Encapsulation of functional moieties within branched star polymers: Effect of chain length and solvent on site isolation

Article abstract of DOI:10.1021/ja003304u

Porphyrin and pyrene photoactive cores have been encapsulated within an isolating polymeric shell using an efficient and general strategy based on the use of dendritic initiators for the ring-opening polymerization of ε-caprolactone to yield functional core star polymers. The isolation of the core functionalities has been studied using fluorescence quenching and fluorescence resonance energy transfer (FRET) techniques as well as solvatochromic probes. With increasing chain length as well as solvent polarity, enhanced site isolation of the core has been observed. These findings have been correlated to actual molecular dimensions independently measured by pulsed field gradient spin - echo (PGSE) NMR. The developed synthetic methodology offers a rapid route to efficient encapsulation of functional moieties and therefore has potential for the design of new materials.

Full text of DOI:10.1021/ja003304u

18
J. Am. Chem. Soc. 2001, 123, 18-25
Encapsulation of Functional Moieties within Branched Star Polymers:
Effect of Chain Length and Solvent on Site Isolation
Stefan Hecht, Nikolay Vladimirov, and Jean M. J. Fr e´ chet*
Contribution from the Department of Chemistry, UniVersity of California, Berkeley, California 94720-1460
ReceiVed September 7, 2000
Abstract: Porphyrin and pyrene photoactive cores have been encapsulated within an isolating polymeric shell
using an efficient and general strategy based on the use of dendritic initiators for the ring-opening polymerization
of ꢀ-caprolactone to yield functional core star polymers. The isolation of the core functionalities has been
studied using fluorescence quenching and fluorescence resonance energy transfer (FRET) techniques as well
as solvatochromic probes. With increasing chain length as well as solvent polarity, enhanced site isolation of
the core has been observed. These findings have been correlated to actual molecular dimensions independently
measured by pulsed field gradient spin-echo (PGSE) NMR. The developed synthetic methodology offers a
rapid route to efficient encapsulation of functional moieties and therefore has potential for the design of new
materials.
Introduction
successfully implemented, the synthesis of highly branched
structures having exotic core functionalities remains challenging.
In recent years, one of the major objectives in dendrimer
1
chemistry has been the encapsulation of active core function-
(6) (a) Dandliker, P. J.; Diederich, F.; Gross, M.; Knobler, C. B.; Louati,
A.; Sanford, E. M. Angew. Chem., Int. Ed. 1994, 33, 1739. (b) Dandliker,
P. J.; Diederich, F.; Gisselbrecht, J.-P.; Louati, A.; Gross, M. Angew. Chem.,
Int. Ed. 1995, 34, 2725. (c) Sadamoto, R.; Tomioka, N.; Aida, T. J. Am.
Chem. Soc. 1996, 118, 3978. (d) Dandliker, P. J.; Diederich, F.; Zingg, A.;
Gisselbrecht, J.-P.; Gross, M.; Louati, A.; Sanford, E. M. HelV. Chim. Acta
1997, 80, 1773. (e) Pollak, K. W.; Leon, J. W.; Fr e´ chet, J. M. J.; Maskus,
M.; Abru n˜ a, H. D. Chem. Mater. 1998, 10, 30. (f) Weyermann, P.;
Gisselbrecht, J.-P.; Boudon, C.; Diederich, F.; Gross, M. Angew. Chem.,
Int. Ed. 1999, 38, 3215. (g) Jiang, D.-L.; Aida, T. Chem. Commun. 1996,
2
alities within dendritic backbones. Numerous examples of such
structures have shown profound effects of the dendrimer shell
on the core properties, and impressively demonstrated the
importance of this site isolation concept. The use of dendritic
nonnatural building blocks to mimic enzyme functions will
ultimately provide a more detailed understanding of biological
processes, and facilitate the design of artificial nanoscale
molecular devices by creating specific microenvironments
around active units. Porphyrin core dendrimers have received
special attention due to their potential as hemeprotein mimics
3
4
1
523. (h) Collman, J. P.; Fu, L.; Zingg, A.; Diederich, F. Chem. Commun.
1997, 193. (i) Bhyrappa, P.; Young, J. K.; Moore, J. S.; Suslick, K. S. J.
5
6
Am. Chem. Soc. 1996, 118, 5708. (j) Bhyrappa, P.; Young, J. K.; Moore,
J. S.; Suslick, K. S. J. Mol. Catal. A 1996, 113, 109. For additional work
on porphyrin core dendrimers see: (k) Jin, R.-H.; Aida, T.; Inoue S. J.
Chem. Soc., Chem. Commun. 1993, 1260. (l) Tomoyose, Y.; Jiang, D.-L.;
Jin, R.-H.; Aida, T.; Yamashita, T.; Horie, K.; Yashima, E.; Okamoto, Y.
Macromolecules 1996, 29, 5236. (m) Pollak, K. W.; Sanford, E. M.; Fr e´ chet,
J. M. J. J. Mater. Chem. 1998, 8, 519. (n) Jiang, D.-L.; Aida, T. J. Am.
Chem. Soc. 1998, 120, 10895. (o) Bhyrappa, P.; Vaijayanthimala, G.;
Suslick, K. S. J. Am. Chem. Soc. 1999, 121, 262. (p) Kimura, M.; Shiba,
T.; Muto, T.; Hanabusa, K.; Shirai, H. Macromolecules 1999, 32, 8237.
(q) Vinogradov, S. A.; Lo, L.-W.; Wilson, D. F. Chem. Eur. J. 1999, 5,
6a-f
6g,h
for electron transport,
dioxygen binding,
and oxidation
catalysis.6i,j
The preparation of structurally perfect dendrimers traditionally
suffers from its time-consuming nature due to the required
repetitive coupling and activation steps, and the necessity for
7
extensive purification. Although accelerated synthetic protocols
and the construction of less perfect analogues8 have been
1338. (r) Vinogradov, S. A.; Wilson, D. F. Chem. Eur. J. 2000, 6, 2456.
(
1) (a) Newkome, G. R.; Moorefield, C. N.; V o¨ gtle, F. Dendritic
(s) Matos, M. S.; Hofkens, J.; Verheijen, W.; De Schryver, F. C.; Hecht,
S.; Pollak, K. W.; Fr e´ chet, J. M. J.; Forier, B.; Dehaen, W. Macromolecules
2000, 33, 2967.
Molecules: Concepts, Synthesis, PerspectiVes; VCH: Weinheim, 1996. (b)
Bosman, A. W.; Jansen, H. M.; Meijer, E. W. Chem. ReV. 1999, 99, 1665.
(
c) Fischer, M.; V o¨ gtle, F. Angew. Chem., Int. Ed. 1999, 38, 885. (d)
(7) Accelerated growth strategies have been reported. For double-stage
convergent growth approaches, see: (a) Wooley, K. L.; Hawker, C. J.;
Fr e´ chet, J. M. J. J. Am. Chem. Soc. 1991, 113, 4252. (b) Ihre, H.; Hult, A.;
Fr e´ chet, J. M. J.; Gitsov, I. Macromolecules 1998, 31, 4061. For a double
exponential growth approach, consult: (c) Kawaguchi, T.; Walker, K. L.;
Wilkins, C. L.; Moore, J. S. J. Am. Chem. Soc. 1995, 117, 2159. The use
of “hypermonomer” building blocks is described in: (d) Wooley, K. L.;
Hawker, C. J.; Fr e´ chet, J. M. J. Angew. Chem., Int. Ed. 1994, 33, 82. (e)
L’abbe, G.; Forier, B.; Dehaen, W. J. Chem. Soc., Chem. Commun. 1996,
1262. Orthogonal monomers have been used by: (f) Spindler, R.; Fr e´ chet,
J. M. J. J. Chem. Soc., Perkin Trans. 1 1993, 913. (g) Zeng, F.; Zimmerman,
S. C. J. Am. Chem. Soc. 1996, 118, 5326. (h) Freeman, A. W.; Fr e´ chet, J.
M. J. Org. Lett. 1999, 1, 685.
Matthews, O. A.; Shipway, A. N.; Stoddart, J. F. Prog. Polym. Sci. 1998,
3, 1. (e) Fr e´ chet, J. M. J.; Hawker, C. J. In ComprehensiVe Polymer
Science, 2nd Suppl.; Aggarwal, S. L., Russo, S., Eds.; Pergamon Press:
Oxford, 1996; p 140. (f) Fr e´ chet, J. M. J. Science, 1994, 263, 1710. (g)
Tomalia, D. A.; Naylor, A. M.; Goddard, W. A., III Angew. Chem., Int.
Ed. 1990, 29, 138.
2
(
2) Functional dendrimers have been comprehensively reviewed by: (a)
Chow, H.-F.; Mong, T. K.-K.; Nongrum, M. F.; Wan, C.-W. Tetrahedron
998, 54, 8543. Other reviews include: (b) Newkome, G. R..; He, E..;
1
Moorefield, C. N. Chem. ReV. 1999, 99, 1689. (c) Archut, A.; V o¨ gtle, F.
Chem. Soc. ReV. 1999, 27, 233.
(3) Hecht, S.; Fr e´ chet, J. M. J. Angew. Chem., Int. Ed. 2001, 40, 74.
(
4) Dendrimers as biological mimics have been reviewed in: Smith, D.
(8) Recently, there have been major advances in achieving dendrimer-
like properties using hyperbranched polymers. Although such materials are
easily accessible via one-pot procedures, the covalent encapsulation of a
single entity is not currently practical using this approach. Comprehensive
reviews on hyperbranched polymers include: (a) Kim, Y. H. J. Polym. Sci.,
Part A: Polym. Chem. 1998, 36, 1685. (b) Voit, B. I. Acta Polym. 1995,
46, 87. (c) Reference 1e.
K.; Diederich, F. Chem. Eur. J. 1998, 4, 1353.
5) (a) Wang, P.-W.; Liu, Y.-J.; Devadoss, C.; Bharathi, P.; Moore, J.
S. AdV. Mater. 1996, 8, 237. (b) Kawa, M.; Fr e´ chet, J. M. J. Chem. Mater.
998, 10, 286. (c) Halim, M.; Pillow, J. N. G.; Samuel, I. D. W.; Burn, P.
(
1
L. AdV. Mater. 1999, 11, 371. (d) Freeman, A. W.; Koene, S. C.; Malenfant,
P. R. L.; Thompson, M. E.; Fr e´ chet, J. M. J. J. Am. Chem. Soc. In press.
1
0.1021/ja003304u CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/13/2000

Encapsulation of Functional Moieties
J. Am. Chem. Soc., Vol. 123, No. 1, 2001 19
Scheme 1
Table 1. Molecular Weight Characteristics of Star Polymers
DP(NMR)
Mh n(NMR)
W
Mh / Mh n(GPC)
1
1
1
1
2
2
2
2
3
3
3
3
7a
7b
7
7
8
8
a
b
c
d
a
b
c
d
a
b
c
26.6
33.3
44.6
54.2
24.9
31.9
41.0
50.4
24.9
31.9
41.0
50.4
33.7
56.8
84.5
116.3
31.5
53.6
82.4
107.3
25 400
31 600
41 800
50 600
47 200
59 900
76 600
93 700
50 000
62 700
79 400
96 500
8 000
13 300
19 600
26 900
14 700
24 800
38 000
49 300
1.14
1.13
1.17
1.19
1.18
1.14
1.12
1.10
1.18
1.14
1.12
1.10
1.35
1.37
1.32
1.26
1.21
1.17
1.16
1.18
Recently, we overcame these synthetic obstacles by efficiently
encapsulating porphyrins in a starlike polymer shell. This
9
strategy was based on the “living” ring-opening polymerization
of ꢀ-caprolactone using multifunctional, highly branched por-
phyrin initiators. This very practical approach benefits from the
rapid synthesis, the ease of purification, and the flexibility of
post-modification of both core and chain ends.
d
c
d
a
b
Here we report on the generality and modularity of our
synthetic route to encapsulate active core functionalities that
enable us to use a variety of spectroscopic techniques to evaluate
the degree of site-isolation of the core with regard to chain length
as well as solvent. In particular, we utilize fluorescence
resonance energy transfer (FRET) from terminal donor chro-
mophores to a core acceptor dye as a suitable tool to study the
conformational dynamics in such functional star polymer
architectures. The chain length and solvent dependence of core
encapsulation is supported by pulsed field gradient spin-echo
8c
8d
ing many initiating sites. Therefore, first generation dendrimers
having 8 and 16 hydroxyl functionalities, respectively, were used
as initiators to grow 8-arm and 16-arm stars. Metalation of the
obtained polymers gave two series of zinc porphyrin star
polymers 1a-d and 2a-d that were used in fluorescence
quenching studies employing an external probe (Scheme 2). In
contrast, esterification of the polymer chain ends using cou-
marin-3-carboxylic acid chloride provided star polymers 3a-d
having an internal probe since the coumarin chromophores are
capable of efficient FRET to the porphyrin core (Scheme 3).
Dendrimer 3e and coumarin-terminated star 4 having no
porphyrin acceptor core were also prepared as model compounds
for comparison purposes.
To complement the spectroscopic measurements on the
porphyrin core star polymers as well as to demonstrate the
generality of this encapsulation strategy, pyrene core star
polymers 7a-d and 8a-d having either 2 or 4 arms were
prepared from their respective initiators 5 and 6 (Scheme 4).
The pyrene core is ideally suited for use as a solvatochromic
probe since its fluorescence signal is highly sensitive toward
the local environment as well as aggregation (excimer forma-
tion).
1
4
(PGSE) NMR experiments, which allow a direct correlation with
molecular size.
Results and Discussion
Synthesis. Rapid encapsulation of a functional moiety
involves the preparation of a low-generation hydroxyl-terminated
7
b,10
aliphatic polyester dendritic initiator
for the ring-opening
polymerization of ꢀ-caprolactone11 to afford the desired func-
tional star polymers (Scheme 1). All polymerizations were
carried out in the bulk employing catalytic amounts of tin(II)
2
-ethylhexanoate, as described in detail by Trollsås and Hed-
1
2
rick. The desired polymers were obtained in essentially
quantitative yields and narrow polydispersities after a single
precipitation into methanol (Table 1). Molecular weights (MW)
were easily controlled by adjusting the monomer-to-initiator
ratio, in accord with the “living” nature of the polymerization.
Increasing degree of polymerization (DP)13 generally required
longer polymerization times. Due to the limited solubility of
the initiators in the monomer, polymers with DP > 25 were
generally prepared most effectively.
Fluorescence Quenching Experiments. In our initial com-
munication, we used the interaction of a small molecule
9
fluorescence quencher such as methyl viologen with the interior
zinc porphyrin to evaluate the core accessibility in stars 1a-d
and 2a-d (Scheme 2). The products of the quenching rate
constant kq and the excited-state lifetime τ were determined as
a function of chain length, i.e., degree of polymerization (DP),
using Stern-Volmer analysis (Figure 1). A direct evaluation
of kq to measure core accessibility is possible since the
absorption and fluorescence spectra as well as the fluorescence
quantum yields (in the absence of the quencher) remain
We focused on porphyrin and pyrene cores due to their unique
spectroscopic characteristics that allowed us to probe their
microenvironment, and therefore to evaluate the degree of site
isolation of the core. The use of tetrakis(4-hydroxyphenyl)-
porphyrin and tetrakis(3,5-dihydroxyphenyl)porphyrin having
multiple reactive groups offers the advantage of rapidly generat-
1
5
(
9) Hecht, S.; Ihre, H.; Fr e´ chet, J. M. J. J. Am. Chem. Soc. 1999, 121,
16
essentially constant, indicating no significant change of τ.
A strong shielding of the core moiety in the star polymers
compared to zinc tetraphenylporphyrin (ZnTPP) as the reference
9
6
239.
(10) (a) Ihre, H.; Hult, A.; S o¨ derlind, E. J. Am. Chem. Soc. 1996, 118,
388.
(
11) For a recent review, consult: L o¨ fgren, A.; Albertsson, A.-C.; Dubois,
P.; Jer oˆ me, R. J. Macromol. Sci., ReV. Macromol. Chem. Phys. 1995, C35,
79.
12) (a) Trollsås, M.; Hedrick, J. L.; Mecerreyes, D.; Dubois, P.; Jer oˆ me,
(14) In all described polymer series, the index a-d indicates increasing
DP. See Table 1 for further details.
3
(
(15) Consult standard photochemistry textbooks: (a) Turro, N. J. Modern
Molecular Photochemistry; University Science Books: Sausalito, 1991. (b)
Gilbert, A.; Baggott, J. Essentials of Molecular Photochemistry; CRC
Press: Boca Raton, 1991. (c) Klessinger, M.; Michl, J. Excited States and
Photochemistry of Organic Molecules; VCH: Weinheim, 1995.
(16) Turro, N. J.; Barton, J. K.; Tomalia, D. A. Acc. Chem. Res. 1991,
24, 332.
R.; Ihre, H.; Hult, A. Macromolecules 1997, 300, 8508. (b) Trollsås, M.;
Hedrick, J. L.; Mecerreyes, D.; Dubois, P.; Jer oˆ me, R.; Ihre, H.; Hult, A.
Macromolecules 1998, 31, 2756. (c) Trollsås, M.; Hedrick, J. L. J. Am.
Chem. Soc. 1998, 120, 4644 and references therein.
(13) Values given in Table 1 refer to the average DP of individual chains,
as indicated in Schemes 1-4.

20
J. Am. Chem. Soc., Vol. 123, No. 1, 2001
Hecht et al.
Scheme 2
Scheme 3
Scheme 4
was observed, thereby demonstrating inhibited penetration of
the small molecule quencher through the polymeric backbone.
The magnitude of site isolation is significantly higher than that
in the dendritic counterparts employed in similar quenching
studies.6 The degree of shielding is clearly dependent on the
chain length: with increasing DP the accessibility of the core
decreases. Extrapolation suggests a rather steep decline in core
accessibility in the early stages where DP is less than 25.
Surprisingly, the quenching efficiency is comparable for 8-arm
and 16-arm stars having similar DPs, suggesting that the chain
length, rather than the number of arms, is crucial for isolation
of the core unit. This finding is in agreement with existing
experimental studies17 as well as theoretical models describing
the shape of star polymers using an asphericity factor δ that
has a value of 1 for cylindrical molecules and vanishes for
molecules with spherical symmetry.18 In good solvents, only
minor differences are predicted between the 8-arm star (δ )
e,k
expected to be diminished even further in poor solvents,20 such
as the acetonitrile employed in this experiment.
1
9
0
.14) and 16-arm star (δ ) 0.07), and this difference is
However, the fluorescence quenching experiments were
somewhat limited due to the low solubility of the external probe
(
17) (a) Willner, L.; Jucknischeke, O.; Richter, D.; Farago, B.; Fetters,
L. J.; Huang, J. S. Europhys. Lett. 1992, 19, 297. (b) Willner, L.;
Jucknischeke, O.; Richter, D.; Roovers, J.; Zhou, L.-L.; Topowski, P. M.;
Fetters, L. J.; Huang, J. S.; Lin, M. Y.; Hadjichristidis Macromolecules
(19) Calculated by using δ ) (150f - 140f )/(135 - 120f - 4f-2),
where f is the number of arms. For further details, consult: Wei, G.;
Eichinger, B. E. J. Chem. Phys. 1990, 93, 1430.
-
1
-2
-1
1
994, 27, 3821.
18) (a) Rudnick, J.; Gaspari, G. J. Phys. A 1986, 19, L191. (b)
Aronowitz, J. A.; Nelson, D. R. J. Phys. 1986, 47, 1445.
(
(20) Sikorski, A.; Romiszowski, P. J. Chem. Phys. 1998, 109, 6169.

Encapsulation of Functional Moieties
J. Am. Chem. Soc., Vol. 123, No. 1, 2001 21
Figure 2. Photophysical characteristics of compounds 3 in chloroform.
Increasing optical density (concentrations: 0.2-4 µM) of the sample
(top) leads to enhanced self-absorption (trivial energy transfer) in the
system as indicated by the shape of the donor emission spectrum
(bottom).
Figure 1. Fluorescence quenching for two series of zinc porphyrin
star polymers having 8 arms (1a-d) and 16 arms (2a-d). The k τ
q
values are derived from Stern-Volmer analysis in acetonitrile employ-
ing methyl viologen as the quencer.
strating that FRET was facile but not quantitative in this system.
In a separate control experiment, excitation of a porphyrin core
star polymer having no coumarin donor chromophores attached
led to no observable emission due to the negligible absorption
at this particular excitation wavelength. It is important to note
that at concentrations above 0.2 mM, the extremely high
in the less polar solvents that are good solvents for the
investigated polymers. Rather than employing different neutral
quenchers, we sought to take advantage of the unique design
of the system with its multiple reactive chain ends to attach an
internal probe enabling the study of site isolation as a function
of solvent.
6k,q
-1
-1
extinction coefficient of the Soret band (ꢀ ) 519 000 M cm )
promotes self-absorption, i.e., trivial radiative energy transfer.
Since the coumarin emission spectrum is very indicative of this
effect (Figure 2), it could be shown that the portion of the
radiative energy transfer increased linearly with increasing
concentration and therefore optical density, assuming constant
efficiency of the nonradiative transfer, i.e., FRET. In all
experiments concentrations of approximately 0.1 mM were
employed to avoid this self-absorption process, and to enable
the use of both good and poor solvents in the experiments. On
the basis of the very low concentrations employed, aggregation
phenomena are unlikely to occur. Further experimental evidence
for this assumption arises from the excellent agreement of the
FRET Studies. FRET involves the nonradiative transfer of
singlet excitation energy from a donor chromophore to an
acceptor chromophore. According to F o¨ rster’s mechanism,
the magnitude of the interaction, i.e., the FRET efficiency
2
1a
(
ΦFRET), is inversely proportional to the sixth power of the
13,21
donor-acceptor separation.
Mainly due to this sensitive
distance dependence, FRET has proven to be an exceptional
tool for the study of the conformation and motion of biological
2
2
macromolecules. To monitor interactions over the range of
several nanometers, a high intrinsic probability of FRET and
therefore maximized spectral overlap between the emission of
1
3,21
the donor and the absorption of the acceptor is required.
In the present study, the porphyrin core served as acceptor,
whereas multiple donors were introduced by modification of
the chain ends (Scheme 3). In our system, coumarin-3-
carboxylates were chosen as the donor chromophores since their
emission centered around 420 nm strongly overlaps with the
Soret absorption band (λmax ) 420 nm) of the porphyrin acceptor
1
MW determined independently by light scattering and H NMR
as well as the constant gel permeation chromatography (GPC)
elution volume of polymers 3a-d within a wide concentration
-8
-4
range (10 -10 M in THF and CHCl3). The exclusion of
aggregation is essential since only intramolecular FRET provides
valuable information about the intrinsic polymer dynamics.
The dependence of FRET on polymer chain length was
investigated (Figure 3). As the chain length increases and the
donor chromophores are placed at further average distances from
the acceptor core, the donor emission intensity increases as the
result of the reduced probability of FRET in the system. In one
extreme case, quantitative FRET is observed in 3e due to
extremely short average donor-acceptor distances. In contrast,
no FRET occurs in 4 as a result of the absence of an acceptor
chromophore.
2
3
core (Figure 2). In addition, coumarin-3-carboxylic acid is a
commercially available, rather inexpensive dye that can be
conveniently linked to the hydroxyl-chain ends via esterifica-
9
tion. When the coumarin donors in polymers 3a-d were
excited selectively (λexc ) 350 nm), emission from both the
coumarins and the porphyrin acceptor was observed, demon-
(21) (a) F o¨ rster, T. Fluoreszenz Organischer Verbindungen; Vandenhoech
and Ruprech: G o¨ ttingen, 1951. (b) Van der Meer, W. B.; Coker, G., III;
Chen, S.-Y. Resonance Energy Tranfer, Theory and Data; VCH: Weinheim,
1
994.
22) Representative recent reviews are given in the following: (a) Weiss,
The values of ΦFRET were determined by the method of
(
2
4
quenched donor emission, which compares the coumarin
S. Science 1999, 283, 1676. (b) Sz o¨ ll o¨ si, J.; Damjanovich, S.; M a´ tyus, L.
Cytometry 1998, 34, 159.
emission in the presence of the acceptor (3a-e) to the emission
(23) To our knowledge, this is the first time the emission originating
25
in the absence of an acceptor, as in model compound 4. The
from a coumarin dye is quenched by the porphyrin Soret absorption. One
example of energy transfer from a coumarin donor (7-diethylaminocoumarin-
results shown in Figure 4 obtained with different solvents reflect
the chain length dependence, as described above, and comple-
ment our earlier findings using fluorescence quenching (vide
supra). Interestingly, a pronounced solvent effect is observed.
In good solvents for the poly(ꢀ-caprolactone) stars, such as
CHCl3 or toluene, a steep decline of ΦFRET is observed with
3
-carboxylate) to a metalloporphyrin acceptor (palladium(II) tetraarylpor-
phyrin) has been reported by: Kaschak, D. M.; Lean, J. T.; Waraksa, C.
C.; Saupe, G. B.; Usami, H.; Mallouk, T. E. J. Am. Chem. Soc. 1999, 121,
3
435. However, due to the red-shifted emission of the employed coumarin,
sensitization of the Q-band and not the Soret band of the metalloporphyrin
most likely occurred.

22
J. Am. Chem. Soc., Vol. 123, No. 1, 2001
Hecht et al.
Figure 3. Corrected emission of the terminal coumarin donor chromo-
phores in compounds 3a-e and model compound 4 in chloroform as
a function of chain length (DP).
Figure 5. Illustration of the solvation induced encapsulation of the
core functionality. While the polymer adopts a more extended
conformation in a good solvent (top), it collapses in a poor solvent
Figure 4. Resonance energy transfer from the terminal coumarin donor
chromophores to the free-base porphyrin core acceptor in compounds
(bottom) leading to a reduced aVerage donor-acceptor distance and
therefore enhanced energy transfer efficiency. The terminal coumarin
chromophores are shown in cyan, the poly(caprolactone) backbone in
green, the dendritic branching unit in gray, and the porphyrin core in
magenta.
3
a-e employing different solvents. The energy transfer efficiencies
were determined from the amount of quenched donor emission as
compared to model compound 4.
increasing DP, whereas in poor solvents, such as acetonitrile
or DMSO, ΦFRET levels off at high DP. Although other
coumarins such as 7-methoxycoumarin display solvent-depend-
ent emission characteristics,26 the observed effect is clearly due
to conformational changes since the chain length is the only
variable for measurements in each solvent.
density, the same changes in volume will be associated with
smaller and smaller changes in radius. This behavior should be
most pronounced in the solvent collapsed and therefore more
densely packed structure since an increase in DP will have only
a minor effect on the average donor-acceptor separation.
Recent work on dendritic systems with nearly quantitative
ΦFRET (ΦFRET ) 0.93-0.98) related interchromophoric distances
estimated from molecular modeling to the rate of FRET
We suggest that the collapse of star polymers 3 in poor
solvents leads to a reduced aVerage donor-acceptor distance
and therefore increased ΦFRET as compared to the extended
conformations present in good solvents (Figure 5).27 In a crude
approximation, the different slopes in Figure 4 can be rational-
ized by considering a sphere (δ ) 0.07, vide supra) for which
the volume is directly related to the third power of the radius,
2
4a
measured by time-resolved techniques. In contrast, the star
polymer counterparts studied here cover a much wider range
of donor-acceptor distance and therefore ΦFRET (ΦFRET )
0
.46-0.98). In the special case when ΦFRET is 0.5, the
corresponding distance known as the F o¨ rster radius R0 can be
3
i.e., V ∝ r . By progressing outward and assuming a similar
calculated from the spectral overlap of the participating chro-
2
1
(24) We recently found this method to be well suited for studying FRET
mophores. Since theory validates the assumption that, in the
absence of any favorable interactions between chain ends and
in dendritic systems: (a) Adronov, A.; Gilat, S.; Fr e´ chet, J. M. J.; Ohta,
K.; Neuwahl, F. V. R.; Fleming, G. R. J. Am. Chem. Soc. 2000, 122, 1175.
1f
inner building blocks, the terminal groups are located near the
(
1
b) Gilat, S. L.; Adronov, A.; Fr e´ chet, J. M. J. Angew. Chem., Int. Ed.
999, 38, 1422. (c) Adronov, A.; Malenfant, P. R. L.; Fr e´ chet, J. M. J.
2
8
periphery, this allows a rough correlation of molecular size
Chem. Mater. 2000, 12, 1463. For a detailed discussion, see also: Devadoss,
C.; Bharati, P.; Moore, J. S. J. Am. Chem. Soc. 1996, 118, 9635. Quantifying
the enhanced acceptor emission was found to be associated with a much
greater error arising from the almost negligible emission of the model
compound having no donors chromophores.
with measured ΦFRET. For this particular system, R0 is calculated
(27) A similar result has been observed in a dendritic system having a
water-soluble carboxylate-functionalized periphery and a palladium por-
phyrin core. Different quenching constants of molecular oxygen were found
with varying pH of the medium suggesting conformational changes of the
dendrimer backbone. See refs 6q,r for further details.
(28) (a) Grest, G. S.; Kremer, K.; Milner, S. T.; Witten, T. A.
Macromolecules 1989, 22, 1904. (b) Li, H.; Witten, T. A. Macromolecules
1994, 27, 449. (c) Grest, G. S. Macromolecules 1994, 27, 3493.
(
25) Almost identical results were obtained using different model
compounds such as methyl coumarin-3-carboxylate.
26) (a) Seixas de Melo, J. S.; Becker, R. S.; Macanita, A. L. J. Phys.
(
Chem. 1994, 98, 6054. (b) Schade, B.; Hagen, V.; Schmidt, R.; Herbrich,
R.; Krause, E.; Eckardt, T.; Bendig, J. J. Org. Chem. 1999, 64, 9109.

Encapsulation of Functional Moieties
J. Am. Chem. Soc., Vol. 123, No. 1, 2001 23
2
2 2
Figure 7. Plot of polymer peak area versus g γ δ (∆ - δ/3) and linear
fit for star polymers 3 in CDCl (open symbol, dotted lines) and CD CN
(solid symbols, solid lines) at 298 ( 1 Κ.
3
3
1
Figure 6. Stacked plot H NMR spectra of star polymer 3b in CDCl
3
(top) and CD
3
CN (bottom) acquired with increasing gradient strength
at 298 ( 1 Κ.
an ideal opportunity to study the effect of chain length as well
as solvent on the actual size of the star polymers.
2
to be 8.6 nm using eq 1, where κ is the orientation factor, J is
When a pulse sequence producing a spin-echo is repeated
under conditions of increasing gradient strength, the observed
signal intensities will decrease depending on the diffusion rate
of the corresponding spin-labeled species. In this process the
gradient encodes the spatial location of the observed molecule,
similar to magnetic resonance imaging (MRI). Employing such
experimental setup, star polymers 3a-d were measured in
CDCl3 and CD3CN, respectively. Figure 6 shows two repre-
sentative 2D NMR spectra, displaying chemical shift in one
dimension and gradient strength, which is correlated to molec-
ular diffusion and therefore size, in the other dimension.
Depending on the solvent, the same polymer sample exhibits a
distinctly different decay behavior. As a result of a larger
diffusion coefficient, which in turn has its origin in a smaller
size of the diffusing object, decay is faster in a bad solvent
6
2
0
.5291κ J
R0 )
(1)
4
x
n N
A
the overlap integral of the fluorescence intensity of the donor
and the molar extinction coefficient of the acceptor normalized
by the frequency expressed in wavenumbers, n is the index of
15b
refraction of the solvent, and NA is Avogadro’s constant. This
suggests a diameter of 3d in CHCl3 or toluene on the order of
8-20 nm. This value provides a reasonable explanation of why
our attempts to directly obtain molecular size data by light
scattering were not successful, since reliable data can usually
only be obtained if the diameter of the investigated molecule is
at least 1/20 of the wavelength of the laser (488 nm) used, i.e.,
1
(CD3CN) than in a good solvent (CDCl3).
2
9
at least 25 nm.
Quantitative analysis was accomplished by linear regression
PGSE NMR Experiments. In addition to its conventional
role of providing detailed information about the local electronic
environment of a molecule, NMR also represents a powerful
tool for the study of overall molecular properties. Since sizes
and shapes of molecular objects are related to friction factors
for reorientational and translational motion, a correlation with
nuclear relaxation rates is possible. Translational motion can
effectively be studied by NMR using PGSE methods, relating
signal intensities to diffusion rates and therefore enabling the
resolution of molecular dimensions in solution. With regard to
nonbiological macromolecules, PGSE NMR has mainly been
applied to the study of surfactant systems and polymer
32
of the Stejskal-Tanner equation (eq 2), where g is the gradient
2
2 2
g γ δ (∆-δ/3)D
I ) I0e
(2)
strength, γ is the gyromagnetic ratio, δ is the duration of gradient
pulse, ∆ is the delay between consecutive gradient pulses, and
3
3
D is the diffusion coefficient (Figure 7).
30
From the obtained diffusion coefficients, the hydrodynamic
radii RH were estimated using the Stokes-Einstein equation (eq
3
), where kB is the Boltzmann constant, T is the absolute
k T
3
0b
B
mixtures. Surprisingly in view of its great potential, the use
of this method to study branched polymer architectures has so
far been rather limited.31 In our case, PGSE NMR represents
RH )
(3)
6
πηD
temperature, and η is the viscosity of the solvent. The results
clearly reflect the dependence of size and therefore site-isolation
on chain length and solvent (Table 2). In both solvents the
increasing DP corresponds to an increase in RH giving rise to a
“thicker” polymer shell around the core. In comparison to the
(
29) (a) Light Scattering from Polymer Solutions; Huglin, M. B., Ed.;
Academic Press: New York, 1972. (b) Wyatt, P. J. Anal. Chim. Acta 1993,
72, 1 and references therein.
30) Representative reviews include: (a) Johnson, C. S., Jr. Prog. Nucl.
Magn. Reson. Spectrosc. 1999, 34, 203. (b) S o¨ derman, O.; Stilbs, P. Prog.
Nucl. Magn. Reson. Spectrosc. 1994, 26, 445. (c) Stilbs, P. Prog. Nucl.
Magn. Reson. Spectrosc. 1987, 19, 1. (d) Price, W. S. Concepts Magn.
Reson. 1998, 10, 197. (e) Price, W. S. Concepts Magn. Reson. 1997, 9,
99.
31) (a) Young, J. K.; Baker, G. R.; Newkome, G. R.; Morris, K. F.;
Johnson, C. S., Jr. Macromolecules 1994, 27, 3464. (b) Ihre, H.; Hult, A.;
S o¨ derlind, E. J. Am. Chem. Soc. 1996, 118, 6388. (c) Valentini, M.; Pregosin,
P. S.; R u¨ egger, H. Organometallics 2000, 19, 2551.
2
(
(32) Stejskal, E. O.; Tanner, J. E. J. Chem. Phys. 1965, 42, 282.
(33) Although the investigated system is not monodisperse, satisfactory
data (R > 0.99) were obtained using a linear regression procedure.
Presumably, the low polydispersity (PDI ) 1.10-1.20) and the high
sphericity contribute to a narrow diffusion coefficient distribution. For details
about polydisperse samples consult: (a) Reference 30. (b) Chen, A.; Wu,
D.; Johnson, C. S., Jr. J. Am. Chem. Soc. 1995, 117, 7965.
2
(

24
J. Am. Chem. Soc., Vol. 123, No. 1, 2001
Hecht et al.
Table 2. Diffusion Coefficient D and Hydrodynamic Radii R
Star Polymers 3a-d in CDCl and CD CN
H
of
3
3
2
solvent
CDCl
polymer
D (m /s)
R
H
(nm)
-
11
11
11
11
10
10
10
10
3
3a
8.62 × 10
7.31 × 10
5.86 × 10
4.40 × 10
3.20 × 10
2.36 × 10
1.85 × 10
1.39 × 10
4.7
-
-
-
-
-
-
-
3
3
3
b
c
d
5.5
6.9
9.2
2.0
2.7
3.4
4.6
3
CD CN
3a
3
3
3
b
c
d
good solvent, the poor solvent leads to largely reduced sizes
indicating a collapse of the poly(caprolactone) backbone around
the core unit leading to a more densely packed shell. These
results are in good agreement with the FRET studies described
above. For instance, the RH and ΦFRET values of 3a in CDCl3
and 3d in CD3CN are comparable. Furthermore, the size of 3d
in CDCl3 as measured by PGSE NMR agrees well with the
calculated F o¨ rster radius (vide supra) validating the assumption
Figure 8. Solvatochromic probing using the I
ratio in two series of pyrene core star polymers having 2 arms (7a-d)
and 4 arms (8a-d) in toluene and DMSO.
1
:I
3
fluorescnece intensity
approaches a polarity reflecting that of the poly(ꢀ-caprolactone)
shell, which is more polar than toluene but less polar than
DMSO. As observed with the FRET measurements (Figure 4),
the response to a change in chain length is more pronounced in
the good solvent, providing further evidence in support of
solvation-induced site encapsulation. Furthermore, the change
in the number of arms, going from 7a-d to 8a-d (2 vs 4 arms),
contributes more to site isolation than the change from 1a-d
to 2a-d (8 vs 16 arms), especially when the polymer is extended
that the terminal ends are close to the periphery in the
investigated polymer architectures.34
Solvatochromic Probes. In an alternative approach to study
core isolation that does not involve porphyrin chromophores,
star polymers 7a-d and 8a-d with a pyrene core were
investigated as solvatochromic probes of their own core
environment. Pyrene itself has been used extensively to help
determine local environments since the vibrational fine structure
of its fluorescence spectrum, namely the ratio of the I1 peak
(
good solvent). This is not surprising since the shape differences
1
8
19
(
∆δ) between the 2-arm star (δ ) 0.53) and the 4-arm star
(
0-0 band) and the I3 peak (0-2 band), is very sensitive to
3
5
(δ ) 0.27) are more pronounced than those for the 8-arm star
solvent polarity due to vibronic coupling (Ham effect). With
increasing polarity of the probe’s microenvironment, the I1:I3
ratio increases since the I1 band is markedly enhanced whereas
the I3 band remains unaffected. In addition, pyrene can be used
to study aggregation phenomena due to its characteristic excimer
emission.36
(δ ) 0.14) and 16-arm star (δ ) 0.07), recalling that δ ) 0 for
a sphere (vide supra).
Conclusions
As expected, when substitution is introduced r to attach the
initiating functional groups, the pyrene moiety becomes less
sensitive to changes in solvent polarity. The decreased probe
sensitivity is not a result of the lowered symmetry of the system,
but instead it derives from the introduction of the nearby
carbonyl functionality, which presumably interferes with the
vibronic coupling mechanism.37 This substitution also prevents
excimer formation, since even at relatively high concentrations
of approximately 0.5 mM no excimer emission could be
detected. However, when solvents such as toluene and DMSO
were employed, representing extremes in polarity, a clear trend
could be deduced from the I1:I3 ratios (Figure 8).
The construction of star polymers via ring-opening poly-
merization of ꢀ-caprolactone using functional dendritic initiators
has been demonstrated to be an efficient route for the encap-
sulation of active core functionalities. While this approach lacks
3
the precision of dendrimer encapsulation, it is synthetically
simpler providing access to a larger range of sizes and allowing
the products to be easily purified by precipitation. Moreover,
this route seems to be fairly general, since in principle any
functional core carrying phenol or alcohol functionalities (or
conceivably amines) could be encapsulated using similar
chemistry. The possibility of post-modification affords access
to a wide variety of materials, which may be used to probe the
effect of variables such as chain length, number of arms, or
solvent on core isolation. Using three different types of assays,
namely fluorescence quenching, FRET, and solvatochromic
probes, it has been shown that core encapsulation is mainly
dependent on DP with only minor contributions arising from
the number of arms. In addition, poor solvents seem to further
increase the degree of site isolation due to a structural collapse
of the polymer backbone giving rise to a more dense shielding
around the core unit. This solvation-induced encapsulation effect
and the chain length dependence are supported by PGSE NMR
experiments that allow for direct determination of the molecular
sizes of the polymers in different solvents. FRET as well as
PGSE NMR have been shown to be particularly useful
techniques to study the dynamics of star polymers in solution.
These studies concerned with the general principles of site-
isolation and its practically useful realization will ultimately lead
to the design of future functional materials.
In agreement with the experiments described above, a
significant chain length dependence on core isolation is found.
With increasing DP, the local environment around the core
(34) The theoretical maximum radius assuming a fully extended con-
formation was estimated using molecular modeling that predicts a length
of ∼8.6 Å for each monomer repeat unit. Therefore, polymer 3d (DP )
5
0) would have a theoretical maximum radius of approximately 44 nm,
1
i.e., 50 × 8.6 Å + /2 core size (∼10 Å). The measured radius of 9.2 nm
Table 1) is much smaller, presumably due to substantial participation of
random coil conformations.
35) (a) Kalyanasundaram, K. Photochemistry in Microheterogeneous
(
(
Systems; Academic Press: London, 1987; Chapter 2, p 36. (b) Kalyana-
sundaram, K. In Photochemistry in Organized and Constraint Media;
Ramamurthy, V., Ed.; VCH: Weinheim, 1991; p 39.
(
36) (a) Selinger, B. K.; Watkins, A. R. Chem. Phys. Lett. 1978, 56, 99.
b) Infelta, P. P.; Gr a¨ tzel, M. J. Chem. Phys. 1979, 70, 179. (c) Atik, S. S.;
Nam, M.; Singer, L. A. Chem. Phys. Lett. 1979, 67, 75.
37) This can be shown by a comparison of the changes in I1:I3 ratio
∆I1:I3) when going from acetonitrile to CHCl3: for pyrene ∆I1:I3 ≈ 0.5, for
initiators 5 and 6 ∆I1:I3 ≈ 0.1, whereas for 1-pyrenemethanol ∆I1:I3 ≈ 0.6.
(
(
(

Encapsulation of Functional Moieties
J. Am. Chem. Soc., Vol. 123, No. 1, 2001 25
Acknowledgment. Henrik Ihre is acknowledged for his help
in the early stages of the project as well as Seth Bush and Rudi
Nunlist for their invaluable help with the PGSE NMR experi-
ments. Financial support from the AFOSR-MURI program and
the National Science Foundation (NSF-DMR 9816166) is
acknowledged with thanks.
Supporting Information Available: Experimental details
and chemical characterization data (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA003304U