Article

pubs.acs.org/Langmuir

Covalent Functionalization of Silica Surface Using “Inert”

Poly(dimethylsiloxanes)
Gabriel Graffius,† Frank Bernardoni,‡ and Alexander Y. Fadeev*,†
†
Department of Chemistry and Biochemistry, Seton Hall University, South Orange, New Jersey 07079, United States
‡
Analytical Development and Commercialization, Merck Inc., Rahway, New Jersey 07065, United States
*
S Supporting Information

ABSTRACT: Methyl-terminated poly(dimethylsiloxanes)

(PDMSs) are typically considered to be inert and not suitable
for surface functionalization reactions because of the absence
of readily hydrolyzable groups. Nevertheless, these siloxanes
do react with silica and other oxides, producing chemically
grafted organic surfaces. Known since the 1970s and then
forgotten and recently rediscovered, this reaction provides a
versatile yet simple method for the covalent functionalization
of inorganic surfaces. In this work, we have explored the
reactions of linear methyl-terminated and cyclic PDMS and
bis-fluoroalkyl disiloxanes for the surface functionalization of
mesoporous silica (Dpore ≈ 30−35 nm). The optimal reaction
conditions included 24 h of contact of neat siloxane liquids and
silica at 120−250 °C (depending on the siloxane). A study of the reactions of silicas with different extents of hydration
demonstrated the critical role of water in facilitating the grafting of the siloxanes. The proposed reaction mechanism involved the
hydrolysis of the adsorbed siloxanes by the Lewis acidic centers (presumably formed by water adsorbed onto surface defects)
followed by the coupling of silanols to the surface to produce grafted siloxanes. For rigorously dehydrated silicas (calcination
∼1000 °C), an alternative pathway that did not require water and involved the reaction of the siloxanes with the strained siloxane
rings was also plausible. According to FTIR and chemical analysis, the reactions of bis-fluoroalkyl disiloxanes and cyclic PDMS
(D3−D5) produced covalently-attached monolayer surfaces, and the reactions of high-MM methyl-terminated PDMS produced
polymeric grafted silicas with a PDMS mass content of up to 50%. As evidenced by the high contact angles of ∼130°/100° (adv/
rec) and the negligible amount of water adsorption over the entire range of relative pressures, including saturation (p/p0 → 1),
the siloxane-grafted porous silicas show uniform, high-quality hydrophobic surfaces. An overall comparison of siloxanes with
classical silane coupling agents (i.e., silanes with readily hydrolyzable functionalities such as chloro, amino, etc.) demonstrated
that the reactions of siloxanes produced surfaces of similar quality and, although requiring higher temperatures, used
noncorrosive, less hazardous reagents, thereby providing an environmentally benign alternative to the chemical functionalization
of metal oxide surfaces.

■ INTRODUCTION
Organosilanes with the general formula R4−nSiXn (n = 1, 2, 3),
organosilanes “work” with a wide range of substrates of
different geometries (single crystals, fibers, fine particles, porous
where X = Cl, OC(O)R′, OR′, NR′2, or H, are the most widely solids) and chemistries (silicas, metals and metal oxides,
used reagents for the chemical functionalization of mineral polymers, etc.). The principles of silane coupling have been
surfaces that find numerous applications as materials for extensively covered in the literature.1−4,8,9
adsorption, separations, catalysis, wetting and adhesion control, Organosiloxanes (R3Si−O−SiR3) are typically not regarded
hydrophobic and superhydrophobic coatings, polymer fillers as silane coupling agents10 because of the absence of readily
and composites, the immobilization of biological molecules, hydrolyzable groups. The siloxanes, however, do react with
sensors, modified electrodes, and others.1−7 Organosilanes with surfaces in a similar fashion as do the “reactive” organosilanes,
a large variety of R groups are available commercially, enabling which was recognized in pioneering works on covalent surface
the preparation of surfaces of nearly all major organic functionalization in the 1970s and 1980s. The reactions of
functionalities either through direct deposition or subsequent hexamethyldisiloxane and linear and cyclic poly-
modification reactions. Robust Si−O linkages between the (dimethylsiloxane) (PDMS) producing silica-supported grafted
silane and the surface (covalently attached monolayers, CAMs)
and between neighboring silane molecules (self-assembled Received: August 8, 2014
monolayers, SAMs) ensure good thermal and chemical stability Revised: November 21, 2014
of the silane-modified surfaces. Many of the surface reactions of Published: November 24, 2014

© 2014 American Chemical Society 14797 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807
Langmuir Article

siloxane surfaces have been reported.11−16 Although the direct Reaction of Siloxanes with Silicas. Silica (∼0.5 g) was placed in
cleavage of Si−O−Si bond by surface silanols occurred only at a glass vial, and neat silicone oil was added to the vessel dropwise in an
high at temperatures14,15 of ∼300−350 °C, the reactions were amount sufficient to soak through and fully cover the powder (∼1
mL). Hexamethylcyclotrisiloxane (D3), a solid at room temperature,
facilitated greatly by adsorbed water.11−13 The nonreversible
was melted in the vessel by heating to 80 °C prior to the addition of
adsorption (chemisorption) of hexamethyldisiloxane on silica in silica. The vials were capped and placed in an oven at the desired
the presence of physisorbed water was observed at room temperature for the desired amount of time (generally 24 h). After the
temperature.11 The mechanism of siloxane bond cleavage on completion of the reaction, the reactors were allowed to cool and the
silica involving the six-membered ring formed by surface silanol, silicas were transferred to a glass filter and rinsed with toluene (three
siloxane, and molecular water has been proposed.13 In more to five times with 30 mL), acetone (three to five times with 30 mL),
recent works,17,18 the use of the siloxanes for the covalent water−acetone, and acetone again (two to three times with 30 mL).
functionalization of surfaces has been rediscovered. With solid- The silicas were air dried on the filter and then dried in an oven at 60
state NMR and DSC it was demonstrated17 that the reaction of °C for 1 h.
To study the effect of surface hydration (silanol content) on the
linear PDMS with silica produced surfaces of chemically grafted reaction, a series of progressively dehydrated silicas were prepared by
PDMS loops of four to eight repeat units. The use of methyl- calcination at 400−1000 °C overnight. After calcination, the silicas
terminated PDMS and PDMS copolymers as a general method were cooled to room temperature in nitrogen and used immediately in
for the covalent functionalization of inorganic surfaces (Si, Ti, the reactions with siloxanes. To study the effect of atmospheric
Al, and Ni) has been reported.18 moisture, a series of reactions were carried out in pressure Ace glass
The adsorption of PDMS on silica and the local structure of sealed reactors and in sealed glass ampules. The rest of the procedure
the PDMS−silica interface have received significant attention was the same as described above. For the majority of the reaction
because of the importance of this interface in applications of the conditions, the reactions were repeated at least twice. The results of
chemical analysis for independent runs were within 10% (or better)
silica-reinforced composites and silicone coatings. It should be agreement.
noted, however, that despite the significant research effort there Chemical Analysis. The extent of surface reactions was quantified
is no general agreement in the literature on the adsorption through chemical analysis (%C, %H, and %F when present), which
mechanism and the type of molecular interactions at the was performed by Robertson Microlit Laboratories (Ledgewood, NJ).
PDMS−silica interface. On the basis of ellipsometry, contact To allow direct comparison between the reactions using linear and
angles, adsorption kinetics, and rheological studies, the cyclic PDMS of different molecular masses, the grafting density was
authors19−21 concluded that the PDMS adsorption on silica calculated as the number of dimethylsiloxane (DMS) repeat units per
was governed by the H-bonding between siloxane oxygen and nm2 using equation
surface silanols, yet no direct evidence of such H-bonding was
6 × 105 × (%C) 1
ever demonstrated. Some chemical grafting during the ρDMS =
adsorption of methyl-terminated PDMS on glass was [nC × 1200 − MM × (%C)] S(BET)
mentioned;22 however, the major adsorption interaction of where MM and nC represent the molecular mass and number of
PDMS with the surface was attributed to H-bonding, and the carbons in the DMS group (74 g/mol and 2), %C is the carbon weight
chemical grafting was attributed to the presence of OH- percentage in the modified silica, and S(BET) is the BET surface area
terminated PDMS as a contaminant. Alternatively, compre- of bare silica (m2/g). For silicas reacted with fluoroalkyl disioxanes, the
hensive studies23−25 using inelastic neutron scattering, IR, grafting density was calculated as the number of monomer fluoroalkyl
quantum chemistry, and MD simulations demonstrated that siloxy groups per nm2. The grafting density was determined separately
interactions of PDMS with silica were similar to those in bulk from %C and from %F
PDMS (i.e., weak, nonspecific van der Waals type). In terms of
6 × 105 × (%C) 1
interaction energies, classical H-bonding made only a minor ρC =
contribution, and van der Waals forces mainly governed the [nC × 1200 − MM × (%C)] S(BET)
system.24,25
In this article, we explored the reactions of methyl- 6 × 105 × (%F) 1
ρF =
terminated linear and cyclic PDMS and bis-fluoroalkyl [nF × 1900 − MM × (%F)] S(BET)
disiloxanes for the covalent surface functionalization of
mesoporous silicas. The main focus of our work was placed where MM is the molecular mass of grafted group −OSi-
on (1) the optimization of the reaction conditions for the (CH3)2(CH2)2CnF2n+1 and nC and nF represent the number of carbon
and fluorine atoms in it. The ρC and ρF data are in close agreement.
covalent functionalization of porous silicas using siloxanes with
Typical standard deviations in carbon analysis and surface area
different structures, (2) the clarification of the role of silica measurements were ∼0.05% C and ∼5−10 m2/g (for Davisil 250),
surface hydration in its reactivity toward siloxanes, and (3) the respectively, which translated to ∼0.05 group/nm2 standard deviation
characterization of the grafting density, surface hydrophobicity, for the grafting density of the siloxane-modified silicas.
and thermal stability of the siloxane-grafted porous silicas. Thermogravimetric analysis (TGA) was performed using a TA

■
Instruments thermogravimetric analyzer operated between room
temperature and 1000 °C at a heating rate 10 °C/min in a flow of
EXPERIMENTAL SECTION dry nitrogen (20 cm3/min).
Chemicals. All solvents were purchased from Thermo Fisher FTIR measurements were performed using a PerkinElmer Spectrum
Scientific (Waltham, MA). The siloxanes were obtained from Gelest One instrument with a Harrick Seagull accessory. Samples of bare and
(Morrisville, PA) and used as received (unless stated otherwise). siloxane-grafted silica were packed tightly in the sample compartment,
Silicas. Wide-pore silica gel Davisil 250 was obtained from Grace and the spectra were collected in reflectance mode after 128 scans at a
(Columbia, MD). The pore structure of Davisil 250 was assessed by resolution of 4 cm−1. The subtraction of spectra was done using the
nitrogen adsorption (77 K) using an ASAP 2020 analyzer (Micro- PerkinElmer software.
meritics, Norcross, GA): specific surface area S(BET) = 252 m2/g, Water contact angles for the modified silicas were measured using a
total pore volume Vpore = 1.82 cm3/g, and average pore diameter Dpore Rame-Hart contact angle goniometer with pressed silica pellets. Eight-
≈ 30−35 nm. millimeter-diameter pellets of modified silicas were prepared using a

14798 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807

Langmuir Article

Figure 1. Schematics of the reaction of siloxanes with the silica surface and structures of the siloxane reagents studied.

hand press. The probe fluid (DI water) was added/withdrawn using a Table 1. Structure and Properties of Siloxanes Studied in
Gilmont microsyringe. This Work
Adsorption of Water Vapor. The adsorption−desorption
isotherms of water (293 K) were obtained using a Quantachrome molecular mass, no. of repeat
Autosorb One automatic volumetric adsorption instrument (Boynton chemical structure Da unitsb des.
Beach, FL). Prior to the analysis, samples were outgassed at 100 °C Cyclic PDMS
overnight. [OSi(CH3)2]3 222 3 D3

■
[OSi(CH3)2]4 296 4 D4
RESULTS AND DISCUSSION [OSi(CH3)2]5 371 5 D5
Linear PDMS (Methyl-Terminated)a
The main focus of this work was to investigate the reactions of −[OSi(CH3)2]n− 400 5 T02
poly(dimethylsiloxanes) of different structures for the covalent −≫− 750 10 T05
surface functionalization of mesoporous silica gel Davisil 250 −≫− 1250 16 T11
(Dpore ≈ 30−35 nm). By using the mesoporous silica, we −≫− 2000 27 T12
intended to explore the potential of the siloxane reagents for −≫− 3750 50 T15
the surface functionalization of porous materials for applica- −≫− 13 500 182 T23
tions in adsorption and separations, water purification, catalysis, −≫− 52 500 710 T35
and energy storage, among others. Three groups of siloxanes −≫− 115 000 1550 T46
were evaluated: (1) a series of methyl-terminated linear PDMSs −≫− 250 000 3400 T53
with molecular mass (MM) ranging from 0.4 to 225 kDa, (2) Bis(perfluoroalkyl)tetramethyl-disiloxane
cyclic PDMS (D3, D4, and D5), and (3) bis-fluoroalkyl O[Si(CH3)2(CH2)2CF3]2 326 2 bis-CF3
disloxanes (bis-CnF2n+1, n = 1, 6, 8). The reaction schematics, O[Si(CH3)2(CH2)2C6F13]2 826 2 bis-
C6F13
the structure of the siloxanes, and their notation are
O[Si(CH3)2(CH2)2C8F17]2 966 2 bis-
summarized in Figure 1 and Table 1. C8F17
Reaction of Siloxanes with Silica by IR. The presence of a
Linear PDMS were mixtures of oligomers. bThe molecular mass and
the methylsiloxane species on the siloxane-grafted silicas was number of repeat units were provided by the manufacturer and
confirmed by FTIR spectroscopy. As a representative example, represented average values.
Figure 2 shows an overlay of a spectrum of neat PDMS T23
and a spectrum of silica reacted with PDMS T23 (after the The changes in the spectra in the 3000−3800 cm−1 region
subtraction of the spectrum of bare silica). The absorption (OH stretching) provided insight into the surface chemical
bands at 1261 and 786 cm−1 were attributed to bending and bonding mode. For silicas pretreated at room temperature,
rocking modes of the Si−CH3 groups. The bands at 2961 and however, the picture was not very informative because the
1420 cm−1 (weak) were stretching and deformation modes of 3000−3800 cm−1 region was dominated by the strong
the CH3 groups in PDMS. Similar spectral changes that were absorption from physisorbed water. After the reactions with
consistent with the structure of the siloxanes used for the siloxanes, the subtraction spectra (Figure 2, top) showed only
reactions were also observed for silicas reacted with cyclic broad negative peaks at ∼3400 and 1620 cm−1, which was
PDMS and bis-Rf (Figures 1S and 2S). attributed to the removal of physisorbed water, consistent with
14799 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807
Langmuir Article

PDMS (D3, D4, and D5) and linear PDMS with MM = 0.4 and
0.75 kDa, a single band at ∼1003 cm−1 was observed. These
bands were assigned to the stretching of Si−O (νAS) in the
grafted PDMS species. From the literature, it is known that the
Si−O mode is present as a single band in the spectra for cyclic
dimethylsiloxanes with small rings (n < 5), whereas for larger
siloxane rings (n ≥ 6) and/or linear PDMS this band splits in
two.26 The analysis of this spectral region for the PDMS-grafted
silicas led us to conclude that the major products of the
reaction of the high-MM PDMS were linear PDMS chains and/
or larger PDMS loops (n ≥ 6), whereas for the reactions of
cyclic and low-MM PDMS the major products were PDMS
cycles with smaller rings (n < 5). The characteristic bands for
Si(CH3)2OH groups (970 and 840 cm−1, Si−OH stretching)
were not noted in the spectra for silicas reacted with either
linear or cyclic PDMS, which argued that the presence of the
OH-terminated PDMS species was insignificant, thereby
further supporting the conclusion that the major products for
the reaction were PDMS loops grafted to the surfaces by two
ends. In our previous work,27 we studied the reactions of α,ω-
dichloro-terminated PDMS with hydrated silicas that produced
PDMS loops covalently attached to silica by two SiS−O−Si
bonds. A comparison of the results27 with the results of the
present work showed that the IR spectra for silicas reacted with
methyl-terminated PDMS and silicas reacted with α,ω-
dichloro-terminated PDMS27 were nearly identical, demon-
strating the similarity in composition of these grafted layers.
Role of Reaction Temperature and Structure of
Siloxane. The effect of the reaction temperature was studied
for all three types of siloxanes by assessing the number of
grafted siloxanes, which was quantified through chemical
analysis (%C, H, F). The results are summarized in Figure 3.

Figure 2. FTIR spectra of Davisil 250 reacted with PDMS T23. Top:
Spectrum of neat PDMS T23 (1) and difference spectrum of Davisil
250 (prepared at room temperature) reacted with T23 (2). Bottom:
The OH-stretching region for bare Davisil 250 prepared at 400 °C (3),
spectra of this silica reacted with T23 (4), and difference spectrum (5).

the hydrophobic nature of the siloxane-grafted silicas. For silicas

pretreated at elevated temperatures, physisorbed water was
largely removed and the spectra showed bands from the Figure 3. Effect of reaction temperature on the amount of grafted
isolated (3747 cm−1) and hydrogen-bonded (∼3550 cm−1) siloxane (%C, wt). For the designation of siloxanes, see Table 1. The
silanols. After the reaction with siloxanes, the intensity of these points are connected to guide the eyes; some points are omitted for
bands was significantly reduced (Figure 2, bottom), indicating clarity.
the involvement of the surface silanol groups in the reactions
with siloxanes and suggesting the formation of surface covalent First, we noted that for all of the siloxanes tested, the reactions
bonds SiS−O−Si. required elevated temperature. Very low organic loadings were
The analysis of the Si−O stretching region (1000−1100 observed at room temperature, arguing for thermally activated
cm−1) of the spectra provided additional insight into the chemical bonding of the siloxanes on the silica surface.
structure of the grafted PDMS chains (Figure 1S). For the Depending on the structure of the siloxane, however, the
silicas reacted with the linear PDMS of high MM (≥1.25 kDa), optimal reaction temperatures that ensured maximal grafting
two prominent bands at ∼1010 and 1080 cm−1 were present in were different. Linear PDMSs, with an average MM greater
the spectra, whereas for the silicas reacted with the cyclic than 1.25 kDa, were clearly the most active siloxanes studied in
14800 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807
Langmuir Article

Table 2. Comparison of [CnF2n+1(CH2)2Si(CH3)2]2O, CnF2n+1(CH2)2Si(CH3)2X, and CnF2n+1(CH2)2SiX3 as Reagents for the

Surface Functionalization of Silicas
ρmax,
group/
reagent, n = 1−6 nm2 comments and refs
[CnF2n+1(CH2)2Si(CH3)2]2O 1.4− noncorrosive reagents; effective in vapor phase and as neat liquids; T ≈ 200 °C; works well for single and porous surfaces
2.3a (this work)
CnF2n+1(CH2)2Si(CH3)2X 1.6− best results obtained for X = Cl and N(CH3)2 that are corrosive; effective in solution and vapor phase and as neat liquids;
X = Cl, N(CH3)2, CH3O, 2.4a T ≈ 60−100 °C; works well for single and porous surfaces28
C2H5O
CnF2n+1(CH2)2SiX3, 2.7− best results obtained for X = Cl, which is highly corrosive; poor reproducibility due to extreme sensitivity to reaction
X = Cl, CH3O, C2H5O 3.5a conditions and moisture content; T ≈ 25 °C; solvent is needed and ρmax can be achieved only for flat surfaces; does not
work well for porous silicas28
a
For all of the reagents, the grafting density decreased as the size of the fluoroalkyl component increased. The small number corresponds to the
reagent with n = 6, and the large number corresponds to the reagent with n = 1.

this work: the maximal grafting was observed at 120 °C. The For silicas reacted with cyclic PDMS, the maximal grafting
reactions of cyclic PDMS and bis-Rf disiloxanes required higher densities were in the range of five to six dimethylsiloxane
temperatures: ∼200 °C (bis-Rf) and ∼250 °C (cyclic PDMS). (DMS) repeat units per nm2, which was consistent with the
Further increases in temperature resulted in lower grafting or monomolecular grafting and thereby ruled out possible ring-
gave rise to side reactions, e.g., bulk polymerization of cyclic opening polymerization. A minor decrease in the grafting
PDMS at temperatures exceeding 275 °C. The relatively high density was observed in the range of D3 > D5 ≈ D4. The
reactivity of the high-MM PDMS was attributed to the somewhat higher reactivity of D3 was attributed to the strained
molecular mass effect: the higher-MM polymer adsorbed ring effect.26
more strongly and therefore was conceivably more available As opposed to the reactions of bis-Rf disiloxanes and cyclic
for the surface reactions. We cannot, however, completely rule PDMS, the reactions of high-MM PDMS produced polymeric
out the probability of the presence of OH-terminated PDMS in surfaces. The data in Figure 4 show that the grafting density
the samples as contaminants. The OH-terminated PDMSs are increased as the MM of PDMS increased and then leveled off.
known to be much more active19,22 than CH3-terminated The saturation was observed for PDMS of 115 kDa and was
PDMSs in their reactions with silica, and their presence, even in higher with the grafting density plateau at ∼45 DMS group/
small quantities, would affect the surface grafting. For nm2. Using the grafting density data and the molecular
comparison, Figure 3 shows data for the OH-terminated dimensions of the DMS group, the average thickness of the
PDMS. The reaction of the OH-terminated PDMS with silica PDMS grafted layers was estimated as follows. The area
occurred at room temperature and resulted in notably larger occupied by the DMS group grafted onto silica has been
grafting compared to that for methyl-terminated PDMS. The reported27 to be σ = 0.40 nm2. This value was close to the
higher reactivity of bis-Rf disiloxanes versus that of cyclic molecular cross-section of the PDMS chain in PDMS crystals,30
PDMS was attributed to the electron-withdrawing effect of the σ = 0.36 nm2. When 0.40 nm2 was used for the molecular cross-
section, the maximal grafting density for a monolayer of DMS
fluoroalkyl group that made the siloxane bond more reactive.26
groups on silica was obtained as ρmax = σ−1 = 2.50 group/nm2,
Grafting Density for the Siloxane-Modified Silicas:
which served as an upper limit for a closely packed single layer
Comparison with Other Silane Coupling Agents. The
of PDMS lying flat on the surface. The diameter of the DMS
effectiveness of the surface functionalization reactions is
group was determined to be (4σ/π)1/2 = 0.71 nm (spherical
primarily determined by the extent of surface coverage or the
shape), which was used as an estimate of the thickness of a
maximal achievable grafting density ρmax, group/nm2. Also, such single layer of PDMS. Assuming uniformly packed layers, the
parameters as the ease of the procedure, the reproducibility of average thickness of the PDMS surface with the grafting density
the reaction, and the hazards of the reagents play an important ρ was calculated to be h = (ρ/ρmax) × 0.71 nm. The plot of the
role when the effectiveness of different reactions is evaluated. calculated thickness for silicas reacted with PDMS as a function
To compare the effectiveness of the siloxanes, we examined of their MM is shown at the bottom of Figure 4. From that
their maximal grafting densities along with the other reaction graph, we determined that the amount of grafted PDMS that
parameters with respect to those for chloro- and alkoxysilanes, was equivalent to a single layer of PDMS was observed for the
the most studied class of organosilicon reagents used for the reaction of 0.75 kDa PDMS. The saturation grafting density of
surface functionalization of silicas and other substrates. The 45 DMS group/nm2 corresponded to ∼18 layers of PDMS with
most straightforward comparison was possible between bis-Rf a total thickness of ∼12.8 nm. Figure 4 also shows the
disiloxanes (this work) and Rf monofunctional and trifunctional ellipsometric thickness data reported for the PDMS layers
silanes28 because they reacted with silica surfaces in a similar grafted onto Si wafers.17 The reaction conditions (24 h at 100
fashion to produce chemically grafted monolayers of the Rf °C) used in ref 17 were similar to those used in this work, thus
groups (Table 2). allowing for a direct comparison of the results. As seen in
The analysis in Table 2 shows that the ρmax values obtained Figure 4, two sets of data matched, thereby demonstrating that
for the disiloxanes were similar to those reported for the best- the reactions of PDMS with porous silica worked as well as with
quality monolayers derived from chloro- and aminosilanes.28 smooth silica (Si wafer), producing surfaces of similar thickness
We noted that the reaction of siloxanes, although it required a (amount of grafted PDMS). The difference between a flat
higher temperature, used noncorrosive reagents and produced surface substrate (Si wafer) and porous silica becomes notable
water as the only byproduct, thereby providing a cleaner and for PDMS with MM ≥ 115 kDa when the thickness of porous
less hazardous environment than amino- and chlorosilanes. silica reached the plateau, which was attributed to space
14801 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807
Langmuir Article

not differentiate between the various kinds of silanol groups

(isolated, vicinal, geminal) nor did it take into account their
accessibility (internal vs external silanols). Nevertheless, this
model accurately described the overall dehydration process and
provided a simple method to determine the amount of
physisorbed water and the number of silanol groups by
TGA. 30,31 To determine the surface concentration of
physisorbed water and silanol groups, the BET surface areas
were assayed for each silica dehydrated at different temperature.
The results are shown in Table 3. Davisil 250 dehydrated at 100

Table 3. Amount of Physisorbed Water and Total Silanols in

Davisil 250 Prepared at Different Temperatures
calcination T, S(N2, BET), H2O, SiOH, group/
°C m2/g molecule/nm2 nm2
100 252 2.3 5.5
500 273 0 2.5
600 253 0 1.2
800 232 0 0.5
1000 118 0 0.2

°C had 2.3 molecules of physisorbed water per nm2 and 5.5

silanols per nm2, indicating a fully hydroxylated surface.30
Calcination at T ≥ 500 °C produced silicas that were free from
molecularly adsorbed water with a gradually decreasing
population of silanols as the calcination temperature increased,
in a good agreement with the literature.30
Figure 5 presents grafting densities plotted as functions of
silica calcination temperature. The grafting density data are

Figure 4. Top: Effect of the average molecular mass of linear PDMS

on the grafting density (DMS group/nm2). Reaction conditions: 24 h
at 100 °C. Bottom: Squares, thickness of PDMS layers calculated from
the grafting density and diameter of the PDMS chain (this work).
Triangles, ellipsometric thickness of PDMS grafted onto Si wafers.18
The dashed line corresponds to a single layer of PDMS lying flat on
the surface.

limitations for the reactions inside the pores as compared to the Figure 5. Effects of silica calcination temperature on the reactions with
reaction with smooth substrates. The average thickness at the different siloxanes. Grafting density data (y axes) is shown as a % from
plateau (12.8 nm) was ∼80% of the average pore radius of the maximal value observed for a given siloxane. Closed symbols
Davisil 250, ∼15−17.5 nm. represent data for the reactions carried out at atmospheric pressure
Role of Surface Water. The importance of the physisorbed using siloxanes as received (containing 60−100 ppm water). Open
water for the reactions of siloxanes with silicas has been symbols stand for the reactions carried out under rigorously dry
highlighted in early work.11−13 To investigate further, we have conditions. The temperature and time of all of the reactions were 100
studied the reactions of siloxanes with a series of progressively °C and 24 h. The points are connected to guide the eyes.
dehydrated silicas. It is well documented30 that the amount of
water on the silica surface can be thoroughly controlled through shown as a percentage of the maximal value (observed for silica
the thermal treatment of silica at different temperatures. pretreated at 100 °C), thereby allowing for direct comparison
According to the Zhuravlev model,30 the initial dehydration of the reactions of the siloxanes of different structures. As
of silica involved the removal of physically adsorbed water, anticipated, for all of the siloxanes the grafting density
most of which was completed by ∼200 °C. The subsequent loss decreased as the temperature of silica calcination increased.
of water in the 200−1000 °C range was attributed to the The decrease in grafting density, however, was not consistent
process of dehydroxylation of surface silanols with the with the decreased number of surface silanols. Indeed, for silicas
formation of siloxane bridges. We note that this was a rather pretreated at 800 °C, the number of silanols was less than 10%
simplified description of the dehydration process because it did of that for fully hydroxylated silica, yet only a minor decrease in
14802 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807
Langmuir Article

grafting density was observed (Figure 5). The removal of

physisorbed water (treatment at 200 °C) did not seem to have
any significant impact on the reactions. Only for silicas
calcinated at 1000 °C did the number of grafted siloxanes
show a notable decrease reflecting the nearly complete removal
of silanols. However, even for these silicas, the grafting densities
were unusually high. Initially puzzled by these results, we later
realized that silica was not the only source of water for the
reaction. Normally, the reactions of silica with siloxanes were
carried out in capped vials at atmospheric pressure, so there was
a possibility of atmospheric moisture entering the reaction According to the reaction scheme shown above, one
media. We also found that a notable amount of water could be molecule of surface water is consumed and two water molecules
introduced into the reaction system with the siloxanes. are produced per two grafting points. To examine this further,
According to chemical analysis (Oven Karl Fischer), the we ran the reactions in open vessels while monitoring the total
weight of the reaction system over time using TGA. By using
equilibrium amount of water present in all of the siloxanes “as silicas and nonvolatile PDMS T23 (both dried to a constant
received” was ∼60−100 ppm. The use of dried siloxanes weight at the temperature of the reaction, 100 °C, prior to
(heating in vacuum) did show some reduction of the grafting mixing), the weight loss observed was associated with the
density, but the effect was not significant. Therefore, our next removal of the water byproduct of the surface grafting reaction.
experiments were aimed at eliminating the possible impact of The corresponding weight loss curves shown in Figure 6
atmospheric moisture by carrying out the reactions under
rigorously dry conditions. These included thorough outgassing
of both silica and siloxane at elevated temperature and running
the reactions under vacuum (using nonvolatile siloxanes). The
results are also shown in Figure 5 with open symbols. The
difference in the reactivity for the systems in vacuum versus
being exposed to the atmosphere was remarkable: the grafting
densities for the reactions carried out in vacuum were reduced
greatly as compared to the reactions exposed to the
atmosphere, confirming that atmospheric moisture did play
an important role in the mechanism facilitating the reaction of
siloxanes with silica. Although the reduction of grafting under
dry conditions was observed for all types of siloxanes, the effect
was more prominent for low-MM siloxanes (bis-CF3 and D3)
than for high-MM PDMS. We attributed this to differences in
the surface reactions for monomers (one grafting point per
molecule) and polymers, where multiple grafting points per
Figure 6. Isothermal weight loss kinetic plots observed for bare Davisil
molecule were possible. In the case of low-MM siloxanes (bis- 250 and Davisil 250 mixed with PDMS T23 (closed symbols).
CF3 and D3), grafting was directly related to the concentration
of the surface reaction centers (physisorbed water and silanols) demonstrate that the weight loss was notable after the first hour
that were reduced significantly as the dehydration temperature of contact of silica with PDMS, indicating that the reaction was
increased. For high-MM PDMS, the decrease in the number of fast. Interestingly, the reaction was not complete after 24 h
reaction centers probably decreased the number of grafting because the weight loss curve did not level off (Figure 6). In
points per molecule, but it had less of an effect on the total this regard, we point out that, according to chemical analysis (%
number of grafted molecules, which remained approximately C, H), the reactions of T23 PDMS were completed by 12 h
because no further increase in carbon loading was observed
constant at ∼65−70% from the maximum for silicas dehydrated
after that time. When comparing these two sets of kinetic data,
at T > 400 °C. we speculate that the initial stage (minutes to hours) of the
The results presented above further supported the critical reaction included the hydrolysis of the siloxanes into silanols
role of physisorbed water for the reaction of siloxanes with and the immediate strong adsorption of these silanols on the
silicas, as first proposed in ref 13. We believe that the reactions silica surface through H-bonding. These processes resulted in
of siloxanes with silica are initiated by the Lewis acid centers the immobilization of the siloxanes on the surface but did not
formed by water molecules adsorbed on the silica surface release any water and therefore were not reflected in the weight
loss curves shown in Figure 6. In a later stage (many hours and
defects. The acidic centers are known to catalyze the hydrolysis
possibly days, depending on the temperature) of the reaction,
of the siloxane bond.26 The silanols produced after the the adsorbed silanols reacted with surface silanols to release
hydrolysis of the siloxane further react with surface silanols, water, which was observed in the weight loss curves. We
yielding grafted siloxanes: surmise that (especially for the high-MM PDMS) the
14803 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807
Langmuir Article

rearrangements of the adsorbed silanol-terminated siloxanes on Table 4. Hydrophobic Surface Properties of Silicas Grafted
the surface, possibly including their deeper hydrolysis and with Siloxanes
subsequent reactions with surface silanols and with each other
ρ, water contact angles for
(cross-linking) continue to take place well after the reactions group/ water vapor adsorption at p/p0 pressed pellets (adv/
are quenched and considered to be terminated. siloxane nm2a = 0.3, mg per g of silica rec), deg
The data presented in Figure 5 demonstrated that the none 36.19 spreads
reaction of siloxanes with silica did occur, albeit to a lower D3 9.31 4.26 126/95
extent, even with silica dehydrated at 1000 °C and under dry bis- 1.85 6.02 130/105
conditions (i.e., when supposedly no water was present in the C6F13
system). The residual reactivity observed even for rigorously T12 6.51 13.86 110/10
dried systems deserves special explanation. According to the T15 9.20 12.88 115/25
reaction scheme proposed above, the reaction of siloxanes with T23 22.15 12.26 115/70
silica produced water, so, with respect to water, the reaction T35 33.33 11.98 115/84
a
was autocatalytic. It was therefore plausible to assume that even Calculated as the number of dimethylsiloxy groups per nm2. For bis-
traces of water would be sufficient to initiate the reaction. C6F13 disiloxane, calculated as the number of C6F13 group per nm2.
Although we cannot completely rule out the presence of trace
amounts of water in the reactions, we would like to offer an
alternative explanation for the reactivity of dehydrated silicas
with siloxanes that did not require water. This involves the
reaction of the siloxanes with dehydrated silica surface via the
opening of strained siloxane rings:

The dehydration of silicas at high temperatures (∼400−1000

°C) produced surfaces containing strained siloxane rings.32,33
These strained rings readily participated in various reactions
and could be used as reactive centers for the covalent
functionalization of surfaces. The reaction of siloxanes with
strained siloxane rings was first described in ref 11. The Figure 7. Adsorption−desorption isotherms of water vapor (293 K)
reactions of strained siloxane rings with water, alcohols, amines, for bare Davisil 250 silica and silica reacted with T23 and D3 siloxanes.
silanes, and siloxanes have also been reported.32−36 For the Open symbols, adsorption; closed symbols, desorption.
concentration of the siloxane rings, the following numbers have
been reported (with the temperature of the silica treatment in silicas reacted with cyclic PDMS and bis-fluoroalkyl disiloxanes.
parentheses): 0.15 nm−2 (1200 °C),34 0.2−0.4 nm−2 (1080 The water contact angles for these surfaces were ∼130°/100°
°C),33 0.04−-0.13 nm−2 (900 °C),35 and 0.004−0.024 nm−2 (adv/rec), indicating low-energy surfaces with nonpolar
(300−700 °C).36 In this work, for the reaction of dried bis-CF3 functionalities. The high values of both advancing and receding
siloxane with silica dehydrated at 1000 °C, the grafting density angles were characteristic of densely packed uniform surfaces
of the CF3 groups was 0.26 nm−2. Assuming that the reaction that provided efficient blockage of polar centers of bare silica.
occurred exclusively through the strained rings, this corre- We consistently observed that the water contact angles for the
sponded to a concentration of 0.13 siloxane rings per nm2, pressed powders were a bit higher (∼5−10°) than those
which is well within the range reported in the literature. This reported for surfaces of covalently attached and self-assembled
result indicated that the reaction of siloxanes with rigorously monolayers of similar chemical composition supported on
dehydrated silicas could be explained solely by the strained ring smooth (Si wafer) substrates,28 which was attributed to the
centers, thereby demonstrating an alternative pathway for the surface roughness of the pressed pellets. The contact angles for
reaction of siloxanes with surfaces that did not require water. the silicas reacted with linear PDMS were strongly affected by
Surface Properties. To gain insight into the uniformity of the MM of the PDMS used (Table 4). For the surfaces derived
the coatings, the hydrophobic surface properties of the siloxane- from low-MM PDMS (≤3.75 kDa), the advancing contact
modified silicas were examined next. Two methods were used angles were high (>100°) yet the receding angles were
to assess the hydrophobicity of the surfaces prepared: (1) water particularly low (∼0−40°), indicating a substantial presence
contact angles using silica pressed pellets and (2) adsorption of uncovered silanols. For PDMS with MM greater than 13.5
isotherms of water vapor at room temperature (293 K). The kDa, the hydrophobicity of the surfaces greatly improved with
data are summarized in Table 4 and Figure 7. the advancing angles leveling off ∼115°. The receding angles,
Water spreads over the surface of bare silica, indicating highly however, were less than 90°, which was consistent with the
polar surfaces of silanol groups. The grafting of nonpolar hydrophobic surfaces of grafted PDMS chains and, possibly, a
organosiloxanes on silica produced hydrophobic surfaces that small number of polar groups available to water (Si−OH).
were evidenced by a remarkable increase in the water contact The adsorption isotherms of water vapor corroborated the
angles. The best-quality hydrophobic surfaces were obtained for contact angle data. Figure 7 presents the water adsorption
14804 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807
Langmuir Article

isotherms for bare silica and silicas grafted with various

siloxanes. As demonstrated by the remarkable reduction of the
water adsorption observed over the entire range of pressures,
including the saturation (p/p0 → 1), the reactions of siloxanes
produced high-quality hydrophobic surfaces. According to the
adsorption of nitrogen (data not shown), all of the modified
silicas demonstrated type IV isotherms, i.e., the silica pore space
was largely preserved albeit the pore volume was notably
reduced by the grafting. Even though the pore space was
available, no pore filling was observed by water. The adsorbed
amounts of water at saturation corresponded only to a small
fraction (∼2-5%) of the available pore volume. Assuming a
uniform distribution and 0.106 nm2 for the cross-section of the
adsorbed water molecule,37 these adsorbed amounts corre-
sponded to approximately one-third of the monolayer capacity.
Such a low value of water adsorption and the absence of pore
filling at saturation were indicative of uniform, hydrophobic
surfaces of the pores. The absence of pore filling by water was Figure 8. TGA plots for bare silica and silica grafted with D3 (offset by
observed, not only for silicas reacted with bis-Rf disiloxanes and 5 wt %), bis-CF3 (offset by 10 wt %), and T23 (offset by 15 wt %).
cyclic PDMS (where only monomolecular grafting was
possible) but also for silicas reacted with the high-MM
PDMS arguing for the uniform distribution of grafted polymers
over the pore surfaces. As shown in Figure 7, the desorption
was always incomplete, and small residual amounts of water
were retained by the silicas. This water, however, was adsorbed
reversibly because it was completely removed upon outgassing
at elevated temperature. On the basis of the amount of water
adsorbed at p/p0 = 0.3, the hydrophobicity of the siloxane-
grafted silicas was as follows:

cyclic PDMS ≈ bis‐R f siloxane

≥ high MM linear PDMS
> low MM linear PDMS
We noted that this range of hydrophobicity was consistent with
that obtained by contact angles (Table 4). The agreement
between the two techniques in the characterization of the
surface hydrophobicity of porous silicas was remarkable,
Figure 9. Weight fraction of carbon in the TGA total weight loss
considering that the contact angles probe the external surfaces
(200−600 °C region) for silicas grafted with a linear PDMS of
of the particles and the vapor adsorption isotherms probe the different molecular mass. The dashed line is at 0.32, the weight fraction
entire surface, ∼98% of which is the surface of the pores. of carbon in pure PDMS.
Thermal Gravimetric Analysis. One of the superior
qualities of the siloxanes as surface functionalization agents is
related to the excellent thermal stability of the siloxane-grafted rization via scission of the Si−O−Si backbone with the
silicas as assessed by TGA. The representative TGA plots for formation of volatile cyclic siloxanes (D3 and higher). For the
bare silica and silica reacted with different siloxanes are shown PDMS chains grafted to silica,27 the degradation mechanism
in Figure 8. For all siloxane-grafted silicas, the main weight loss was proposed to be similar; however, the process was affected
feature was observed above ∼300 °C, which was attributed to by the grafting points, where the depolymerization of grafted
the degradation of grafted organic groups. The temperature of PDMS was interrupted by the silica matrix. For monolayers of
the onset of the weight loss and the temperature of the maximal DMS groups grafted to silica, the degradation pathway was
rate of the weight loss (Tmax) showed some variation for different because it involved breaking the Si−CH3 bonds rather
different siloxanes. The total weight loss varied significantly, than the desorption of dimethylsiloxy species via breaking the
reflecting the difference in mass fraction of grafted siloxanes: Si−O−SiS bonds.27 The Tmax for PDMS-silicas was ∼450 °C,
from ∼10% for silicas reacted with cyclic PDMS and bis- whereas for DMS-silicas Tmax was higher,27 ∼550 °C. With this
fluoroalkyl disiloxanes up to ∼50% for silicas reacted with the data, the differences in TGA curves observed for silicas reacted
high-MM PDMS. For silicas reacted with cyclic PDMS, the with high- and low-MM PDMS can be explained by different
weight loss occurred over a wide range of temperature (∼340− numbers of grafts per molecule and therefore different
670 °C), with Tmax centered at ∼550−600 °C. For silicas contributions of different degradation mechanisms. We tried
grafted with high-MM PDMS, the weight loss occurred in a to differentiate these processes based on the analysis of the
similar range (∼300−600 °C) yet at a notably lower Tmax ∼ carbon percent (determined separately via chemical analysis) in
450 °C. From the literature on the thermal degradation of the total weight loss. For the thermal degradation occurring
PDMS,38 the mechanism of this process involved depolyme- mainly through depolymerization, the carbon fraction in the
14805 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807
Langmuir Article

weight loss would approach 0.32, the value for pure PDMS. As Notes
the number of grafts per chain increased, the contribution from The authors declare no competing financial interest.

■
the pyrolysis of Si−CH3 groups would also be increasing,
leading to an increase in the percentage of carbon in the total ACKNOWLEDGMENTS
weight loss. The experimental data (Figure 9) demonstrated a
notable decrease in the percentage of carbon in the total weight Support from the Seton Hall Research Council and from Merck
loss as the MM of grafted PDMS increased. This supported the Inc. is acknowledged.
conclusion of fewer attachment points (per PDMS chain) for
silicas reacted with high-MM PDMS. For silicas reacted with
bis-Rf disiloxanes, the total weight loss increased with the size of
■ REFERENCES
(1) Leyden, D. E., Ed.; Silanes, Surfaces, and Interfaces; Gordon and
the fluoroalkyl group, reflecting an increase in the grafted Breach: New York, 1986.
amount, and onset and Tmax were similar at ∼490−500 °C for (2) Plueddemann, E. W. Silane Coupling Agents, 2nd ed.; Plenum:
all of the bis-Rf disioxanes studied.

■
New York, 1991.
(3) Mittal, K. L. Silanes and Other Coupling Agents; VSP: Utrecht,
CONCLUSIONS 1992.
The reactions of bis-Rf disloxanes, cyclic PDMS (D3−D5), and (4) Owen, M. J. Coupling Agents: Chemical Bonding at Interfaces. In
linear methyl-terminated PDMS (MM 0.4−250 kDa) with Adhesion Science and Engineering; Dillard, D. A., Pocius, A. V, Eds.;
Elsevier: Amsterdam, 2002; Vol. 2, pp 403−431.
mesoporous silica (Dpore ≈ 30−35 nm) were investigated. (5) Park, S.-E.; Prasetyanto, E. A. Morphosynthesis and Catalysis by
According to FTIR and chemical analysis, the reactions of bis- Organofunctionalized Mesoporous Materials. In Organosilanes: Proper-
Rf and cyclic PDMS produced closely packed monomolecular ties, Performance, and Applications. Wyman, E. B., Skief, M. C., Eds.;
surfaces of chemically grafted fluoroalkyl groups and small Nova Science: New York, 2010; pp 101−131.
PDMS cycles (n < 5), respectively. Using linear methyl- (6) Krumpfer, J. W.; Gao, L.; Fadeev, A. Y.; McCarthy, T. J. Using
terminated PDMS of high-MM (≥1.25 kDa) produced Surface-Attached Organosilanes to Control and Understand Hydro-
chemically grafted polymeric surfaces. As assessed by water phobicity and Superhydrophobicity. In Advances in Silicon Science;
contact angles and adsorption isotherms (293 K) of water Dvornic, P. R., Owen, M. J., Eds.; Springer: New York, 2012; Vol. 4,
vapor, quality hydrophobic surfaces were obtained for silicas pp 95−114.
grafted with bis-Rf disiloxanes, D3−D5, and linear PDMS of (7) Shokoohi, S.; Arefazar, A.; Khosrokhavar, R. Silane Coupling
Agents in Polymer-based Reinforced Composites: A Review. J. Reinf.
MM ≥ 13.5 kDa. According to TGA, the siloxane-grafted silicas
Plast. Compos. 2008, 27, 473−485.
showed superior thermal stability: the temperatures of the (8) Fadeev, A. Y.; McCarthy, T. J. Self-assembly is Not the Only
maximal weight loss rate were ∼450 °C (PDMS), ∼500 °C Reaction Possible Between Alkyltrichlorosilanes and Surfaces:
(bis-Rf), and ∼575 °C (D3−D5). Monomolecular and Oligomeric Covalently Attached Layers of
A study of the reactions of silicas with different extents of Dichloro- and Trichloroalkylsilanes on Silicon. Langmuir 2000, 16,
hydration and controlled moisture content in the system 7268−7274.
demonstrated the critical role of water in facilitating the grafting (9) Haensch, C.; Hoeppener, S.; Schubert, U. S. Chemical
of the siloxanes. The proposed reaction mechanism involved Modification of Self-Assembled Silane Based Monolayers by Surface
the hydrolysis of the siloxanes by the Lewis acid centers formed Reactions. Chem. Soc. Rev. 2010, 39, 2323−2334.
by the adsorbed water (presumably on surface defects). (10) Silane Coupling Agents: Connecting across Boundaries. , Gelest:
According to this scheme, the hydrolysis of the siloxanes Morrisville, PA, 2006, www.gelest.com.
(11) Aristova, V. G.; Zimmer, I. M.; Gorbunov, A. I.; Galitskaya, O. I.
produced silanols that further reacted with surface silanols to Effect of the Surface State of Aerosil on the Nature of
yield the siloxane-grafted surfaces. For the rigorously Hexamethyldisiloxane Adsorption. Zh. Fiz. Khim. 1974, 48, 134−137.
dehydrated silicas (calcination ∼1000 °C), an alternative (12) Aristova, V. G.; Popkov, K. K.; Galashina, M. L. Interaction of
pathway that involved the reaction of the siloxanes with the M e t h y l c h l o r os i l a n e s a n d L o w - M o l e c u l a r - W e i gh t P o l y -
strained siloxane rings and did not require water was also (methylcyclosiloxanes) with an Aerosil Surface. Kolloidn. Zh. 1974,
plausible. 36, 123−126.
An overall comparison of siloxanes and classical silane (13) Aristova, V. G.; Zimmer, I. M.; Gorbunov, A. I.; Zhilenkov, I. V.;
coupling agents (i.e., silanes with readily hydrolyzable Saushkin, V. V. Mechanism of Siloxane Bond Splitting on a Hydrated
functionalities such aschloro, amino, etc.) demonstrated that Aerosil Surface. Doklady Akademii Nauk SSSR 1980, 255, 131−134.
the reactions of siloxanes with silicas produced surfaces of (14) Tertykh, V. A.; Pavlov, V. V. Reactivity of molecules attacking a
functional center. Adsorp. Adsorb. 1978, 6, 67−75.
similar quality and, although they required higher temperatures, (15) Pavlov, V. V.; Guba, G. Y.; Tertykh, V. A.; Chuiko, A. A. Study
used noncorrosive reagents and generated only water as a of the interaction of alkylsiloxanes with the surface of dispersed silicas.
byproduct, thereby providing a viable and environmentally Adsorp. Adsorb. 1980, 8, 35−39.
benign alternative to the chemical functionalization of (16) Li, Y.; Xia, Y.; Xu, D.; Li, G. Surface Reaction of Particulate
mesoporous metal oxides.

■
Silica with Polydimethylsiloxanes. J. Polym. Sci., Polym. Chem. Ed. 1981,
19, 3069−3079.
ASSOCIATED CONTENT (17) Litvinov, V. M.; Barthel, H.; Weis, J. Structure of a PDMS Layer
*
S Supporting Information Grafted onto a Silica Surface Studied by Means of DSC and Solid-State
FTIR spectra of silicas reacted with cyclic PDMS and bis-CF3 NMR. Macromolecules 2002, 35, 4356−4364.
(18) Krumpfer, J. W.; McCarthy, T. J. Rediscovering Silicones:
disiloxane. This material is available free of charge via the
“Unreactive” Silicones React with Inorganic Surfaces. Langmuir 2011,
Internet at http://pubs.acs.org.

■
27, 11514−11519.
(19) Cohen-Addad, J. P.; Huchot, P.; Jost, P.; Pouchelon, A.
AUTHOR INFORMATION Hydroxyl- or Methyl-Terminated Dimethylsiloxane Chains: Kinetics of
Corresponding Author Adsorption on Silica in Mechanical Mixtures. Polymer 1989, 30, 143−
*E-mail: alexander.fadeev@shu.edu. 146.

14806 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807

Langmuir Article

(20) Levresse, P.; Feke, D. L.; Manas-Zloczower, I. Analysis of the

Formation of Bound Poly(dimethylsiloxane) on Silica. Polymer 1998,
39, 3919−3924.
(21) Leger, L.; Hervet, H.; Deruelle, M. Adsorption of PDMS Chains
on Plane Silica Surfaces. In Adsorption on Silica Surfaces; Papirer, E.,
Ed.; Surface Science Series: Marcel Dekker: New York, 2000; Vol. 90,
pp 597−619.
(22) Marinova, K. G.; Christova, D.; Tcholakova, S.; Efremov, E.;
Denkov, N. D. Hydrophobization of Glass Surface by Adsorption of
Poly(dimethylsiloxane). Langmuir 2005, 21, 11729−11737.
(23) Sheka, E.; Natkaniec, I.; Khavryutchenko, V.; Nikitina, E.;
Barthel, H.; Weis, J. INS study of Intermolecular Interaction at the
Silicone-Fumed Silica Interface. Physica B 2000, 276-278, 244−246.
(24) Barthel, H.; Nikitina, E. INS and IR study of Intermolecular
Interactions at the Fumed Silica-Polydimethylsiloxane Interphase, Part
3. Silica-Siloxane Adsorption Complexes. Silicon Chem. 2004, 1, 261−
279.
(25) Smith, J. S.; Borodin, O.; Smith, G. D.; Kober, E. M. A
Molecular Dynamics Simulation and Quantum Chemistry Study of
Poly(dimethylsiloxane)-Silica Nanoparticle Interactions. J. Polym. Sci.,
Part B: Polym. Phys. 2007, 45, 1599−1615.
(26) Voronkov, M. G.; Mileshkevitch, V. P.; Yuzhelevskii, Yu. A. The
Siloxane Bond; Consultants Bureau: New York, 1978.
(27) Fadeev, A. Y.; Kazakevich, Y. V. Covalently Attached
Monolayers of Oligo(dimethylsiloxane)s on Silica: A Siloxane
Chemistry Approach for Surface Modification. Langmuir 2002, 18,
2665.
(28) Fadeev, A. Y. Hydrophobic Monolayer Surfaces: Synthesis and
Wettability. In Encyclopedia for Surface and Colloid Science, 2nd ed.;
Somasundaran, P., Ed.; Taylor & Francis: New York, 2006; Vol. 4, p
2854.
(29) Polymer Handbook, 3rd ed.; Brandrup, J., Immergut, E. H., Eds.;
Wiley: New York, 1989.
(30) Zhuravlev, L. T. The Surface Chemistry of Amorphous Silica.
Zhuravlev Model. Colloids Surf. 2000, 173, 1−38.
(31) Kozlova, S. A.; Kirik, S. D. Post-Synthesis Covering in
Mesostructured Silicate Materials MCM-41 and SBA-15. Microporous
Mesoporous Mater. 2010, 133, 124−133.
(32) Dubois, L. H.; Zegarski, B. R. Reaction of Alkoxysilane Coupling
Agents with Dehydroxylated Silica Surfaces. J. Am. Chem. Soc. 1993,
115, 1190−1191.
(33) Grabbe, A.; Michalske, T. A.; Smith, W. L. Strained Siloxane
Rings on the Surface on Silica: Their Reaction with Organosiloxanes,
Organosilanes, and Water. J. Phys. Chem. 1995, 99, 4648−4654.
(34) Morrow, B. A.; Cody, I. A. Infrared Studies of Reactions on
Oxide Surfaces. 5. Lewis Acid Sites on Dehydroxylated Silica. J. Phys.
Chem. 1976, 80, 1995−1998.
(35) Bunker, B. C.; Haaland, D. M.; Michalske, T. A.; Smith, W. L.
Kinetics of Dissociative Chemisorption on Strained Edge-shared
Surface Defects on Dehydroxylated Silica. Surf. Sci. 1989, 222, 95−118.
(36) McCrate, J. M.; Ekerdt, J. G. Titration of Free Hydroxyl and
Strained Siloxane Sites on Silicon Dioxide with Fluorescent Probes.
Langmuir 2013, 29, 11868−75.
(37) Livingston, H. K. Cross-sectional Areas of Molecules Adsorbed
on Solid Surfaces. J. Am. Chem. Soc. 1944, 66, 569−573.
(38) Clarkson, S. J.; Semlyen, J. A. Studies of Cyclic and Linear
Poly(dimethylsiloxanes): 21. High Temperature Thermal Behavior.
Polymer 1986, 27, 91−95.

14807 dx.doi.org/10.1021/la5031763 | Langmuir 2014, 30, 14797−14807