Tailoring Polyethers for Post-polymerization Functionalization by Cross Metathesis

Article abstract of DOI:10.1021/acs.orglett.8b00595

Olefin cross metathesis is reported for the first time to attach small molecules to a range of novel polyethers with a poly(ethylene glycol) backbone and pendent alkene groups, allowing for a loading of up to one compound per monomer unit. These polymers are tailored to prevent the occurrence of self metathesis (reaction of the polymer with itself) by varying the substitution on the pendent alkenes, thus steering their reactivity toward olefin cross metathesis. Efficient functionalization has been observed for a range of coupling partners as a proof of concept for the use of olefin metathesis to graft small and larger molecules to polyethers for drug delivery. This approach also paves the way for the use of olefin cross metathesis as an efficient method to functionalize a wide variety of polymers with pendent olefin groups.

Full text of DOI:10.1021/acs.orglett.8b00595

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Tailoring Polyethers for Post-polymerization Functionalization by
Cross Metathesis
Stephen D. Morrison, Rob M. J. Liskamp, and Joelle Prunet*
̈
WESTCHEM, School of Chemistry, University of Glasgow, Joseph Black Building, University Avenue, Glasgow G12 8QQ, U.K.
S
* Supporting Information
ABSTRACT: Olefin cross metathesis is reported for the first
time to attach small molecules to a range of novel polyethers
with a poly(ethylene glycol) backbone and pendent alkene
groups, allowing for a loading of up to one compound per
monomer unit. These polymers are tailored to prevent the
occurrence of self metathesis (reaction of the polymer with
itself) by varying the substitution on the pendent alkenes, thus
steering their reactivity toward olefin cross metathesis. Efficient functionalization has been observed for a range of coupling
partners as a proof of concept for the use of olefin metathesis to graft small and larger molecules to polyethers for drug delivery.
This approach also paves the way for the use of olefin cross metathesis as an efficient method to functionalize a wide variety of
polymers with pendent olefin groups.
iocompatible polymers such as poly(ethylene glycol)
(PEG) have been widely used for drug conjugation,1 but
metathesis was minimized with a high loading (8 equiv) of the
olefin cross partner.
B
they lack functional handles along the polymer backbone. This
limits the possible modifications they can undergo. The
synthesis of pertinent and useful polymers frequently requires
post polymerization modification in order to incorporate
architectures bearing functional groups not compatible with
the polymerization process.2 Side-chain functionalization of
polymers is commonly achieved by employing azide−alkyne
cycloaddition,3 terminal functional group modification,4 thiol−
ene addition,5 Michael-type addition,6 and amidation,7 among
many others.2 Olefin cross metathesis (CM) is a powerful
carbon−carbon bond-forming reaction8 performed under very
mild conditions with catalysts compatible with most heter-
oatom functional groups and, therefore, could be used for the
conjugation of polyethers possessing pendent olefin handles.
However, conjugation of polymers by olefin CM remains a
relatively unexplored area.9 Functionalization by CM was
pioneered by Coates et al., who showed that moderate
conversion could be achieved when reacting various alkene-
contining polyolefins with small olefins.10 Hoogenboom and
Meier and co-workers also reported that acrylate derivatives
could successfully be coupled to a poly(2-oxazoline) (POx)
with pendent olefins.11 This approach was successful, but the
occurrence of self metathesis (SM) was observed, which is the
process by which a pendent olefin of the polymer undergoes
CM with another pendent olefin, either intramolecularly or
intermolecularly. This undesired process was limited by using a
large excess (7−12 equiv) of the acrylate coupling partners.
Edgar et al. then described CM of cellulose esters with 20 equiv
of acrylate derivatives.12 Our group has performed successful
CM reactions of hindered polyesters with several olefinic
partners.13 The Shaver group further reported CM reactions
between the polymer of β-heptenolactone and an extensive
range of small olefinic partners.14 In the latter case, self
Polyethers have a flexible backbone and therefore should be
prone to SM, as is the case for the POx derivatives.11
Intermolecular SM is especially deleterious to the efficiency of
CM functionalization, as the resulting cross-linking leads to
highly increased dispersities in the grafted polymers but using a
large excess of an expensive coupling partner is not a
sustainable solution. Herein, we report tailoring the polyethers
according to the olefinic partners in order to favor CM over
SM, by modifying the substitution on the pendent alkene of the
polymer.
As CM with alkene-containing polyethers had not been
reported, preliminary studies were conducted on the known
poly(allyl glydidyl ether) p(AGE).15 Anionic ring-opening
polymerization of commercially available allyl glycidyl ether
(neat) with potassium benzoxide as initiator produced the
desired polymer (Mn = 7990 g/mol, Đ = 1.08). When this
polyether was submitted to metathesis with methyl acrylate in
the presence of Hoveyda−Grubbs second-generation catalyst
HG216 in refluxing dichloromethane, the desired CM reaction
took place (Scheme 1, x units), but some unreacted olefins
were recovered (y units), and a substantial amount of SM was
observed (z units). Under optimized conditions, the x/y/z ratio
was 85:0:15 for methyl acrylate 1 (R1 = COOMe, R2 = H) and
90:5:5 for Z-1,4-butenediol diacetate 2 (R1, R2 = CH2OAc),
which behaves as an allyl acetate surrogate.
1
These ratios were determined by H NMR spectroscopy,
which clearly shows the different olefinic protons for each unit
(Figure 1). Although the amount of SM is not very high,
especially in the case of CM with the dimer of allyl acetate, the
Received: February 19, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b00595
Org. Lett. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Organic Letters
Letter
Scheme 1. First Attempt of Olefin CM Reactions with
p(AGE)17
Figure 2. Novel polyethers with varying olefin handles.
and type 3 olefins do not dimerize, although they can
participate in CM reactions.20 Under the previously optimized
conditions, coupling of p(CGE) with methyl acrylate 1 (entry
1) saw a slight decrease in conversion in comparison to
p(AGE) (entry 2) with an overall increase in SM, indicating
that the added methyl group is not efficient in inhibiting
formation of the SM product. The reaction of p(PGE) with
methyl acrylate (entry 3) led to a significant increase in
successful CM (95%). Unsurprisingly, no reaction occurred
between p(MAGE) and methyl acrylate (entry 4) due to the
low reactivity of both olefin species.
Coupling of the polymer range with the dimer of allyl acetate
2 saw good conversion with p(AGE) (entry 5), p(CGE) (entry
6), and p(PGE) (entry 7) and a strong correlation between the
degree of SM and the increase in dispersity. In each case, a
substantial degree (up to 30%) of the original pendent olefins
remained unreacted, which is probably due to intramolecular
SM, resulting in conformations of the polymer chain that makes
some olefins less accessible. Most interestingly, p(MAGE)
underwent 23% successful CM with no evidence of SM (entry
8). For the first three polymers, the E/Z ratio of olefin isomers
was 10:1, and for p(MAGE), it was 3:1. These numbers are in
line with the selectivities observed for CM between small
molecules of the same type.20
Coupling more complex molecules, such as peptides with the
polymer range is of considerable interest, as these peptides
moieties could serve as targeting devices in the drug delivery
process. Studies were thus conducted on an amino acid
substrate. CM of the derivative of Boc-protected glycine 3 (see
Scheme 1 for structure) with p(AGE) (entry 9) and p(CGE)
(entry 10) yielded similar results with 30% and 40% successful
CM, respectively, along with a significant amount of SM. With
p(PGE) (entry 11), the same percentage of desired CM was
observed, but with 27% unreacted prenyl handles remained.
With p(MAGE) (entry 12), a conversion of 30% was achieved,
but once again no SM was detected for this polymer, as seen by
the maintained dispersity on the GPC trace.
The results observed for the CM reactions can be explained
by the different reactivity of the polymers. Both p(AGE) and
p(CGE) possess type 1 olefins and, therefore, show good
reactivity with any coupling partner. However, a certain degree
of SM is seen as a result of their reactivity. The ruthenium
catalyst reacts primarily with the pendent olefin to form the
metal−polymer complex 4,21 which can then undergo
successful CM (Scheme 2). Carbene 4 can also react more
slowly with another pendent olefin, yielding SM product 6. In
the case of a type 1 olefin coupling partner, the desired product
5 is also a type 1 olefin and can react with the catalyst to
regenerate the carbene 4, thus leading to the SM product 6 by a
secondary metathesis reaction. For these two polymers,
successful CM is the major outcome of metathesis, along
with 10−20% of SM. We had hoped that SM would be slower
in the case of p(CGE) because of the additional steric
Figure 1. 1H NMR spectrum (400 MHz, CDCl3) showing x, y, and z
units.
dramatic increase in dispersity (Đ > 2) indicates mainly
intermolecular SM and, hence, formation of branched
polymers. Surprisingly, increasing the amount of CM partner
up to 8 equiv did not improve the efficiency of the desired CM
process.
In order to limit the degree of SM, we designed novel
polyethers with additional steric hindrance around the pendent
olefin. Using potassium tert-butoxide as initiator,18 which is
easier to handle than potassium benzoxide, p(AGE), poly-
(crotyl glydidyl ether) p(CGE), poly(prenyl glydidyl ether)
p(PGE) and poly(methallyl glydidyl ether) p(MAGE) were
synthesized in high molar masses and low dispersity (Figure 2)
from the corresponding epoxide monomers.19 It is worth
noting that the grafted polymer units resulting from CM with
p(AGE), p(CGE) and p(PGE) will be the same, while CM
with p(MAGE) will lead to a trisubstituted olefin.
CM was then performed between these polymers and various
type 1 and type 2 coupling partners (Table 1). Grubbs’
classification of olefins according to their reactivity toward
metathesis with a specific catalyst states that type 1 olefins
undergo fast homodimerization, type 2 olefins dimerize slowly,
B
DOI: 10.1021/acs.orglett.8b00595
Org. Lett. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Organic Letters
Letter
a
Table 1. CM of Polymer Range with Diverse Coupling Partners
entry
polymer
Mn before CM (by GPC)
Đ before CM
coupling partner
yield (%)
x/y/z
Mn after CM (by GPC)
Đ after CM
b
1
p(AGE)
p(CGE)
p(PGE)
p(MAGE)
p(AGE)
p(CGE)
p(PGE)
p(MAGE)
p(AGE)
p(CGE)
p(PGE)
p(MAGE)
10000
9200
1.16
1.28
1.35
1.15
1.16
1.28
1.35
1.15
1.16
1.28
1.35
1.15
1
1
1
1
2
2
2
2
3
3
3
3
92
95
84
85:5:10
80:0:20
95:2:3
13900
2.11
2.21
2.00
b
2
13900
b
3
10000
15800
10000
9200
10300
4
0:100:0
87:4:9
c
5
90
83
95
95
94
82
95
97
12300
1.46
1.98
1.76
1.09
1.81
1.72
1.84
1.09
c
6
73:12:15
60:30:10
23:77:0
30:7:63
40:7:53
38:27:35
30:70:0
15600
c
7
10000
15800
10000
9200
12800
b
8
15900
c
9
22700
c
10
11
12
17100
c
10000
15800
16400
b
19000
a
b
All reactions were performed with 5 mol % of HG2 in refluxing dichloromethane for 18 h at 0.4 M with 4 equiv of coupling partner. Monomodal.
c
Multimodal.
Scheme 2. Metathesis Pathways for p(AGE) and p(CGE)
Scheme 3. Metathesis Pathways for p(PGE)
hindrance in metallacyclobutane 8 compared to that of 7, but
this hypothesis was not validated.
Polyether p(PGE) bears type 2 pendent olefins, so formation
of carbene 4 is slow, but in the case of a type 2 olefin partner
such as methyl acrylate, formation of carbene 9a is even slower
(Scheme 3). The catalyst reacts first with p(PGE), and
unhindered carbene 4 undergoes rapid CM reaction. SM is
very slow because of the high steric hindrance of metal-
lacyclobutane 10, and since the desired product 5 is a type 2
olefin, no secondary metathesis is observed. This explains why
p(PGE) is the optimal polymer for coupling with a type 2
olefin.
Scheme 4. Metathesis Pathways for p(MAGE)
In the case of a type 1 coupling partner, the catalyst reacts
first with the small molecule giving carbene 9b. This carbene
can then couple with a polymer olefin to give the desired
product 5. No SM metathesis product is formed by this
pathway, but the CM product 5 is a type 1 olefin that can lead
to the SM product 6 by secondary metathesis reaction, which
explains why p(PGE) produces the same amount of SM
product as the less hindered p(AGE) and p(CGE) polymers
with type 1 olefin partners. Polymer p(MAGE) is a type 3
olefin and does not react with the catalyst, so carbene 11
cannot be formed (Scheme 4). With a type 2 olefin, formation
of carbene 9a is very slow, and no metathesis reaction is
observed. With a type 1 coupling partner, formation of carbene
9b is fast, followed by a slow reaction with p(MAGE) to give
the desired CM product 12. Secondary SM cannot occur at all,
as this product is also a type 3 olefin.
C
DOI: 10.1021/acs.orglett.8b00595
Org. Lett. XXXX, XXX, XXX−XXX

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Organic Letters
Letter
Notes
Having established that the best candidate in terms of
preventing SM in the presence of type 1 olefins is p(MAGE),
we then turned to a more challenging coupling partner, such as
the protected tripeptide RGD. When using the same linker as
for Boc-glycine, the conversion was only 9%, but with only 1
equiv of the hexenyl ester of RGD, 31% successful CM was
observed, with no trace of SM products (Scheme 5).
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank Prof. Craig Hawker (UC Santa Barbara) and
Prof. Karol Grela (University of Warsaw) for a generous gift of
polymer samples and isocyanide scavenger, respectively, as well
as Prof. Michael Shaver (University of Edinburgh) and Prof.
Dave Adams (University of Glasgow) for helpful discussions.
Financial support for this work was provided by the University
of Glasgow, the EPSRC (Doctoral Training Allocation for S.M.
EP/K503058/1), and Tenovus Scotland (Project S17-04).
Scheme 5. Successful CM between p(MAGE) and Protected
RGD22
REFERENCES
■
(1) Banerjee, S. S.; Aher, N.; Patil, R.; Khandare, J. J. Drug Delivery
2012, 2012, 1.
(2) Gauthier, M.; Gibson, M.; Klok, H.-A. Angew. Chem., Int. Ed.
2009, 48, 48−58.
(3) Delaittre, G.; Guimard, N. K.; Barner-Kowollik, C. Acc. Chem. Res.
2015, 48, 1296−1307.
(4) Willcock, H.; O’Reilly, R. K. Polym. Chem. 2010, 1, 149−157.
(5) Hoyle, C. E.; Bowman, C. N. Angew. Chem., Int. Ed. 2010, 49,
1540−1573.
(6) Rieger, J.; Van Butsele, K.; Lecomte, P.; Detrembleur, C.; Jer
́ ̂
ome,
R.; Jer
́ ̂
ome, C. Chem. Commun. 2005, 274−276.
(7) Singha, N. K.; Gibson, M. I.; Koiry, B. P.; Danial, M.; Klok, H.
Biomacromolecules 2011, 12, 2908−2913.
(8) For a review on metathesis, see: Prunet, J. Eur. J. Org. Chem.
2011, 2011, 3634−3647.
(9) For reviews, see: (a) Sinclair, F.; Alkattan, M.; Prunet, J.; Shaver,
M. P. Polym. Chem. 2017, 8, 3385−3398. (b) Dong, Y.; Matson, J. B.;
Edgar, K. J. Biomacromolecules 2017, 18, 1661−1676.
(10) Mathers, R. T.; Coates, G. W. Chem. Commun. 2004, 422−423.
(11) de Espinosa, L. M.; Kempe, K.; Schubert, U. S.; Hoogenboom,
R.; Meier, M. A. R. Macromol. Rapid Commun. 2012, 33, 2023−2028.
(12) Meng, X.; Matson, J. B.; Edgar, K. J. Polym. Chem. 2014, 5,
7021−7033.
(13) Fournier, L.; Robert, C.; Pourchet, S.; Gonzalez, A.; Williams,
L.; Prunet, J.; Thomas, C. M. Polym. Chem. 2016, 7, 3700−3704.
(14) Sinclair, F.; Chen, L.; Greenland, B. W.; Shaver, M. P.
Macromolecules 2016, 49, 6826−6834.
(15) Lee, B. F.; Kade, M. J.; Chute, J. A.; Gupta, N.; Campos, L. M.;
Fredrickson, G. H.; Kramer, E. J.; Lynd, N. A.; Hawker, C. J. J. Polym.
Sci., Part A: Polym. Chem. 2011, 49, 4498−4504.
In conclusion, we have synthesized three novel polymers
with pendent alkenes, where these handles are designed for
optimum CM with various olefinic compounds, avoiding
undesired SM without using a large excess of the small
molecule partner. The best polymer for coupling with type 2
olefins is the p(PGE), while no SM occurs during CM of
p(MAGE) with type 1 olefins, maintaining a good dispersity
throughout the functionalization process. Furthermore, we
report the first successful CM between a polymer and a
coupling partner of biological relevance, RGD, which is
commonly used for targeting tumor cells.23 Further studies
for conjugation of p(MAGE) with drugs such as paclitaxel are
in progress.
(16) First- and second-generation Grubbs catalysts were also
investigated but gave lower percentages of the desired CM (x unit).
(17) DMSO or the isocyanide is used to deactivate the metathesis
catalyst. For the isocyanide scavenger, see: Szczepaniak, G.; Urbaniak,
́
K.; Wierzbicka, C.; Kosinski, K.; Skowerski, K.; Grela, K.
ChemSusChem 2015, 8, 4139−4148.
ASSOCIATED CONTENT
* Supporting Information
■
(18) Kuhlmann, M.; Groll, J. RSC Adv. 2015, 5, 67323−67326.
(19) See the Supporting Information for the synthesis of the
monomers.
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b00595.
(20) Chatterjee, K.; Choi, T.-L.; Sanders, D. P.; Grubbs, R. H. J. Am.
Chem. Soc. 2003, 125, 11360−11370.
(21) The catalyst may also react with the coupling partner if it is a
type 1 olefin, resulting in successful CM when coupling with the
polymer.
(22) Pbf = 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl.
(23) Wang, F.; Li, Y.; Shen, Y.; Wang, A.; Wang, S.; Xie, T. Int. J. Mol.
Sci. 2013, 14, 13447−13462.
Experimental procedures for preparation of polymers and
CM partners and for CM reactions, NMR spectra, and
GPC chromatograms (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: joelle.prunet@glasgow.ac.uk.
ORCID
Rob M. J. Liskamp: 0000-0001-8897-8975
Joelle Prunet: 0000-0002-9075-971X
̈
D
DOI: 10.1021/acs.orglett.8b00595
Org. Lett. XXXX, XXX, XXX−XXX