Hierarchical nanostructures in semifluorinated norbornene block copolymers

Article abstract of DOI:10.1021/ma102195y

Structure and dynamics of two microphase-separated norbornene block copolymers containing one comb-like component are investigated by X-ray scattering, relaxation spectroscopy as well as differential scanning calorimetry. It is shown that two self-assembling processes on different length scales coexist for these copolymers leading to hierarchically structured systems. Microphase separation of both block copolymer components appears on a scale of 20-25 nm while norbornene main and long semifluorinated side chains of the comb-like block nanophase separate within their domains on a scale of about 3 nm. This hierarchical structure is accompanied by a complex relaxation behavior which is characterized by two glass transitions. The glass temperature of the block with long bulky side chains is expectedly lower due to internal plasticization. An additional relaxation process showing typical features of a dynamic glass transition is related to an independent dynamics in small nanodomains formed by long aggregated side chains similar to the findings for many other comb-like polymers. Indications for the disappearance of the nanophase separation of main and semifluorinated side chains at high temperatures are discussed.

Full text of DOI:10.1021/ma102195y

958 Macromolecules 2011, 44, 958–965
DOI: 10.1021/ma102195y
Hierarchical Nanostructures in Semifluorinated Norbornene
Block Copolymers
B. Pulamagatta,*,† S. Pankaj,‡ M. Beiner,*,‡,§ and W. H. Binder†
†Chair of Macromolecular Chemistry, Institute of Chemistry, Division of Technical and Macromolecular Chemistry,
Faculty of Natural Sciences II, Martin-Luther-University Halle-Wittenberg, Von-Danckelmann-Platz 4, D-06120 Halle,
Germany, ‡Faculty of Natural Sciences II, Martin-Luther-University Halle-Wittenberg, D-06099 Halle, Germany, and
§Fraunhofer Institute for Mechanics of Materials, Walter-Hu€lse-Str. 1, D-06120 Halle, Germany
Received September 24, 2010; Revised Manuscript Received December 31, 2010
ABSTRACT: Structure and dynamics of two microphase-separated norbornene block copolymers contain-
ing one comb-like component are investigated by X-ray scattering, relaxation spectroscopy as well as
differential scanning calorimetry. It is shown that two self-assembling processes on different length scales
coexist for these copolymers leading to hierarchically structured systems. Microphase separation of both
block copolymer components appears on a scale of 20-25 nm while norbornene main and long semifluori-
nated side chains of the comb-like block nanophase separate within their domains on a scale of about 3 nm.
This hierarchical structure is accompanied by a complex relaxation behavior which is characterized by two
glass transitions. The glass temperature of the block with long bulky side chains is expectedly lower due to
internal plasticization. An additional relaxation process showing typical features of a dynamic glass
transition is related to an independent dynamics in small nanodomains formed by long aggregated side
chains similar to the findings for many other comb-like polymers. Indications for the disappearance of the
nanophase separation of main and semifluorinated side chains at high temperatures are discussed.
1. Introduction
X-ray diffraction pattern in the range 2 e q e 6 nm-1 (often
called prepeaks in the recent literature) corresponding to an
equivalent Bragg spacing of 1-3 nm and (ii) an independent
polyethylene-like dynamics within the alkyl nanodomains, seen
as RPE process in relaxation data at low temperatures, which
shows typical features characteristic for other cooperative relaxa-
tion processes.12,14 Both effects are observed for various series of
side chain polymers containing long alkyl groups. They are
basically independent of microstructure and softening behavior
of the main chain. Hence it was concluded that nanophase
separation is a general phenomenon in such systems.14,15 Tem-
perature dependent measurements on selected series show that
the behavior is not changing at high temperatures, i.e., that the
underlying structure is relatively unaffected by temperature
changes.13 A theoretical description of nanophase separation
phenomena, however, is still not available, although first ap-
proaches have been developed and applied to systems where the
side chains are noncovalently bonded to the backbone16 or
weakly grafted systems.17
Nanophase separation effects due to demixing of incompatible
main and side chains are also expected and found in cases where
the side chains are no alkyl groups. Lamellar structures with
typical periodicities in the 1-3 nm range are reported in parti-
cular for polymers with fluorinated or semifluorinated side
chains18-20 as well as long poly(ethylene oxide) sequences as side
chains.21 From the structural point of view the situation at room
temperature seems to be similar to that in systems with alkyl side
chains but it is unclear (i) whether or not there is also an
independent dynamics within the side chain domains in these
cases and (ii) to what extent changes in the nanophase separation
behavior appear with temperature. One aim of this paper is to
answer such questions based on a systematic study of two
norbornene diblock copolymers containing a comb-like semi-
fluorinated component. The morphology of these hierarchically
structured systems is studied by scattering experiments and their
Self-assembled polymers are an important class of materials
for various applications requiring well-defined structures with
typical dimensions below 50 nm.1 Archetype examples are diblock
copolymers containing two incompatible components which
aggregate in domains with typical dimensions comparable to
the radius of gyration of the individual blocks (typical values are
5-30 nm). Depending on the volume fraction of both blocks,
systems with different phase morphologies like lamellar, cylind-
rical, spherical, or gyroidal structures will form. Phase behavior
and phase separation processes in block copolymers are well
understood and thermodynamic models to describe the situation
are established.2,3
In the recent years, there has been an increasing interest in
systems with more complicated phase structure and in particular
in self-assembled polymers with a hierarchical structure incor-
porating different well-defined length scales.4-6 One approach to
design such systems is a combination of classical microphase
separation effects in block copolymers with self-assembling
tendencies which exist in comb-like polymers leading to lamellar
structures with typical dimensions in the 1-3 nm range.7-10
Lamellar packing of main and side chains is well-known for side
chain polymers containing stiff units at the end of the side chains,
which can appear in a liquid crystalline state7,8 or systems with
long alkyl side chains where the methylene units far away from
the backbone can crystallize.9,10 However, nanophase separation
of incompatible main and side chain parts is also indicated for
comb-like polymers containing shorter, noncrystallizable alkyl
groups in the side chain.11-13 The alkyl groups in these systems
also aggregate to form small alkyl nanodomains surrounded by
main chains. Nanophase separation in fully amorphous side
chain polymers is indicated by (i) relatively broad peaks in
*Corresponding author.
pubs.acs.org/Macromolecules
Published on Web 01/25/2011
r 2011 American Chemical Society

Article
Macromolecules, Vol. 44, No. 4, 2011 959
relaxation behavior is investigated by different relaxation spec-
troscopy methods as well as calorimetry in order to understand
structure-property relationships. Temperature-dependent changes
in the structure observed above 30 °C might be interesting for
a better understanding of nanophase separation phenomena in
the future.
Scheme 1. Structure of the Investigated Am-b-Cn Block Copolymers
2. Experimental Section
2.1. Methods. X-ray Scattering Experiment. X-ray scattering
measurements were performed on a small angle instrument as-
sembled by JJ X-rays based on a 2D Hi-Star detector by Bruker
and a Rigaku rotating anode with focusing optics. Cu KR radia-
chloride. Further 1H,1H,2H,2H-perfluoro-1-octanol (15.78 mL,
34.3 mmol) and pyridine (3.68 mL, 50.2 mmol) were dissolved in
40 mL of dry CH2Cl2. Under ice cooling endo,exo[2.2.1] bicyclo-
2-ene-5,6-dicarboxylic acid chloride (2.42 g, 11.42 mmol) was
dropped into the reaction mixture and stirred overnight at room
temperature. The reaction mixture was filtered to remove the
pyridinium salt and extracted with CH2Cl2. The organic layer
was extracted with 2 N HCl solution, saturated sodium bicar-
bonate and dried with sodium sulfate. The solvent was removed
under reduced pressure. Finally, the product was purified using
column chromatography with petrolether/ethyl acetate (10:1) as
the solvent mixture to yield 6.96 g (70%) of monomer C as a
white solid.
˚
tion with a wavelength λ=1.54 A was used for all measurements.
The scattering angles θ of the instrument were calibrated using
silver behanate as a reference material. For temperature-dependent
measurements a Linkam hot stage was used with equilibration
time of 15 min. Small angle X-ray scattering (SAXS) experiments
were performed at scattering vectors q=4π(sin θ)/λ in the range
-1
-1
˚
˚
0.009 A e q e 0.15 A and medium angle X-ray scattering
.
-1
-1
˚
(MAXS) data were recorded in the range 0.15 A e q e 0.60 A
˚
Dynamic Mechanical Analysis. Dynamic shear modulus G*=
G0 þ iG00 was measured using an Anton Paar MCR501 instru-
ment. A control strain of 0.015% was used. Isotherms were
measured at angular frequencies ω = 2πf in the range 0.1 rad/se
ω e100 rad/s with five points per decade. Equilibration time at
each temperature was 180 s with a temperature step of 4 K. For
the temperature scans, a rate of 1 K/min was used at a constant
angular frequency of 10 rad/s with point density of 2 points/min.
All the measurements were performed in a controlled nitrogen
gas atmosphere.
Differential Scanning Calorimetry. A Perkin-Elmer Diamond
DSC was used for measurements with heating rates of dT/dt =
10 K/min. Samples with a mass of about 5 mg were encapsulated
in standard 10 μL pans.
1HNMR(δppm, CDCl3, 400 MHz): 6.27 (dd, J= 5.59, 3.12 Hz,
1H), 6.06 (dd, J= 5.61, 2.81 Hz, 1H), 4.36 (td, JH-F=30.16, 6.35 Hz,
4H), 3.40-3.20 (m, 2H), 3.17-3.06 (m, 1H), 2.66 (dd, J= 4.57,
1.65 Hz, 1H), 2.44 (tq, J1FC=19.41, J=6.32 Hz, 4H), 1.64-1.41
(m, 2H).
Polymer Synthesis. The homopolymer A200 was synthesized
via ROMP using Grubbs first generation catalyst according to
the procedures developed previously in our laboratory.22,23 The
substituted norbornene block copolymers, Am-b-Cn, in which
one block (block A) containing two methyl ester groups
(COOCH3) per repeating unit and the other (block C) contain-
ing two fluorinated alkyl ester (COOCH2CH2(CF2)5CF3) side
chains were synthesized by the sequential addition of respective
monomers using Grubbs first generation catalyst. The general
synthetic procedure of block copolymer Am-b-Cn is described below,
for A50-b-C13 as an example. Monomer A (240.2 mg, 1.2 mmol) in
2 mL of CH2Cl2 was added to the Grubbs first-generation
catalyst, [RuCl2(PCy3)2(CHPh)] (18.8 mg, 0.022 mmol) dis-
solved in 2 mL of CH2Cl2 in a heated and argon-flushed glass
vial equipped with a magnetic stir bar. The polymerization was
carried out at room temperature for 2 h until all of monomer
A was consumed, as checked by TLC. Monomer C (259.8 mg,
0.3 mmol) as a solution in 2 mL of CH2Cl2 was then added to the
above reaction mixture and stirred for 7 h at room temperature
until all of the monomer C was consumed, as checked by TLC.
The polymerization was quenched by adding cold ethyl vinyl
ether. The produced polymer was isolated by precipitating in to
cold methanol. Finally the product was dried under high
vacuum overnight to yield 485 mg (97%) of A50-b-C13. The
other block copolymer A50-b-C8 with different block composition
was synthesized using the above stated procedure but adopting
the different monomer/initiator ratio. The general structure of
Am-b-Cn is given in Scheme 1.
Before establishing the above stated polymerization proce-
dure, the increase in molecular weight (Mn) with the poly-
merization time was monitored by GPC. Monomer A was first
polymerized for 2 h to obtain living polymer A50 chain with
molecular weight (Mn(GPC)=8.4 kg/mol) which was comparable
to the calculated molecular weight (Mn(CAL)) with low polydisper-
sity index (PDI=1.17). Furthermore, monomer C was added to
the above reaction mixture. After 4 h of adding monomer C the
Mn was increased to 16.4 kg/mol according to GPC measure-
ments suggesting addition of ∼10 units of monomer C onto the
polymer A50 chain and confirming the crossover reaction.
However further proceeding with the reaction no remarkable
increase in the molecular weight was noticed, instead there was
an increase in the PDI value and also precipitation occurred.
Dielectric Spectroscopy. The dielectric function ε*(ω) =
ε0(ω) - iε00(ω) was measured using a Novocontrol Alpha analyzer
in the frequency range 10-1 Hz e f e 105 Hz. Measurements
were performed on films with 25 μm thickness pressed between
gold plated brass electrodes with a diameter of 20 mm using kapton
spacers. Isotherms were measured at temperatures between -120
and þ140 °C with a temperature step of 5 K.
2.2. Synthesis of the Samples. Materials. Grubbs first gen-
eration catalyst was obtained from Sigma-Aldrich. Dichloro-
methane (CH2Cl2) was freshly distilled over CaH2 and degassed
with argon prior to use. The other solvents like petrolether and
ethyl acetate were used after distillation. All other reagents were
purchased from Sigma-Aldrich (Germany) and were used with-
out further purification.
Sample Characterization. The chemical characterization of
synthesized compounds was done via 1H NMR measurements
performed on a Varian Gemini 200 or 400 MHz FT-NMR
spectrometer using MestRec (4.9.9.9) for the data evaluation.
Molecular weight distributions and polymerization kinetics were
determined by GPC measurements done on a Viscotek VE2001
system using polystyrene standards for conventional external
calibration with Viscotek VE3580 refractive index detector.
Monomer Synthesis. endo,exo-Bicyclo[2,2,1]-hept-5-ene-2,
3-dicarboxylic acid dimethylester (monomer A) was prepared
according to the procedure from ref 22, endo,exo-bicyclo[2,2,1]-
hept-5-ene-2,3-dicarboxylic acid bis(3,3,4,4,5,5,6,6,7,7,8,8,8-tri-
decafluorooctyl) ester (monomer C) was prepared according to
modified procedure from above-mentioned ref 22 (described
below). A dry round-bottom flask equipped with magnetic stir
bar was flushed with argon and charged with endo,exo-bicyclo-
[2,2,1]-hept-5-ene-2,3-dicarboxylic acid (2.5 g, 13.73 mmol) and
excess thionyl chloride (SOCl2) (8.23 mL, 0.114 mol). The
mixture was refluxed for 4 h at 90 °C and subsequently excess
of thionyl chloride was removed under reduced pressure to
obtain endo,exo-bicyclo[2,2,1]-hept-5-ene-2,3-dicarboxylic acid

960 Macromolecules, Vol. 44, No. 4, 2011
Pulamagatta et al.
Table 1. Characterization of the Polymers
Mn(GPC)
Mn(cal) (kg/mol) (kg/mol) PDI Tg,A (°C) Tg,C (°C)
sample
A200
A50-b-C13
A50-b-C8
42.0
21.9
17.5
40.0
16.4
13.0
1.17
1.20
1.20
72
70
67
25
22
Figure 2. X-ray scattering curve in the mediate q range (0.15 A-1 e q e
˚
0.50 A-1) for (a) A50-b-C13 and (b) A50-b-C8. Heating scans (from -50
˚
to þ90 °C) and subsequent cooling (þ90 °C to -50 °C) scans are shown
from the top to the bottom. The curves are vertically shifted.
samples does not show a significant change with temperature;
i.e., the block copolymer morphology is preserved in the inves-
tigated temperature range. Both block copolymers are obviously
strongly segregated at temperatures below 90 °C and far away
from any order-to-disorder transition (ODT). The observed
change from cylindrical to lamellar morphology upon decreasing
the length of block C from 13 to 8 monomeric units indicates that
the C phase forms the matrix in block copolymer A50-b-C13, i.e.,
that cylindrical A domains are surrounded by a continuous C
phase in this case. Although block C has a lower degree of
polymerization compared to block A, the volume fraction of the
C phase seems to be large due to the bulkiness of the semifluori-
nated alkyl side chains.24
Figure 1. Small angle X-ray scattering curves for (a) A50-b-C13 and (b)
A50-b-C8 measuredat differenttemperatures. Thedata setsare vertically
shifted for the sake of clarity.
An analysis of integral values in 1H NMR spectra suggested that
only 13-16 units of C had added onto the polymer A chain (see
Supporting Information for GPC and NMR results). The precipi-
tation indicates that the limited solubility of the block copolymer is
responsible for stopping the polymerization. Thus, a maximum
of 13-16 repeating units of monomer C could be polymerized
onto the polymer A chain irrespective of the chain length of A.
Two block copolymers were synthesized keeping the block A
length constant to 50 units. The length of the C block is varied.
The C block of sample A50-b-C8 contains 8 C units, while that of
the A50-b-C13 contains 13-16 C units. Further information
about the microstructure of the synthesized polymers is sum-
marized in Table 1.24
Nanophase separation of the semifluorinated side chains from
the norbornene main chains in the domains containing block C is
evidenced by X-ray scattering data in the intermediate q range
-1
(0.15 A e q e 0.50 A-1). Figure 2 shows corresponding
scattering pattern for A50-b-C13 and A50-b-C8 respectively. Heat-
ing (-50 °C f þ90 °C) and cooling (þ90 °C f -50 °C) scans are
performed in order to check the reversibility of temperature
dependent changes. A middle angle X-ray scattering (MAXS)
˚
˚
3. Results
Microphase separation and morphology of the synthesized
Am-b-Cn block copolymers are analyzed by small-angle X-ray
˚
scattering (SAXS) experiments at scattering vectors in the range
peak at qnps=0.22 A-1 is seen at temperatures below 70 °C for
-1
of 0.009 A e q e 0.15 A-1. Figure 1 shows SAXS pattern for
the block copolymers A50-b-C13 and A50-b-C8 measured at
temperatures between -50 and þ90 °C. The presence of prominent
higher order reflections up to the fourth order for both samples is
a clear evidence for the existence of well ordered periodic structures
formed due to microphase separation of A and C blocks. In case
both the samples. This maximum position corresponds to equiva-
lent Bragg spacings dnps of about 2.9 nm. Broad peaks in this
q range are commonly observed for amorphous comb-like poly-
mers containing long alkyl side chains due to nanophase separa-
tion of main and side chain parts.13-15,25 For the investigated
block copolymer systems the peak in the MAXS range indicates
demixing of norbornene backbone and semifluorinated alkyl side
chains of block C. Note that the MAXS peaks are extremely
broad indicating that there is no pronounced long-range order.
Interestingly, the MAXS peak gradually vanishes upon heating
between 30 and 90 °C and reappears upon cooling in the same
temperature range (Figure 2). This observation suggests that the
nanophase separation, i.e. the demixing of semifluorinated alkyl
side chains from the main polymer chains, occurs only below this
temperature interval while both subunits are mixed in the C phase
˚
˚
-1
˚
of A50-b-C13, the first order peak appears at q1 = 0.026 A
corresponding to a long period of d1=24 nm according to Bragg’s
law (d1 = 2π/q1). The relative peak positions of the higher order
√
√ √
peaks with q1 are in the ratio of 1: 4: 7: 9 suggesting a
hexagonally ordered cylindrical structure. In case of the A50-b-C8
-1
˚
sample a narrow first order peak appears at q1 = 0.030 A
related to a long period of d1 = 21 nm. The higher order peaks are
equidistantly spaced relative to q1 (1:2:3:4) confirming a lamellar
morphology of this sample. The scattering pattern of both

Article
Macromolecules, Vol. 44, No. 4, 2011 961
relaxation processes of the A50 block are expectedly quite
comparable to that for the A200 homopolymer. Additional relaxa-
tion processes in the shear curves for the two block copolymers
at -29 and þ22 °C originate obviously from block C. DSC
measurements on Am-b-Cn block copolymers give glass tempera-
tures for block C in the range Tg,C=23 ( 2 °C (Figure 4, Table 1).
Hence, the relaxation process occurring at TR,C=22 °C can be
assigned to the segmental relaxation RC of block C. The glass
temperature of block C is reduced in comparison to that of block
A due to internal plasticization of the main chains by the highly
flexible side chains as discussed by Heijboer for poly(n-alkyl
methacrylates).26 The Tg shift in norbornene with semifluorinated
side chains, however, seems to be smaller compared to that for
PnAMAs with similar side chain length. The Tg difference between
atactic PMMA and PnOMA is about 120 K12 while the Tg
difference between the A and C components of our block copo-
lymers with similar side chain lengths is only 55 K. The additional
relaxation process Rfsc,C at -29 °C is probably related to co-
operativemotions withinsmall self-assembledsidechain domains
occurring due to nanophase separation within phase C at low
temperatures. The Rfsc,C process can be understood as an addi-
tional indication for the presence of nanoscopic domains formed
by segregation of long semifluorinated side chains. Unfortunately,
it was not possible to prove this interpretation based on relaxa-
tion data for a neat C homopolymer since the synthesis of such
systems is not feasible yet due to solubility limitations. Note that
it remains open here whether the β relaxation appearing in both
block copolymers near -115 °C is only caused by localized
motions in the A phase. It may also incorporate localized motions
of similar nature in the C phase which are strongly superimposed
with contributions from the A phase (β=βA þ βC). A clarification
of the nature of the underlying motions requires studies with
more selective methods like solid state NMR spectroscopy.
Considering the microstructure of our Am-b-Cn block copoly-
mers (Scheme 1) dielectric spectroscopy should be also a suitable
method to study the relaxation behavior since the polar carboxyl
groups (-C(dO)O-) in both blocks act as dielectrically active
centers. An additional dipole arises in the side chains of the block
C originating from the uncompensated end fluorine atom.27
Hence dielectric measurements in the frequency range from 10-1
to 105 Hz have been performed at temperatures between -140
and þ140 °C.
Figure 3. Shear storage G0 (0) and shear loss modulus G00 (O) as
function of temperature for (a) A200, (b) A50-b-C13, and (c) A50-b-C8.
The vertical dotted lines indicate the temperature positions of the
different relaxation processes (RA, RC, Rfsc,C, β).
Figure 4. DSC heating scans measured with a rate of dT/dt = þ10 K/
min for homopolymer A200 and both block copolymers A50-b-C8 as well
as A50-b-C13. The heat capacity curves for A50-b-C8 and A200 are
vertically shifted by 0.5 and 1 J/gK, respectively.
The discussion here will be focused on results obtained at low
temperatures (-140 °C e T e þ10 °C) showing two secondary
relaxations (β and γ) as well as the Rfsc,C process (Figure 5). The
dielectric loss curves ε00(ω) measured at temperatures above 10 °C
are dominated by strong conductivity contributions (ε00 ∼ σ/ε0ω).
Hence the RA, RC, and Rfsc,C processes are practically hidden in
these isotherms or only seen as a weak shoulder which cannot be
seriously evaluated. The low temperature isotherms ε00(ω) for the
homopolymer A200 show a very broad and weak relaxation peak
which shifts expectedly to higher frequencies if the temperature
increases (Figure 5a). Its peak width decreases with increasing
temperature. On the basis of a comparison of the peak maxima
positions with those taken from shear curves (Figure 6) as well as
a closer inspection of peak shape and width as a function of
temperature one can conclude that this peak contains contribu-
tions from two secondary relaxation processes, β and γ, indicating
the appearance of two different types of localized motions in
glassy A200. Unfortunately, it is not possible to deconvolute the
contributions of both relaxation processes in this case. Thus, only
the maxima of the β process appearing at the low frequency end
of the broad peak will be incorporated in the further discussion.
The maxima positions are taken directly from the ε00 isotherms as
indicated by ticks in Figure 5a.
at higher temperatures. Similar behavior has not been reported to
our knowledge for nanophase-separated polymers containing alkyl
side chains which are covalently bonded to different backbones.
Dynamic shear modulus G*(T) = G0(T) þ iG00(T) data, with G0
being the storage and G00 the loss modulus, for homopolymer
A200 as well as the two block copolymers A50-b-C13 and A50-b-C8
are shown in Figure 3. Two independent relaxation processes at
around -115 and þ77 °C are observed in 10 rad/s isochrones for
homopolymer A200 (Figure 3a). The low temperature process
at -115 °C showing an Arrhenius-like temperature dependence
(see below) corresponds to a local mode β, whereas the process
at 75 °C is the segmental relaxation RA caused by co-operative
motions of complete monomeric units leading to a softening of
the entire system. The glass temperature Tg,A taken from DSC
scans performed with a heating rate of 10 K/min (Figure 4,
Table 1) is comparable to the relaxation temperature for the RA
process, TR,A, as obtained from shear measurements with an
angular frequency ω of 10 rad/s. The β and RA processes, which
are already seen for the neat homopolymer A200, do also appear
in the microphase-separated block copolymers A50-b-C13 and
A50-b-C8 at -120 °C and at 77 °C (Figure 3, parts b andc), i.e., the
Dielectric loss isotherms for the block copolymer A50-b-C13
measured at comparable temperatures are given in Figure 5c.

962 Macromolecules, Vol. 44, No. 4, 2011
Pulamagatta et al.
Figure 6. Arrhenius plot for the A200 homopolymer (squares) and the
block copolymers A50-b-C13 (stars) and A50-b-C8 (circles) obtained from
peak maxima positions of shear isochrones (open symbols) and di-
electric isotherms (half filled symbols). The five different relaxation
processes (RA, RC, RFS, β, γ) appearing in these samples are labeled.
DSC glass temperatures from scans with a heating rate of 10 K/min
(solid symbols) are shown for comparison. Straight lines indicate
Arrhenius fits as discussed in the main text.
parts b and d) in accordance with shear data for both block
copolymers. Like discussed above this relaxation process should
reflect motions within nanodomains formed by aggregated side
chains of the C block. The small dielectric relaxation strength of
the Rfsc,C process and the temperature dependence of the relaxa-
tion frequency ωRsc which corresponds to that obtained for the
Rfsc,C process in shear data (Figure 6) fit to this interpretation.
Unfortunately it is impossible to get quantitative information
about the temperature dependence of relaxation strength and
shape for the Rfsc,C peak in our systems since it is located in the
valley between the RC and β/γ peaks. In this frequency range
various relaxation processes do superimpose. Clear is that the Rfsc,C
peak moves out of the frequency window of our measurements
above 10 °C significantly below the temperature where the MAXS
peak disappears in the diffraction pattern as shown in Figure 2.
Thus, it is impossible to study the influence of structural changes
above 30 °C on the Rfsc,C relaxation dynamics directly. Note that
a relaxation process of semifluorinated side chains in practically
the same frequency-temperature range has been reported for
block copolymers where such side chains with different lengths
(C4F9, C8F17) have been attached to a polybutadiene backbone.
For the longer side chain, this peak has a bimodal relaxation time
distribution in this case which was interpreted as an indication for
the anisotropy of the underlying motions.26
Figure 5. Dielectric loss ε00 isotherms for (a) A200, (b & c) A50-b-C13, and
(d & e) A50-b-C8. The temperature limits are labeled in the individual plots.
The temperature step between the shown isotherms is always 10K.
The results show clearly that the shape of the peak in ε00(ω) has
changed. The peak maximum appears now at the high frequency
end of the broad peak and β contributions dominating in the
isotherms for the A200 homopolymer seem to be strongly sup-
pressed. The γ process dominates obviously in case of this block
copolymer. Similar behavior with slightly stronger contributions
of the β process is observed for the A50-b-C8 block copolymer
(Figure 5e). The experiments show commonly that the γ relaxa-
tion process is amplified and the β relaxation is suppressed if
semifluorinated side chains (C units) are incorporated in the
system. This is interesting since the main chains of both blocks, A
and C, are identical. The only difference between both compo-
nents is the long, semifluorinated side chain carrying only a small
dipole but contributing significantly to the volume of the C block.
Such a significant difference between the dielectric spectra for
A200 and both block copolymers is somehow unexpected since the
number of monomeric units (and (-C(dO)O-) groups) in the C
block is small as compared to that in the A block. However, one
has to take into account that the detected dielectric signals are
tiny and that the long semifluorinated alkyl side chains of the C
block possess an additional dipole. The results seem to indicate
that the bulky side chains either amplify the γ relaxation which
exists also for A polymers without side chains or that an extra
relaxation process γC appears in the C block at a very similar
frequency-temperature position. The latter case seems to be
unlikely but cannot be completely excluded.
An Arrhenius plot summarizing the information about the
temperature dependence of all five relaxation processes (RA, RC,
R
fsc,C, β and γ) in our An-b-Cm block copolymers is given in
Figure 6. The RC and Rfsc,C relaxations appear only in the C block
and have not been found for the A200 homopolymer. The results
from shear and dielectric spectroscopy do basically coincide.
Small shifts of the order of half a decade between the loss maxima
positions in dielectric and shear curves might be due to the fact that
both methods measure different susceptibilities and accordingly
also different fluctuations.28 Note that the frequency temperature
position of all relaxation processes is nearly identical if A50-b-C13
and A50-b-C8 samples are compared, i.e. the dynamics is practi-
cally independent of the block copolymer morphology. The
segmental dynamics (RA) of the A block in both microphase-
separated block copolymers corresponds to that in homopolymer
A200 containing the same monomeric units but having higher
molecular weight. The segmental relaxation RC of the C block
with long semifluorinated side chains appears at significantly
lower temperature. The glass temperatures from DSC scans with
The Rfsc,C process of the C block is observed in the dielectric
data at intermediate temperatures (-50 °C e T e 10 °C, Figure 5,

Article
Macromolecules, Vol. 44, No. 4, 2011 963
Figure 7. Scheme of the structural situation in the block copolymers A50-b-C13 (left) and A50-b-C8 (right) with cylindrical and lamellar morphology,
respectively. The zoom in the middle shows the situation at the interface and the internal structure of the nanophase separated phase C.
a heating rate of 10 K/min are comparable with estimates deter-
mined based on an extrapolation of RA and RC traces assuming
that the relaxation time at Tg is approximately 100 s. Interestingly,
main and side chains in the C domains results in an internal
structure with typical periodicities dnps of about 2.9 nm. Like
in side chain polymers containing long alkyl groups, nano-
phase separation of the C block with long partly fluorinated
side chains is indicated by (i) a peak in the MAXS range at q
the estimated steepness indices m = -[log ω]/d(Tg/T)|T=T
g
are seemingly quite different for both blocks. The A block
shows the behavior of a fragile glass indicated by large
mR,Α ∼ 136 values while block index of mR,C ∼ 48 corresponding
to values obtained for strong glasses. A reduction of m with side chain
length has been also observed for other polymer series where
internal plasticization appears like poly(n-alkyl methacrylates)
containing alkyl groups with different length.29 The β and γ
processes of A and C block show the typical features of ordinary
secondary relaxation processes related to localized motions.
Their temperature dependence can be described by an Arrhenius
law ω=Ω exp(-EA/RT) with limiting frequencies Ω which are in
both cases close to 1015 rad/s and activation energies which are
∼ 0.22 A-1 and (ii) a main chain independent dynamics in the
˚
side chain domains (Rfsc,C) showing a non-Arrhenius-like
temperature dependence. Obviously, nanophase separation
can also occur under confinement, i.e., in self-assembled C
block domains which are only 10 times larger than dnps.
The morphology of the investigated block copolymers
with cylindrical (A50-b-C13) and lamellar (A50-b-C8) overall
structure should be similar to that sketched in Figure 7. A
similar overall structure has been proposed for other block
copolymers containing comb-like blocks with long side
chains in the amorphous, semicrystalline or liquid-crystalline
state.7-10,19 Note that the backbones of the C block are
aligned basically perpendicular to the interfaces between
both microphases and that the fluorinated side chains are
noninterdigitated. Only under this condition one can explain
the observed dnps = 2.9 nm values which are nearly twice
larger than the extended length of the partly fluorinated side
chains (1.4 nm) of the C block. Noninterdigitated side-chains
have been also reported for comb-like homopolymers with
partly fluorinated side chains18-20 in contrast to the findings
for many systems with alkyl side chains.12,14,25,30 This might
be reasonable considering the fact that in case of noninterdi-
gitating semifluorinated side chains the contacts between
alkylated and fluorinated units are minimized which should
be energetically favorable.
Summarizing the discussed findings one can conclude that
the two structure formation mechanisms in the investigated
Am-b-Cn block copolymers can occur more or less indepen-
dently. These copolymers can optimize seemingly their pack-
ing on two different length scales in parallel and form a
structure-in-structure morphology. The competition between
both structure formation mechanisms is obviously limited,
i.e., side and mains chains of the comb-like block C can
manage to nanophase separate like under bulk conditions. In
some sense the situation seems to be different from that in
microphase-separated block copolymers containing crystal-
lizable components like polyethylene or poly(ethylene
oxide).31,32 In this case there is often a clear competition of
both structure formation mechanisms, microphase separa-
tion and crystallization, since the thickness of the folded
crystalline lamellae is typically similar to the size of the
domains formed due to microphase separation (∼10 nm).
EA,β=21.4 ( 3 kJ/mol and EA,γ=13.3 ( 3 kJ/mol for the β and γ
processes, respectively. More unusual is the temperature depen-
dence of the Rfsc,C process. There are clear indications for a non-
Arrhenius-like behavior although the Rfsc,C trace appears linear
in the Arrhenius plot in the accessible frequency window. An
Arrhenius fit as shown in Figure 6 gives an apparent activation
energy of about EA,fsc=68.8 ( 6 kJ/mol but a limiting frequency
Ω of about 1035 rad/s. This value is an unphysical extrapolation
since it is much higher than the upper frequency limit for relaxation
processes of about 1014 rad/s. The indicates clearly that the Rfsc,C
relaxation frequencies must deviate from the discussed Arrhenius
extrapolation, i.e., a non-Arrhenius-like temperature dependence
must occur above our measurement window and the slope of the
Rfsc,C trace changes at high temperatures. The steepness index
calculated from the trace of the Rfsc,C process is mfsc,C ∼ 30
corresponding to that of strong glasses. Similar behavior has been
reported for the RPE process in self-assembled alkyl nanodomains
reported for various nanophase-separated side chain polymers
containing long alkyl groups.14,15 We conclude that the Rfsc,C
process shows indeed typical features of a dynamic glass transition
and reflects most likely cooperative motions in nanodomains
formed by aggregated side chains of the partly fluorinated C block
with a typical size of about 2 nm.
4. Discussion
Hierarchy of Length Scales. The presented results show
clearly that the investigated, Am-b-Cn block copolymers are
self-assembled systems with a hierarchy of length scales in the
nanometer range. Classical microphase separation of incom-
patible A and C blocks leads to a large scale structure with
spacings of about 20-25 nm while nanophase separation of

964 Macromolecules, Vol. 44, No. 4, 2011
Pulamagatta et al.
Hence crystallization can destroy the original block copolymer
morphology and strong nonequilibrium effects do appear.33
The coexistence of two structure formation mechanisms
which do not strongly compete like in our hierarchically
structured block copolymers can be a useful strategy to
design functional materials which are close to equilibrium
but able to combine the properties of individual domains
with different dimensions. Note that nanophase separated
side chain polymers containing alkyl groups of different
length are materials with truly predictable structure-property
relations. Damping can be introduced in a controlled way
since the frequency temperature position of the RPE process
in the alkyl nanodomains depends systematically on the side
chains length (and nanodomain size dnps) but basically not
on the main chain properties.12,15 Similar behavior would be
expected for comb-like polymers in the nanophase-separated
state containing other side chains, e.g., long partly fluori-
nated sequences. In combination with a systematic variation
of the second block this can be used to fine-tune the mechanical
properties of self-assembled systems with hierarchical struc-
ture. Other properties of such polymers might be optimized
following the same strategy.
Early Stages of Nanophase Separation. An interesting finding
of this work is that the peak in the MAXS range indicating a
nanophase separation in phase Cconnected with an aggregation
of long partly fluorinated side chains shows significant changes
above 30 °C. The MAXS peak starts to broaden at this
temperature during heating, disappears finally at temperatures
around 90 °C but reappears during cooling. This observation
can be interpreted as an indication for a transition from the
nanophase-separated state to a disordered state without separa-
tion of main and side chains. Disappearance of the peak in the
MAXS range at temperatures near 70 °C has been also reported
for other polymeric systems which contain a component with
partly fluorinated side chains.34 Room-temperature ionic liquids
being small molecules containing long alkyl tails show similar
trends.35 Comb-like polymers containing long alkyl groups,
however, show usually a well pronounced scattering peak in
the MAXS range until degradation starts to appear.13 Hence
systems with partly fluorinated side chains might be useful
model systems to study the order-to-disorder transition (ODT)
in nanophase-separated polymers. To have such model systems
seems to be important since a theory which describes the
behavior of nanophase-separated side chain polymers quanti-
tatively is still missing. Information about the temperature
dependence of the scattering intensity close to an order-to-
disorder transition seems to be of special importance for the
development of such theoretical models.
ODT due to a transition to the disordered state without
separated main and side chain domains and superstructure
on the mesoscale. For the block copolymers studied in this
work the Rfsc,C process is unfortunately inaccessible in di-
electric spectra measured at the relevant temperatures (30-
90 °C). Part of the evaluation problem are strong relaxation
processes in the A block appearing at frequencies near the
Rfsc,C process which hide the latter in the ε00 isotherms
(Figure 5, parts c and e). A way out of this problem would
be homopolymers of type C, which are unfortunately not
producible yet due to solubility restrictions. Other comb-like
polymers where semifluorinated side chains are attached to
similar backbones might be an alternative.
Summarizing the discussion in this part we think that
questions related to the nanophase separation behavior should
be in the focus of further investigations on semifluorinated
side chain polymers with a microstructure similar to that of
the C block in our copolymers. One way to clarify details of
their complex structure might be studies on oriented samples.36
Further studies may contribute to a better understanding of
the driving forces causing nanophase separation in side chain
polymers which are still not finally clear. This could also help
to make useful predictions about structure and properties of
comb-like polymers based on the knowledge about their
microstructure.
5. Conclusions
We have shown in this work that norbornene based block
copolymers containing a comb-like component with semifluori-
nated side chains are systems which are hierarchically structured
on the nanoscale. They show a structure-in-structure morphology
since a microphase separation of both blocks (periodicity 20-25 nm)
coexists with a nanophase separation of main and side chains in
the phase formed by the comb-like component (periodicity
∼3 nm). Although the length scales are not too different, both self-
assembling mechanisms can occur. The energetic optimization of
the systems seems to require both. The relaxation behavior of the
investigated block copolymers is rather complex and incorpo-
rates a relaxation process (Rfsc) which is seemingly related to an
independent dynamics in nanodomains formed by semifluori-
nated side chains with typical dimensions of about 2 nm. X-ray
scattering data strongly indicate that the nanophase separation in
the comb-like component disappears at temperatures between 30
and 90 °C. This finding for the semifluorinated block is interest-
ing since polymeric systems showing that the nanophase separa-
tion of main andside chains disappears at accessible temperatures
are not well investigated but important for a deeper understanding
of the driving forces of such structure formation processes on
extremely small length scales. Progress with a theoretical description
of this phenomenon and microstructure-based predictions would
be of importance for commercial users since comb-like polymers
are used as functional materials in various fields of application.
Although clear indications for an ODT belonging to the
nanophase separation are seen in the data for our Am-b-Cn
block copolymers, there are drawbacks preventing a detailed
analysis. The problems have to do with the fact that the
investigated systems are copolymers and only a certain
fraction of the signal is coming from the comb-like C block
which is undergoing nanophase separation. (i) The relatively
Acknowledgment. The authors thank A. Petzold for assistance
with the SAXS data analysis based on the interfacial distribution
function. W.H.B. acknowledges grants from the DFG (INST
271/249-1, INST 271/247-1, and INST 271/248-1). M.B. thanks
for financial support by the DFG (BE 2352/4-1) and the state
Sachsen-Anhalt.
-1
˚
broad MAXS peak in the scattering pattern at q ∼ 0.22 A
has a small intensity since only the C phase contributes.
Moreover, nanophase and microphase separated structures
coexist and the corresponding scattering peaks do partly
superimpose. Both effects lead to a situation that makes it
very complicated to analyze the temperature dependence
of intensity and width of the MAXS peak seriously. (ii)
Changes in the relaxation behavior, in particular the Rfsc,C
process, near the speculated ODT would be also of major
interest. Such changes could support the given interpretation
since an independent dynamics within self-assembled side chain
domains should disappear or change qualitatively above an
Supporting Information Available: Table of molecular weight
(Mn) and PDI measured by GPC and figures showing GPC
curves for homopolymer A50 and block copolymer A50-b-C13, 1H
1
and 13C NMR spectra of monomers A and C and H NMR
spectrum of block copolymer A50-b-C13. This material is available
free of charge via the Internet at http://pubs.acs.org.

Article
Macromolecules, Vol. 44, No. 4, 2011 965
References and Notes
(24) Note that it is extremely hard to determine volume fractions in case
of our block copolymers since homopolymer C has not been success-
fully synthesized yet. Hence its density is unknown and one can not
use a combination of NMR data and density information to
estimate volume fractions. We have tried to estimate the volume
fractions for the A50-b-C8 sample with lamellar morphology based
on the interfacial distribution function K00(z). Assuming that C is
the minority component one gets ΦC = 0.4-0.45. As long the
densities are identical one can estimate based on that information
the volume fraction Φc for the A50-b-C13 sample. This gives values
of about 0.5-0.62. ΦC values up to 0.7 seem to be possible for
the A50-b-C13 sample considering an inverted phase situation in the
lamellar sample A50-b-C8 what is not completely excluded in the
light of the extended length of block C. Further experiments
incorporating additional block copolymer samples are required
to understand the situation in the phase diagram.
(1) Hamley, I. W. Nanotechnology 2003, 14, R39.
(2) Bates, F. S.; Fredrickson, G. H. Annu. Rev. Phys. Chem. 1990, 41,
525.
(3) Hamley, I. W. The physics of block copolymers; Oxford University
Press: Oxford, U.K., 1998.
(4) Muthukumar, M.; Ober, C. K.; Thomas, E. L. Science 1997, 277, 1225.
(5) Ikkala, O.; ten Brinke, G. Science 2002, 295, 2407.
(6) Ober, C. K.; Cheng, S. Z. D.; Hammond, P. T.; Muthukumar, M.;
Reichmanis, E.; Wooley, K. L.; Lodge, T. P. Macromolecules 2009,
42, 465.
(7) Schneider, A.; Zanna, J.-J.; Yamada, M.; Finkelmann, H.; Thomann,
R. Macromolecules 2000, 33, 649.
(8) Ansari, I. A.; Castelletto, V.; Mykhaylyk, T.; Hamley, I. W.; Lu,
Z. B.; Itoh, T.; Imrie, C. T. Macromolecules 2003, 36, 8898.
€
(9) Hempel, E.; Budde, H.; Horing, S.; Beiner, M. Lect. Notes Phys.
2007, 714, 201.
(10) Gitsas, A.; Floudas, G.; Butt, H. J.; Pakula, T.; Matyjaszewski, K.
(25) Arrighi, V.; Triolo, A.; McEwen, I. J.; Holmes, P.; Triolo, R.;
Amenitsch, H. Macromolecules 2000, 33, 4989.
(26) Heijboer, J. In Physics of non crystalline solids; Prins, J. A., Ed.;
North Holland: Amsterdam, 1965; p 231.
Macromolecules 2010, 43, 2453.
(11) Chen, W.; Wunderlich, B. Macromol. Chem. Phys. 1999, 200, 283.
(12) Beiner, M. Macromol. Rapid Commun. 2001, 22, 869.
(13) Arbe, A.; Genix, A.-C.; Colmenero, J.; Richter, D.; Fouquet, P.
Soft Matter 2008, 4, 1792.
(14) Beiner, M.; Huth, H. Nat. Mater. 2003, 2, 595.
(15) Pankaj, S.; Hempel, E.; Beiner, M. Macromolecules 2009, 42, 716.
(16) Stepanyan, R.; Subbotin, A.; Knaapila, M.; Ikkala, O.; ten Brinke,
G. Macromolecules 2003, 36, 3758.
(17) Milner, S. T. Macromolecules 1994, 27, 2333.
(18) Honda, K.; Morita, M.; Otsuka, H.; Takahara, A. Macromolecules
2005, 38, 5699.
(19) Wang, J.; Mao, G.; Ober, C. K.; Kramer, E. J. Macromolecules
1997, 30, 1906.
€
(27) Floudas, G.; Antonietti, M.; Forster, S. J. Chem. Phys. 2000, 113,
3447.
(28) Beiner, M.; Korus, J.; Lockwenz, H.; Schroter, K.; Donth, E.
€
Macromolecules 1996, 29, 5183.
(29) Floudas, G.; Stepanek, P. Macromolecules 1998, 31, 6951.
(30) Mierzwa, M.; Floudas, G.; Stepanek, P.; Wegner, G. Phys. Rev. B
2000, 62, 14012.
(31) Loo, Y.-L.; Register, R. A.; Ryan, A. J.; Dee, G. T. Macromole-
cules 2001, 34, 8968.
(32) Castillo, R. V.; Arnal, M. L.; Muller, A. J.; Hamley, I. W.;
Castelletto, V.; Schmalz, H.; Abetz, V. Macromolecules 2008,
41, 879.
(33) Loo, Y.-L.; Register, R. A.; Ryan, A. J. Macromolecules 2002, 35,
2365.
(20) Kim, B. G.; Chung, J.-S.; Sohn, E.-H.; Kwak, S.-Y.; Lee, J.-C.
Macromolecules 2009, 42, 3333.
(21) Neugebauer, D.; Theis, M.; Pakula, T.; Wegner, G.; Matyjaszewski, K.
(34) Al-Hussein, M.; Serero, Y.; Konovalov, O.; Mourran, A.; Moller,
M.; de Jeu, W. H. Macromolecules 2005, 38, 9610.
(35) Triolo, A.; Russina, O.; Bleif, H.-J.; Di Cola, E. J. Phys. Chem. B
2007, 111, 4641.
Macromolecules 2006, 39, 584.
(22) Binder, W. H.; Pulamagatta, B.; Kir, O.; Kurzhals, S.; Barqawi, H.;
Tanner, S. Macromolecules 2009, 41, 8405.
(23) Adekunle, O.; Tanner, S.; Binder, W. H. Beilstein J. Org. Chem.
(36) Clark, C. G.; Floudas, G. A.; Lee, Y. J.; Graf, R.; Spiess, H. W.;
€
2010, 6, No. 59.
Mullen, K. J. Am. Chem. Soc. 2009, 131, 8537.