C in an atmosphere of nitrogen.
115.8 (ArCH para to NEt), 126.5, 126.5, 127.4, 128.1, 128.5, 128.8,
128.9, 129.3 and 129.4 (ArCH), 132.2 and 132.9 (ArCCH O), 137.9,
2
138.2, 138.4 and 139.2 (ArCCQNNH ), 147.7 (ArCNEt), 149.0
2
+
+
T
g
was calculated using the software by plotting two tangent lines
and 149.0 (CQN); m/z (ESI ) 374.2 ([M + H] , 100%).
above and below the transition and the temperature midway
between the lines was considered as Tg.
N-(2-((4-(Diazo(phenyl)methyl)benzyl)oxy)ethyl)-N-ethylaniline 2c.
To a stirring solution of N-ethyl-N-(2-((4-(hydrazono(phenyl)-
methyl)benzyl)oxy)ethyl)aniline 2b (1.04 g, 2.78 mmol) in methanol
(34 ml), manganese(IV) oxide (0.61 g, 6.96 mmol), potassium
hydroxide (0.16 g, 2.78 mmol) and sodium sulfate (0.56 g,
3.94 mmol) were added. The mixture was stirred for 3 hours when
analysis by mass spectrometry showed complete consumption of
the starting hydrazone. The methanol was removed in vacuo, and
DCM was added (30 ml). The mixture was filtered through a
Celite plug, and concentrated to give product 2c as a dark pink oil
Synthetic preparations
(
4-((2-(Ethyl(phenyl)amino)ethoxy)methyl)phenyl)(phenyl)-
1
9
methanone 2a . The following procedure was carried out under
an inert atmosphere using dry solvents. A solution of 2-(N-ethyl-
anilino)ethanol) (3.32 g, 20.0 mmol) in THF (20 ml) was added to
sodium hydride (0.74 g, 31.0 mmol) in THF (5 ml) at ꢁ78 1C. The
mixture was warmed to room temperature then stirred for 30
19
minutes and re-cooled to 0 1C. 4-Methylbromobenzophenone
1
198 New J. Chem., 2012, 36, 1187–1200
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
ꢁ
1
(
0.91 g, 89%). vmax(film)/cm 2865, 2040 (NQNQN), 1660,
3.05 (2H, d, JHP 20.9, 2 ꢀ ArCH
7.1, 4 ꢀ P(O)(OCH CH ), 6.64 (2H, d, JHH 8.4, ArCH ortho to
NH ), 7.08 (2H, dd, J 8.4, JHP 2.4, ArCH meta to NH ); dC
2
P(O)), 3.95–4.05 (4H, m, JHH
1
3
3
6
505; d
H
(400.2 MHz; CDCl ) 1.18 (3H, t, J 7.0, NCH
3
2 3
CH ),
2
3 2
)
.45 (2H, q, J 7.0, NCH CH ), 3.57 (2H, t, J 6.4, OCH CH N),
2
2
3
2
2
HH
2
.69 (2H, q, J 6.4, OCH CH N), 4.55 (2H, s, ArCH O),
2
(100.6 MHz; CDCl ) 16.4 (d, J 6.0, 2 ꢀ P(O)(OCH CH ) ), 32.7
2
2
3
CP
2
3 2
.66–6.73 (3H, m, 2 ꢀ ArCH ortho to NEt and 1 ꢀ ArCH para
(d, JCP 138.8, CH P(O)), 62.0 (d, J 7.0, 2 ꢀ P(O)(OCH CH ) ),
2
CP
2
3 2
to NEt), 7.19–7.25 (3H, m, 2 ꢀ ArCH meta to NEt and 1 ꢀ
ArCH para to C–N
), 7.27–7.32 (4H, m, 4 ꢀ ArCH ortho to
C–N ); d (100.6
), 7.36–7.43 (4H, m, 4 ꢀ ArCH meta to C–N
MHz; CDCl ) 12.2 (NCH CH ), 45.4 (NCH CH ), 50.1
OCH CH N), 68.0 (OCH CH N), 73.0 (ArCH
115.3 (d, JCP 3.0, ArCH ortho to NH ), 121.0 (d, J 10.1,
2 CP
2
ArCCH
ꢁ ꢁ
2
(ArCNH ); m/z (ESI ) 242.10 ([M–H] , 100%).
2
), 130.6 (d, JCP 7.0, ArCH meta to NH
2
), 145.2
2
2
C
3
2
3
2
3
Synthesis and coupling of diazonium salts with modified
polymer 3 to give polymers 5a–m. The diazonium salt
(12 w/w%) was prepared based on the mass of the beads to be
modified by taking the relevant amine (1 eq.), isopentyl nitrite
(
2
2
2
2
2
O), 111.7 (2 ꢀ
ArCH ortho to NEt), 115.7 (ArCH para to NEt), 125.2 (ArCH
0
para to C–N
2
), 125.6 (2 ꢀ ArCH meta to NEt), 128.6 and 128.9
), 129.1 and 129.3 (4 ꢀ ArCH meta to
(ArCH ortho to C–N
2
(1 eq.) (NB: Care! Heart stimulant)* and tetrafluoroboric acid
C–N ), 129.6 (2 ꢀ ArCCN )), 135.7 (ArCNEt), 147.7 (CQN);
2
2
+
m/z (ESI +) 372.2 ([M + H] , 100%).
(2 eq.) (NB: in the case of 1,4-phenylenediamine, 1.1 eq. of
isopentylnitrite and 2 eq. of tetrafluoroboric acid were used) and
by stirring in EtOH at 0 1C for 2 hours. The presence of the
desired diazonium salt was confirmed using the H-acid test
Modification of polystyrene XAD-4 beads with N-(2-((4-(diazo-
(
phenyl)methyl)benzyl)oxy)ethyl)-N-ethylaniline 2c to give polymer
. The supplied XAD-4 beads were washed with copious
amounts of water, acetone and then dried. N-(2-((4-(Diazo-
phenyl)methyl)benzyl)oxy)ethyl)-N-ethylaniline 2c (5 w/w%,
2
2
3
(see below). The mixture was added to a vial containing the
modified beads 3 and EtOH was added such that the beads were
completely immersed. The bead salt mixture was left to stand for
18 hours in a fridge at 5 1C. The modified beads 5a–m were
filtered from the mixture, washed with water and EtOH and then
left to dry on a sinter funnel under vacuum.
(
relative to bead weight) was dissolved in ether and the Amberlite
XAD-4 beads added. More ether was added such that the beads
were completely submerged, and the mixture was carefully
concentrated to dryness in vacuo. The beads were heated at
H-acid test. A sample of the diazonium salt solution (1 ml) was
1
20 1C until the beads had turned from pink to off yellow/white
colour. The product beads 3 were washed with acetone on a sinter
adjusted to pH 4 using sodium acetate and then H-acid (4-amino-
5-hydroxy-2,7-naphthalene disulfonic acid) was added (10 mg).
funnel and dried under vacuum.
The mixture was mixed thoroughly and left to stand at room
temperature for 15 minutes. The presence of a diazonium salt is
indicated by a dark purple colour (Scheme 2). An alternative
‘‘spot test’’ procedure is as follows: drops of the suspension were
placed onto a filter paper soaked in ‘H-acid’ solution. If the
diazonium species had been generated, intense colour formation is
instantaneous.
6
6
Diethyl (4-nitrobenzyl)phosphonate. 4-Nitrobenzyl bromide
1.09 g, 5.05 mmol), tetrabutylammonium iodide (132 mg,
(
0
.36 mmol) and triethylphosphite (2 ml, 11.7 mmol) were stirred
together at 120 1C for 5 hours. The mixture was cooled and
purified by column chromatography on silica (eluent 4 : 1 EtOAc :
petrol) to yield a yellow oil (1.07 g, 78%). R 0.27 (4 : 1 EtOAc:
f
ꢁ1
petrol (40–60)); vmax(neat film)/cm 3470, 2985, 1520, 1350, 1250
PQO), 1025, 965 (P–O); d (400.2 MHz; CDCl ) 1.20 (6H, t,
HH 7.1, 6 ꢀ OCH CH ), 3.20 (2H, d, JHP 22.4, 2 ꢀ ArCH P),
.96–4.03 (4H, m, 4 ꢀ OCH CH ), 7.42 (2H, dd, JHH 8.7, JHP 2.3,
ArCH meta to NO ), 8.11 (2H, d, JHH 8.7, ArCH ortho to NO );
(100.6 MHz; CDCl CH ), 33.8
) 16.3 (d, JCP 6.0, 2 ꢀ OCH
d, JCP 136.1, 2 ꢀ ArCH P), 62.4 (d, JCP 7.0, 2 ꢀ OCH CH ),
23.6 (d, J 3.0, 2 ꢀ ArCH meta to NO ), 130.6 (d, J 6.0, 2 ꢀ
Protein adsorption on microparticles
(
H
3
J
2
3
2
Protein adsorption on microparticles (single time point).
Protein adsorption was determined using bovine serum albumin
(BSA) with the modified microparticle surfaces. The microparticles
were sterilized in methanol for a few minutes, soaked overnight in
Dulbecco’s Phosphate Buffered Saline (DPBS, Invitrogen), the
DPBS was removed and the particles dried in an oven at 37 1C
3
2
3
2
2
d
C
3
2
3
(
2
2
3
1
CP
2
CP
ꢁ
1
ArCH ortho to NO ), 139.7 (d, J 9.1, ArCCH P), 146.9 (d, J
CP
for 24 h. A stock protein solution of BSA in DPBS (5 mg ml )
2
CP
2
ꢁ
ꢁ
4
.0, ArCNO ); m/z (ESI ) 272 ([M–H] , 100%).
2
was used. Particles (2 mg) were placed in 24 well plates (3 replicates
for each sample group) and 1000 ml of BSA solution was added.
The plates were then incubated at 37 1C under constant shaking
for 3 h and the supernatant was removed, leaving the micro-
particles with adsorbed protein. Sodium dodecyl sulfate (SDS)
solution (1% (w/v), 1 ml) was added to each of the wells and the
plates were incubated at 37 1C for 1 h. After 1 h, the SDS solution
was removed and optical density (OD) read at a wavelength of
6
7
Diethyl (4-aminobenzyl)phosphonate. SnCl ꢃ2H O (1.80 g,
2
2
7
.96 mmol) and HCl (1.2 ml, 39.5 mmol) were added to a
stirring solution of diethyl (4-nitrobenzyl)phosphonate (1.07 g,
.87 mmol) in EtOH (20 ml). The mixture was refluxed for
.5 hours, cooled to room temperature then the pH adjusted to
–9 using aq. 5% NaOH solution. The mixture was extracted with
2
3
8
DCM (1 ꢀ 50 ml, 2 ꢀ 20 ml) and the combined organic layers
280 nm. The concentration of proteins adsorbed onto the particles
washed with water (20 ml) and brine (20 ml). The organic fraction
was calibrated using a standard curve. The DPBS and BSA without
microparticles served as the blank and reference respectively.
4
was dried over MgSO , and concentrated under vacuum to
produce a yellow semi-solid (774 mg). The product was purified
through a silica plug using EtOAc to elute the product as a yellow
Protein adsorption on microparticles (multiple time points).
Protein adsorption at different time points was determined using
bovine serum albumin (BSA) with the modified microparticle
surfaces. The microparticles were sterilized in methanol for
solid (576 mg, 82%). R 0.39 (EtOAc); mp 84–85 1C; vmax(DCM
f
ꢁ
1
film/cm ) 3430 (N–H), 2980, 1615, 1225 (PQO), 1050, 965, 850;
dH (400.2 MHz; CDCl ) 1.24 (6H, t, J
7.1, 6 ꢀ OCH CH ),
3
HH
2
3
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1187–1200 1199

View Article Online
1
0 min, soaked overnight in DPBS, and dried at 37 1C for 24 h
27 K. O. V. Flores, A. P. de Aguiar, M. R. M. P. de Aguiar and
L. C. de Santa Maria, Mater. Lett., 2007, 61, 1190–1196.
28 D. H. Williams and I. Fleming, Spectroscopic Methods in Organic
Chemistry, McGraw Hill, London, 4 edn. 1987.
prior to adsorption. A stock protein solution of BSA in DPBS
ꢁ
1
5 mg ml ) was used. Each type of modified microparticle was
(
divided into 5 groups (3 replicates) with different incubation times
of 1, 3, 6, 12 and 24 h. The microparticles were incubated in 96-well
plates, containing microparticles (2 mg) and BSA solution (200 mL)
in each well at 37 1C with constant shaking. At every time point,
the supernatant was carefully removed from all the samples and
transferred to a fresh 96-well plate. The BSA concentration in the
supernatant was analyzed spectrophotometrically at 280 nm. The
concentration of proteins adsorbed onto the particles can be
calculated by subtracting the concentration in the supernatant
2
3
3
9 J. Bernard, C. Branger, T. L. A. Nguyen, R. Denoyel and
A. Margaillan, React. Funct. Polym., 2008, 68, 1362–1370.
0 K. S. Birdi, Self-Assembly Monolayer Structures of Lipids and
Macromolecules at Interfaces, Springer, 1999.
1 chemicalize.org was used for generating structure property prediction
and calculations (http://www.chemicalize.org).
32 See, for example, http://www.emdbiosciences.com/html/NBC/lit
erature.html.
33 J. Rieger, J. Therm. Anal., 1996, 46, 965–972.
34 J. Anand, S. Palaniappan and D. N. Sathyanarayana, J. Polym.
Sci., Part A: Polym. Chem., 1998, 36, 2291–2299.
ꢁ1
from 5 mg ml , the actual concentration added before incubation.
35 D. Wang and Y. Liu, Front. Chem. China, 2008, 3, 440–444.
6 M. M. Hann, A. R. Leach and G. Harper, J. Chem. Inf. Model.,
001, 41, 856–864.
37 K. J. Miller and J. A. Savchik, J. Am. Chem. Soc., 1979, 101,
206–7213.
3
2
Acknowledgements
7
EMP gratefully acknowledges receipt of an industrially funded
CASE studentship (Materials KTN and OAS).
38 A. K. Ghose and G. M. Crippen, J. Chem. Inf. Model., 1987, 27,
21–35.
3
9 V. N. Viswanadhan, A. K. Ghose, G. R. Reyankar and
R. K. Robins, J. Chem. Inf. Model., 1989, 29, 163–172.
References
40 P. Ertl, B. Rohde and P. Selzer, J. Med. Chem., 2000, 43,
714–3717.
3
1
2
3
P. Vadgama, Annu. Rep. Prog. Chem., Sect. C, 2005, 101, 14–52.
C.-S. Ha and J. A. Gardella, Chem. Rev., 2005, 105, 4205–4232.
D. R. Absolom, W. Zingg and A. W. Neumann, J. Biomed. Mater.
Res., 1987, 21, 161–171.
4
1 P. Ertl, in Molecular Drug Properties, ed. R. Mannhold, Wiley-VCH,
Weinheim, 2007, Editon edn, pp. 111–126.
4
2 P. Ferrara, J. Apostolakis and A. Caflisch, Proteins: Struct.,
Funct., Genet., 2002, 46, 24–33.
3 A. Marmur, Soft Matter, 2006, 2, 12–17.
4
5
D. E. E. MacDonald, Biomaterials, 2002, 23, 1269–1279.
E. Moy, F. Y. H. Lin, J. W. Vogtle, Z. Policova and
A. W. Neumann, Colloid Polym. Sci., 1994, 272, 1245–1251.
A. Caro, V. Humblot, C. Methivier, M. Minier, M. Salmain and
C.-M. Pradier, J. Phys. Chem. B, 2009, 113, 2101–2109.
C.-K. Kang and Y.-S. Lee, J. Mater. Sci.: Mater. Med., 2007, 18,
4
4
4
4
4
4 I. Meiron and S. Sam, J. Colloid Interface Sci., 2004, 274, 637–644.
5 D. Y. Y. Kwok, Langmuir, 1997, 13, 2880–2894.
6 T. Yasuda, Langmuir, 1994, 10, 2435–2439.
6
7
8
9
7 R. J. Good and E. D. Kotsidas, J. Colloid Interface Sci., 1978, 66,
3
60–362.
8 Y. Li, J. Q. Pham, K. P. Johnston and P. F. Green, Langmuir,
007, 23, 9785–9793.
9 K. Holmberg, K. Bergstro
J. M. Harris, J. Adhes. Sci. Technol., 1993, 7, 503–517.
0 G. V. Lubarsky, R. H. Davidson and M. R. Bradley, Appl. Surf.
Sci., 2004, 227, 268–274.
1389–1398.
4
4
5
B. G. Keselowsky, D. M. Collard and A. J. Garcia, J. Biomed.
Mater. Res., 2003, 66, 247–259.
Y. Inoue, T. Nakanishi and K. Ishihara, React. Funct. Polym.,
2
¨
m, C. Brink, E. Osterberg, F. Tiberg and
¨
2011, 71, 350–355.
1
1
1
0 D. R. Schmidt, H. Waldeck and W. J. Kao, in Biological Interactions
on Materials Surfaces, ed. D. A. Puleo and R. Bizios, Springer,
Editon edn, 2009, pp. 1–18.
1 S. J. Dilly, M. P. Beecham, S. P. Brown, John. M. Griffin,
A. J. Clark, C. D. Griffin, J. Marshall, Richard. M. Napier,
P. C. Taylor and A. Marsh, Langmuir, 2006, 22, 8144–8150.
2 B. Sun, M. Yi, C. C. Yacoob, H. T. Nguyen and H. Shen, Acta
Biomater., 2012, 8, 1109–1116.
5
5
1 A. S. Hoffman, J. Biomed. Mater. Res., 1986, 20, ix–xi.
2 F. Boury, H. Marchais, J. P. Benoit and J. E. Proust, Biomaterials,
1
997, 18, 125–136.
3 G. Yin, Z. Liu, J. Zhan, F. Ding and N. Yuan, Chem. Eng. J.
Amsterdam, Neth.), 2002, 87, 181–186.
5
(
5
5
4 G. L. Trainor, Expert Opin. Drug Discovery, 2007, 2, 51–64.
5 G. L. Trainor, in Annu. Rep. Med. Chem., 2007, vol. 42,
pp. 489–502.
1
1
1
1
3 V. Beachley and X. Wen, Prog. Polym. Sci., 2010, 35, 868–892.
4 Z. Ma, Z. Mao and C. Gao, Colloids Surf., B, 2007, 60, 137–157.
5 Y. Ikada, Biomaterials, 1994, 15, 725–736.
5
5
6 M. P. Gleeson, J. Med. Chem., 2007, 50, 101–112.
7 P. Ascenzi, A. Bocedi, S. Notari, G. Fanali, R. Fesce and
M. Fasano, Mini-Rev. Med. Chem., 2006, 6, 483–489.
8 J. Ghuman, P. A. Zunszain, I. Petitpas, A. A. Bhattacharya,
M. Otagiri and S. Curry, J. Mol. Biol., 2005, 353, 38–52.
9 P. J. Hajduk, R. Mendoza, A. M. Petros, J. R. Huth, M. Bures,
S. W. Fesik and Y. C. Martin, J. Comput.-Aided Mol. Des., 2003,
6 J. M. Goddard and J. H. Hotchkiss, Prog. Polym. Sci., 2007, 32,
698–725.
5
5
1
1
7 G. Fourche, Polym. Eng. Sci., 1995, 35, 968–975.
8 P. J. Davis, L. Harris, A. Karim, A. L. Thompson, M. Gilpin,
M. G. Moloney, M. J. Pound and C. Thompson, Tetrahedron
Lett., 2011, 52, 1553–1556.
1
7, 93–102.
1
2
2
2
2
9 K. Awenat, P. J. Davis, M. G. Moloney and W. Ebenezer, Chem.
Commun., 2005, 990–992.
0 H. Wang, J.-P. Griffiths, R. G. Egdell, M. G. Moloney and
J. Foord, Langmuir, 2008, 24, 862–868.
1 C. Choong, J.-P. Griffiths, M. G. Moloney, J. Triffitt and
D. Swallow, React. Funct. Polym., 2009, 69, 77–85.
2 J. P. Griffiths, B. Maliha, M. G. Moloney and A. L. Thompson,
Langmuir, 2010, 26, 14142–14153.
3 P. Luksirikul, B. Ballesteros, G. Tobias, M. G. Moloney and
M. L. H. Green, Carbon, 2010, 48, 1912–1917.
6
0 C. M. Stiff, M. Zhong, R. W. Sarver, H. Gao, A. M. Ho,
M. T. Sweeney, G. E. Zurenkod and D. L. Romero, Bioorg.
Med. Chem. Lett., 2007, 17, 5479–5482.
61 C. Ybert and J.-M. d. Meglio, Langmuir, 1998, 14, 471–475.
62 N.-P. Huang, R. Michel, J. Voros, M. Textor, R. Hofer, A. Rossi,
D. L. Elbert, J. A. Hubbell and N. D. Spencer, Langmuir, 2001, 17,
489–498.
63 H. Noh and E. A. Vogler, Biomaterials, 2006, 27, 5780–5793.
64 H. Noh and E. A. Vogler, Biomaterials, 2006, 27, 5801–5812.
65 D. Rana and T. Matsuura, Chem. Rev., 2010, 110, 2448–2471.
66 T. Suzuki, O. Nagae, Y. Kato, H. Nakagawa, K. Fukuhara and
N. Miyata, J. Am. Chem. Soc., 2005, 127, 11720–11726.
67 S. Gester, J. Pietzsch and F. Wuest, J. Labelled Compd. Radiopharm.,
2007, 50, 105–113.
2
2
4 M. G. Moloney, J. Phys. D: Appl. Phys., 2008, 41, 174006.
5 D. Leonard, M. G. Moloney and C. Thompson, Tetrahedron Lett.,
2009, 50, 3499–3502.
26 T. R. Stratton, B. M. Applegate and J. P. Youngblood,
Biomacromolecules, 2011, 12, 50–56.
1
200 New J. Chem., 2012, 36, 1187–1200
This journal is c The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012