Synthesis, purification, and characterization of "perfect" star polymers via "click" coupling

Article abstract of DOI:10.1002/pola.25864

The copper (I)-catalyzed azide-alkyne cycloaddition "click" reaction was successfully applied to prepare well-defined 3, 6, and 12-arms polystyrene and polyethylene glycol stars. This study focused particularly on making "perfect" star polymers with an ex

Full text of DOI:10.1002/pola.25864

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Synthesis, Purification, and Characterization of ‘‘Perfect’’ Star Polymers
via ‘‘Click’’ Coupling
Yejia Li, Boyu Zhang, Jessica N. Hoskins, Scott M. Grayson
Department of Chemistry, Tulane University, New Orleans, Louisiana 70118
Correspondence to: S. M. Grayson (E-mail: sgrayson@tulane.edu)
Received 6 October 2011; accepted 7 November 2011; published online 15 December 2011
DOI: 10.1002/pola.25864
ABSTRACT: The copper (I)-catalyzed azide-alkyne cycloaddition
‘‘click’’ reaction was successfully applied to prepare well-
defined 3, 6, and 12-arms polystyrene and polyethylene glycol
stars. This study focused particularly on making ‘‘perfect’’ star
polymers with an exact number of arms, as well as developing
techniques for their purification. Various methods of characteri-
zation confirmed the star polymers high purity, and the struc-
tural uniformity of the generated star polymers. In particular,
matrix-assisted laser desorption ionization-time-of-flight mass
spectrometry revealed the quantitative transformation of the
end groups on the linear polymer precursors and confirmed
their quantitative coupling to the dendritic cores to yield star
polymers with an exact number of arms. In addition to prepar-
ing well-defined polystyrene and poly(ethylene glycol)homo-
polymer stars, this technique was also successfully applied to
C
V
amphiphilic, PCL-b-PEG star polymers.
2011 Wiley Periodi-
cals, Inc. J Polym Sci Part A: Polym Chem 50: 1086–1101, 2012
KEYWORDS: atom transfer radical polymerization (ATRP);
MALDI-TOF MS; separation of polymers; star polymers
INTRODUCTION Star polymers are branched polymers that
consist of multiple linear chains connected to a central core.1
Star polymers’ multiarm structure, globular shape, and mul-
tiplicity of end groups imparts on them a unique set of prop-
erties (e.g., crystalline, mechanical, and viscoelastic proper-
ties), when compared with their linear analogs. Perhaps
most significantly, the three-dimensional architecture of
appropriately designed amphiphilic star polymers can be tai-
lored to encapsulate guest molecules providing relatively
rapid access to well-defined nanocarriers for applications in
drug delivery.2–4 There are three general strategies for the
synthesis of star polymers, each of which exhibits particular
advantages and disadvantages: the ‘‘core first’’ approach, the
‘‘arm first’’ approach, and the ‘‘graft onto’’ approach.5–8 The
‘‘core first’’ approach initiates polymerization of the arms
from a polyfunctional core, which likely results in nonuni-
form arms lengths because of potential differences in the
reactivity of each initiating group. Likewise, when comparing
a set of star polymers prepared from different cores, it is dif-
ficult to ensure uniformity of arm length from batch to
batch. The ‘‘arm first’’ approach involves first the polymeriza-
tion of the arms followed by their coupling, usually by con-
tinuation of the polymerization with addition of a crosslink-
ing monomer. Although technically simple, this typically
leads to materials with an inexact number of arms and
broader polydispersities. The third major route, the ‘‘graft
onto’’ approach involves the attachment of preformed arms
onto core molecules via an efficient coupling reaction. This
approach makes the conjugation of arms of uniform length
onto different cores possible but requires highly efficient
conjugation reactions to complete the coupling to the core.
This route also requires an excess linear polymer arms to
drive the coupling reaction to completion, and therefore
involves tedious purification to remove the unreacted linear
reactant. The ‘‘graft onto’’ method typically affords the most
structurally uniform star polymers and provides access to
comparable libraries with different cores yet identical arm
lengths, however, truly well-defined structures can only be
accessed with highly efficient coupling reactions and sub-
stantial purification efforts.
Much of the early work preparing well-defined star polymers
used anionic polymerization methods owing to the excep-
tional control of this technique but suffered from limited
monomer compatibility and extremely tedious experimental
preparation due to the technique’s sensitivity to trace impur-
ities.9 Recently, the development of controlled/living radical
polymerization,10 especially atom transfer radical polymer-
ization (ATRP),11,12 enabled the facile preparation of narrow
dispersed polymers with broad monomer compatibility yet
bearing uniformly end-functionalized polymer groups.
Around the same time, the optimization of dendrimer chem-
istry afforded a number of efficient syntheses that provide
access to well-defined dendritic cores, with exact numbers of
functional groups.13 The combination of these two
Additional Supporting Information may be found in the online version of this article.
C
V
2011 Wiley Periodicals, Inc.
1086
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
approaches provides relatively straight forward routes to the
well-defined precursors to make high purity star polymers
via the ‘‘graft onto’’ approach.
synthesized by described procedure.35 The detail can be
found in the Supporting Information.
The only remaining requirement for preparing well-defined
star polymers, a highly efficient coupling reaction, is pro-
vided by the copper-catalyzed azide-alkyne cycloaddition
(CuAAC) ‘‘click’’ coupling reaction.14–16 This particular reac-
tion is exceptionally powerful in overcoming the steric inhi-
bition that frequently prevents the quantitative coupling of
polymeric substrates. Its usage for polymer conjugations has
been demonstrated widely17–20 but is perhaps most appa-
rent in the near quantitative coupling of hindered macromo-
lecular components. Representative examples include the
synthesis of dendronized polymers by coupling third genera-
tion dendrons onto each repeat unit of either a linear poly(-
vinylacetylene) backbone21 or a cyclic styrenic backbone;22
the synthesis of polymer-peptide conjugates of bovine serum
albumin with polystyrene (4.15 kDa)23 or poly(N-isopropyla-
crylamide) (PNIPAM) (16.3 kDa);24 and the synthesis of graft
polymers via the coupling of 775 Da PEG chains onto a 27.5
k MW poly(hydroxyethyl methacrylate) backbone.25 The
CuAAC reaction has already been used in numerous exam-
ples to make star polymers, including poly(styrene) stars,25
poly(tert-butyl acrylate) and poly(ethylene glycol) stars,26
cyclodextrin-core seven-arm star poly(e-caprolactone)
(PCL),27 cyclodextrin-core seven-arm and 21-arm star PNI-
PAM,28 POSS-core block copolymer stars,29,30 and so forth. In
addition, other types of ‘‘click’’ reactions, such as thiol-ene
reaction31 and Diels-Alder32 reactions have also been used
to synthesize star polymers via the ‘‘graft onto’’ method.
However, most of these materials were not analyzed in suffi-
cient detail to judge their purity with regards to missing
arms. Because of the interest in star polymers for biomedical
applications such as transdermal drug carriers,4,33,34 it is
critical to develop synthetic, purification, and characteriza-
tion methodologies, which can reproducibly yield ‘‘perfect’’
star polymers, those which exhibit the exact number of arms
as were targeted in their synthesis. For these reasons, a
detailed study of the coupling, purification, and characteriza-
tion of star polymers was performed and described below.
Characterization
All 1H NMR (400 MHz) and 13C NMR (100 MHz) were obtained
using a Varian Mercury spectrometer (Palo Alto, CA), using TMS
¼ 0.00 ppm for 1H calibration. Gel permeation chromatography
(GPC) was performed on a Waters model 1515 series pump
(Milford, MA) with three column series from Polymer Laborato-
ries, consisting of PLgel 5 lm Mixed D (300 mm ꢀ 7.5 mm, mo-
lecular weight range 200–400,000), PLgel 5 lm 500 Å (300
mm ꢀ 7.5 mm, molecular weight range 500–30,000), and PLgel
5 lm 50 Å (300 mm ꢀ 7.5 mm, molecular weight range up to
2000) columns. The system was fitted with a Model 2487 dif-
ferential refractometer detector and anhydrous THF was used
as the mobile phase (1 mL minꢁ1 flow rate). The resulting mo-
lecular weight was based on calibration using linear polysty-
rene standards. Data were collected and processed using Preci-
sion Acquire software. Mass spectral data was acquired using a
Bruker Autoflex III matrix-assisted laser desorption ionization-
time-of-flight mass spectrometer (MALDI-TOF MS) with delayed
extraction using both positive ion and reflector detection
modes. For all PCL polymers, THF stock solutions of 9-nitroan-
thracene as the matrix (20 mg mLꢁ1) and NaI as the counterion
(10 mg mLꢁ1) were used. The polymer sample was prepared at
a 2 mg mLꢁ1 concentration in THF. MALDI samples were pre-
pared by combining 50 lL of polymer solution, 100 lL of coun-
terion solution, and 200 lL of matrix solution. For all PEG poly-
mers, THF stock solutions of alpha-cyano 4-hydroxycinnamic
acid (20 mg mLꢁ1) and sodium trifluoroacetate as the counter-
ion (1 mg mLꢁ1) were used. The polymer sample was prepared
at a 2 mg mLꢁ1 concentration in THF. MALDI samples were pre-
pared by combining 20 lL of polymer solution, 0.5 lL of coun-
terion solution, and 10 lL of matrix solution. For all PSt poly-
mers, THF stock solutions of dithranol (20 mg mLꢁ1) and silver
trifluoroacetate as the counterion (1 mg mLꢁ1) were used. The
polymer sample was prepared at a 2 mg mLꢁ1 concentration in
THF. MALDI samples were prepared by combining 20 lL of
polymer solution, 0.5 lL of counterion solution, and 10 lL of
matrix solution.
General Procedure for Synthesis of Alkynylated Dendrons
The alkynylated dendrons were synthesized according to a
previously reported procedure.36 Dendrimers with hydroxyl
group on the periphery, pentynoic acid anhydride, 4-Dime-
thylaminopyridine (DMAP), and pyridine (1 hydroxide:1.67
pentynoic acid anhydride:0.33 DMAP:6 pyridine) were added
to a round bottom flask with dichloromethane. After stirring
for 12 h in room temperature under nitrogen, the mixture
was quenched by DI water for about 2 h. Sequentially the
reaction was washed three times with 1 M NaHSO4 solution,
1 M NaHCO3 solution and of brine. The organic layer was
dried over anhydrous MgSO4, filtered, and concentrated
before dissolve in diethyl ether. After filtration, the filtrate
was concentrated and precipitated into 40 mL of cold hex-
anes. The product was isolated as oil and dried in vacuo. The
specific details are described in the Supporting Information.
EXPERIMENTAL
Materials
All reagents were purchased from Aldrich and used without
further purification, unless otherwise noted. e-CL was dried
over calcium hydride at room temperature for at least 8 h
and distilled under reduced pressure just before use. Sol-
vents were purchased from Pharmaco Aaper, Aldrich, and
Fisher Scientific. Tetrahydrofuran (THF) was dried over so-
dium refluxing overnight. Dichloromethane and hexane were
dried over calcium hydride at room temperature overnight.
All other solvents were reagent grade and used without fur-
R
ther distillation or purification. Bio-BeadsV S-X beads are po-
rous crosslinked polystyrene polymers used for gel permea-
tion separations of hydrophobic polymers in the presence of
organic solvents. The dendrimers with hydroxyl group were
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101
1087

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Synthesis of Polystyrene Bromide
PEG-OMs was dissolved in 25 mL of dried DMF, followed by
the addition of sodium azide. The reaction was heated to 50
ꢂC and stirred under nitrogen for 12 h. The reaction was dis-
solved in 50 mL of DCM, and then washed three times with
100 mL of DI H2O, and three times with 100 mL of brine.
The organic layer was dried over anhydrous MgSO4, filtered,
and concentrated before precipitation from dichloromethane
into 20 mL of cold ethyl ether. The product was isolated via
filtration and dried in vacuo; 78% yield: 1H NMR (400 MHz,
CDCl3) d 1.99 (s, 2H), 3.35 (s, 3H), 3.35–3.37 (m, 4H), 3.62–
3.64 (br, 189H). 13C NMR (100 MHz, CDCl3,) d 50.87, 59.27,
70.25, 70.78, 72.13. MS (MALDI-TOF): (m/z), Mn ¼ 2100;
PDI ¼ 1.01. GPC: Mn ¼ 2700; PDI ¼ 1.02.
A 500 mL round bottomed flask containing 2.73 g (19.06
mmol) CuBr, 0.75 g (3.37mmol) CuBr2, 4.66 g (26.9 mmol)
PMDETA, and 300 g (2.88 mol) styrene was degassed using
two freeze/pump/thaw cycles. After warming to room tem-
perature, 4.15 g (22.4 mmol) of the initiator was added via
syringe. The reaction mixture was placed in a preheated 75
ꢂC oil bath and allowed to stir under nitrogen for 56 min.
The reaction mixture was then cooled to room temperature
and purified by extraction from water into dichloromethane
followed by precipitation into methanol to yield the polymer
as a white solid: 26.9 g (yield 8.86%). 1H NMR (400 MHz,
CDCl3) d 0.9–1.1 (br, 3), 1.3–2.6 (br, 72), 4.4–4.6 (br, 1), 6.4–
7.4 (br, 120). MS (MALDI-TOF): (m/z), Mn ¼ 2450; PDI: 1.07.
GPC: Mn ¼ 2800; PDI ¼ 1.12. FTIR: 3082, 3061, 3026, 2923,
2850, 1943, 1872, 1803, 1685, 1601, 1493, 1453, 1373,
Synthesis of Three Arm PCL-OH
1181, 1154, 1070, 1028, 907, 759, 698 cmꢁ1
.
1,1,1-tris(hydroxymethyl)ethane (0.421g, 3.5 mmol) was
placed in a previously dried 25 mL two neck round bottom
flask equipped with a stir bar and dried under vacuum for
about 1 h. e-Caprolactone (10 g, 87.6 mmol) was added
through the rubber septum via syringe to the initiator. After
all initiator was dissolved in the monomer, tin(II) ethylhexa-
noate (0.177g, 0.4 mmol) was added via syringe, and the
reaction flask was immediately submerged in an oil bath at
Synthesis of Polystyrene Azide
To a 500 mL round bottomed flask was added 26.9 g (10.4
mmol) PS-Br. The polymer was then dissolved into 60 mL
dimethylformamide (DMF), and 3.37 g (0.51.8 mmol) sodium
azide was added as a solid. The solution was allowed to stir
overnight at room temperature before purification by precip-
itation into methanol to give a white solid: 22.3 g (yield
83%). 1H NMR (400 MHz, CDCl3) d 0.9–1.1 (br, 3), 1.3–2.6
(br, 72), 3.8–4.1 (br, 1), 6.4–7.4 (br, 120). MS (MALDI-TOF):
(m/z), Mn ¼ 2500; PDI: 1.04. GPC: Mn ¼ 2800; PDI ¼ 1.12.
FTIR: 3082, 3060, 3026, 2923, 2849, 2094, 1943, 1872,
1803, 1745, 1601, 1493, 1453, 1374, 1181, 1154, 1069,
ꢂ
130 C. After 3 h 10 min, the flask was removed from the oil
bath, and the contents were diluted with dichloromethane,
precipitated twice into a 1:1 mixture of hexanes and diethyl
ether, recovered by filtration, and dried in vacuo before char-
acterization. Characterization data: 1H NMR (400 MHz,
CDCl3) d 1.31 (m, 2n), 1.57 (m, 4n), 1.88 (m, 2, J ¼ 6.3 Hz),
2.27 (t, 2n, J ¼ 7.6), 3.37 (t, 2, J ¼ 6.3), 3.60 (t, 2, J ¼ 6.6
Hz), 4.03 (t, 2n, J ¼ 6.3 Hz), 4.14 (t, 2, J ¼ 6.2 Hz). Calcd Mn
¼ 4600. 13C NMR (CDCl3, d, ppm) 24.78, 25.73, 28.54, 34.32,
64.36, 173.79. MS (MALDI-TOF): (m/z), Mn ¼ 2220, PDI ¼
1.06. GPC: Mn ¼ 4040, PDI ¼ 1.08.
1028, 907, 760, 698 cmꢁ1
.
Synthesis of PEG-OMs
A typical procedure of synthesis of PEG-OMs was previously
reported.37 PEG-OH (Mn ¼ 2000, 4 g, 2 mmol) was heated to
ꢂ
50 C in oil bath and dried in vacuo for 12 h. It was cooled
to room temperature before dissolved in 100 mL of dried
dichloromethane. The solution was cooled to 0 ꢂC in ice-
water bath, and distilled triethylamine (1.53 mL, 11.0 mmol)
and methanesulfonyl anhydride (1.74 g, 10 mmol) were
sequentially slowly added. After stirring for 12 h in room
temperature under nitrogen, the mixture was filtered and
sequentially washed three times with 100 mL of 1 M
NaHSO4 solution, 100 mL of 1 M NaHCO3 solution and 50
mL of brine. The organic layer was dried over anhydrous
MgSO4, filtered, and concentrated before precipitation from
dichloromethane into 40 mL of cold ethyl ether. The product
was isolated via filtration and dried in vacuo; 72% yield. 1H
NMR (400 MHz, CDCl3) d 3.06 (s, 3H), 3.35 (s, 3H), 3.51–
3.53 (m, 2H), 3.61–3.63 (br, 189H), 4.36–4.37 (m, 2H). 13C
NMR (100 MHz, CDCl3) d 37.91, 59.23, 61.80, 69.20, 69.52,
70.47, 72.08, 72.84. MS (MALDI-TOF): (m/z), Mn ¼ 2260;
PDI: 1.01. GPC: Mn ¼ 2800; PDI ¼ 1.02.
Synthesis of Six Arm PCL-OH
G1(OH)6 (0.295g, 0.63 mmol) was placed in a previously
dried 25 mL two neck round bottom flask equipped with a
stir bar and dried under vacuum at 110 ꢂC for 24 h. The
temperature of the flask was then raised to 130 ꢂC under
vacuum and backfilled with nitrogen. e-Caprolactone (2.67 g,
25.2 mmol) was added through the rubber septum via sy-
ringe to the initiator. After all initiator was dissolved in
monomer, tin(II) ethylhexanoate (0.025 g, 0.07 mmol) was
added via syringe. After 40 min after the addition of the cat-
alyst, the flask was removed from the oil bath, and the con-
tents were diluted with dichloromethane, precipitated twice
into a 1:1 mixture of hexanes and diethyl ether, recovered by
filtration, and dried in vacuo before characterization. Charac-
terization data: 1H NMR (CDCl3, d, ppm) 1.31 (m, 2n), 1.57
(m, 4n), 1.88 (m, 2, J ¼ 6.3 Hz), 2.27 (t, 2n, J ¼ 7.6), 3.37 (t,
2, J ¼ 6.3), 3.60 (t, 2, J ¼ 6.6 Hz), 4.03 (t, 2n, J ¼ 6.3 Hz),
4.14 (t, 2, J ¼ 6.2 Hz). calcd Mn ¼ 4600. 13C NMR (CDCl3, d,
ppm) 24.78, 25.73, 28.54, 34.32, 64.36, 173.79. MS (MALDI-
TOF): (m/z), Mn ¼ 5800, PDI ¼ 1.06. GPC: Mn ¼ 7760, PDI
¼ 1.14.
Synthesis of PEG-N3
A typical procedure of synthesis of PEG-N3 was previously
reported.20 PEG-OMs (2.00 g, 1.00 mmol) and sodium azide
(0.33 g, 5.00 mmol) were separately heated to 50 ꢂC under
vacuum for 12 h, and then cooled to room temperature. The
1088
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 1 Synthesis of alkynylated dendrimer cores 2, 3, and 4.
The Preparation of Three Arms PCL-Alkynes
The polymer three arms PCL-OH (1.122 g, 0.51 mmol) was
dissolve in 10 mL of dichloromethane. 4-pentynoic anhydride
(1.000 g, 5.61mmol), pyridine (0.755 g, 9.51mmol), and 4-
(dimethylamino)pyridine (0.069 g, 0.561mmol) were added
to the reaction flask. The reaction was stirred at room
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101
1089

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
temperature for 10 h. Then the crude reaction mixture was
extracted three times from saturated aqueous NaHSO4 and
three times from saturated aqueous NaHCO3 into dichloro-
methane. The organic layer was dried over anhydrous
Na2SO4. After removing most of the solvent, the product was
isolated by precipitation from dichloromethane into cold
1
methanol. Yield (89%). H NMR (400 MHz, CDCl3) d 1.31 (m,
2n), 1.57 (m, 4n), 1.88 (m, 2), 1.98 (s, 1), 2.27 (t, 2n,), 2.51
(t, 4,), 4.03 (t, 2n). MS (MALDI-TOF) (m/z), Mn ¼ 2660, PDI
¼ 1.04). GPC: Mn ¼ 4600, PDI ¼ 1.08.
Preparation of Six Arms PCL-Alkynes
The polymer six arms PCL-OH (0.200g) was dissolve in 10
mL of dichloromethane. 4-pentynoic anhydride (0.114 g),
pyridine (0.098 g), and 4-(dimethylamino)pyridine (0.008 g)
were added to the reaction flask. The reaction was stirred at
room temperature for 10 h. Then, the crude reaction mixture
was extracted three times from saturated aqueous NaHSO4
and three times from saturated aqueous NaHCO3 into
dichloromethane. The organic layer was dried over anhy-
drous Na2SO4. After removing most of the solvent, the prod-
uct was got by precipitation from dichloromethane into cold
1
methanol. Yield (87%) H NMR (400 MHz, CDCl3) d 1.31 (m,
2n), 1.57 (m, 4n), 1.98 (s, 1), 2.27 (t, 2n), 2.51 (t, 4), 4.03 (t,
2n), 4.14 (t, 2). MS (MALDI-TOF) (m/z), Mn ¼ 6330, PDI ¼
1.04). GPC: Mn ¼ 8600, PDI ¼ 1.08.
General Procedure of Cu(I)-Catalyzed ‘‘Click’’ Assembly of
Star Polymers
Alkyne functionalized star cores, azide functionalized macro-
molecule, and PMDETA (1 alkyne:1.05 azido-polymer:2
PMDETA) were added to a round bottom flask with dichloro-
methane. After two freeze-pump-thaw cycles, Cu(I)Br (1:1
with PMDETA) was added, and the third freeze-pump-thaw
cycle was performed. The reaction was stirred for 16 h at
room temperature, then was washed with de-ionized water
and brine, and finally was dried with Na2SO4. Following con-
centration in vacuo, the products were run through a short
silica gel column to remove residual copper. The detail pro-
cedure for synthesis of star polymers is described in the
Supporting Information.
RESULTS AND DISCUSSION
To prepare ‘‘perfect’’ star polymers via the ‘‘graft onto’’
approach, well-defined polymer arms and dendritic cores
must first be prepared with complimentary functionalities.
The dendritic cores were prepared via the highly efficient di-
vergent synthetic approach, using benzylidene protected bis-
MPA.38 After a series of monodisperse dendrimers with 3, 6,
and 12 hydroxyl groups were obtained (‘‘generations’’ G0,
G1, and G2, respectively), they were reacted with pentynoic
acid anhydride to quantitatively introduce alkyne groups
onto their periphery to yield core 2, 3, 4 (Scheme 1).
FIGURE 1 MALDI-TOF mass spectrum of dendritic core 2, 3,
and 4.
The acquired 1H and 13C NMR spectra verified the purity of
the products, which exhibiting exceptional symmetry. The
quantitative esterification of the peripheral hydroxyls could
be confirmed by comparing the integration of the resonances
of the terminal alkyne (1.98 ppm) and those of the methyl-
ene adjacent to the ester oxygen (4.05 ppm) to those of the
methyl groups at the core (G0: 1.03 ppm; G1: 1.05 ppm; G2:
1.12 ppm) (Supporting Information Fig. S1 and Table S1).
Elemental analysis was also used to confirm the purity of
1090
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 2 Synthesis of 3, 6, and 12-arm star PS and PEG.
the dendritic cores used in this study. However, the cores’
MALDI-TOF mass spectra were most valuable for confirming
the truly monodisperse nature of alkynylated dendrimers as
only one signal was observed in each one of their mass spec-
tra (Fig. 1).
With a polymer of unknown end groups, it is impossible to
calculate its end group mass exactly, exclusively from the
mass spectra, because it is unknown how many monomer
units contribute to a given n-mers mass, and how much is
contributed by the end groups. A technique has been devel-
oped, however, to define the set of possible end group
masses. This is performed by determining the smallest possi-
ble end group mass, Mres, by extrapolation from every n-mer
in the mass spectra. From this value of Mres, the series of
possible end group masses can be identified (by adding mul-
tiples of the monomer mass). Because the hypothetical end
group structure is known for these linear and star polymers,
Because the linear arms must also exhibit narrow polydis-
persity and precise end group control, ‘‘living’’ polymeriza-
tion techniques were used to prepare the azido-functional
arms. The linear polystyrene precursors were synthesized
via ATRP from a 2-ethylbromo benzene initiator 5, using
Cu(I)Br/Cu(II)Br2 as catalyst and PMDETA as a ligand
(Scheme 2).
the theoretical end group mass (Mend) can be easily
To ensure near-quantitative functionalization with a bromide
group at the end of each polymer chain, the reaction was
stopped at a low conversion of the monomer (20%) yielding
polystyrene 6 with a Mn of 2800, and PDI ¼ 1.12 by GPC.
Their MALDI-TOF mass spectra exhibited monomodal distri-
butions and the observed signals matched the expected dis-
tribution caused by complexation to Agþ in addition to the
previously reported elimination of HBr [PSBrþAgþ-HBr].39
Calculations from the MALDI-TOF mass spectra confirmed
the well-defined character of the polymers, and agreed
closely with the GPC data (Mn ¼ 2500, PDI ¼ 1.12; Fig. 2
and Table 1).
To confirm the end group identity, a previously reported pro-
cedure was used to calculate from mass spectral data the
‘‘residual mass (Mres)’’ and the ‘‘end group mass (Mend).’’
FIGURE 2 MALDI-TOF mass spectrum of polystyrene 6.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101
1091

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 1 Repeat Unit (Mru) and End Group Mass (Mres) Data
for Polystyrene, 6
case n ¼ 0 yields 104.01, which agrees closely with the theo-
retical value of 104.06. For the larger 12-arm star polymers,
the substantial mass of the core yields solutions where n
>> 0.
Theoretical
(6þAg-HBr)þ
Observed
(2.1aþAg-HBr)þ
Mru
Mres
N
(6n-6n ꢁ 1
)
for 6
To enable their click conjugation, the bromine end groups
were then converted to the azide by reaction with ꢃ5 equiv-
alents of sodium azide in DMF40 The product retained the
narrow polydispersity of the precursor, as judged by GPC
(Mn ¼ 2800, PDI ¼ 1.12). The MALDI-TOF mass spectra pro-
vided additional strong evidence for the near-quantitative
conversion to the azide end group, as the previously
described bromide resonances were completely lost, and a
new set of resonances, dominated by the unique metastable
fragmentation of the azide functionality [PSN3þAg-N2]þ
(metastable ion observed as [PSN3-23]þ) were observed in
reflector mode. Further confirmation of the metastable na-
ture of these signals was gained by acquiring the MALDI-MS
spectra in linear mode (which does not exhibit the effects of
metastable fragmentation), yielding a strong set of signals
corresponding to [PSN3þNa] in addition to other in-source
decay fragments (e.g., [PSN3þAg-N2(in source)]þ observed as
[PSN3-28.0]þ (Supporting Information Fig. S2 and Table S2).
The unusual behavior has been observed for multiple azide
containing polymers and examined in detail elsewhere.41
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
1459.72
1563.78
1667.84
1771.91
1875.97
1980.03
2084.09
2188.16
2292.22
2396.28
2500.34
2604.41
2708.47
2812.53
2916.60
3020.66
3124.72
3228.78
1459.67
1563.79
1667.81
1771.88
1875.93
1979.99
2084.08
2188.12
2292.17
2396.21
2500.27
2604.32
2708.39
2812.44
2916.49
3020.54
3124.59
3228.68
Average
Theoretical
104.02
104.07
104.04
104.05
104.03
104.03
104.06
104.04
104.02
104.00
104.00
103.99
104.00
103.99
103.97
103.96
103.95
103.97
104.01
104.06
104.12
104.03
104.07
104.05
104.05
104.09
104.04
104.05
104.04
104.07
104.05
104.07
104.05
104.05
104.05
104.05
104.09
104.06
104.06
The azido-PEG was prepared following the previously
reported procedure (Scheme 2).42 By reacting with methane-
sulfonyl anhydride with monomethyl polyethylene glycol
(PEG-OH) (Mn ¼ 2090), the hydroxyl end group of PEG-OH 8
could be quantitatively converted to be mesylate group. This
result was verified by their MALDI-TOF spectra, which exhib-
ited a mass increase of 78.07 Da (Dtheo.¼ 77.98 Da) for each
peak (Fig. 3).
determined and compared with the calculated mass. For
example, in the case of the bromide terminated polymer,
which was known to ionize via elimination of HBr and com-
plexation with Agþ, the mass of each monoisotopic peak was
determined, and the trend extrapolated to the smallest possi-
ble positive value, which yielded and Mres of 104.01, and a
repeat unit mass of 104.06. The repeat unit mass agrees
with the expected value (104.0626), while the end group
mass could be one of the series 104.01 þ n(104.06). In this
Subsequently, the mesylated PEG 9 was reacted with 5
equivalent of sodium azide40 to afford PEG with a single az-
ide end group. Again, MALDI-TOF revealed that the azido-
PEG 10 was formed near quantitatively with the expect shift
of molecular weight (Dobs. ¼ ꢁ52.03 Da, Dtheo. ¼ ꢁ52.97
FIGURE 3 MALDI-TOF mass spectra of the PEG methyl ether 2000, 8, mesylated PEG, 9, and azido-PEG, 10.
1092
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 2 MALDI-TOF and GPC Characterization of Purified Polystyrene Stars with Calculated Value in Parenthesis
Compound
PS Stars
Mn by MALDI
PDI
Mn by GPC
PDI
Mru
Mres
7
PS-N3
3-arm
6-arm
12-arm
2,500
1.04
1.02
1.01
1.02
2,800
1.12
1.07
1.04
1.05
104.1 (104.1)
104.2 (104.1)
104.2 (104.1)
N/A
ꢁ0.1 (0)
72.9 (72.9)
60.9 (61.5)
N/A
11
12
13
8,110 (7,860)
15,000 (15,900)
30,000 (32,100)
8,100
13,600
18,300
Da) and the appearance of the characteristic azide metasta-
ble ion using reflector mode. These end group transforma-
tion were also confirmed by the 13C NMR spectrum, where
the signal of CH2OH (59 ppm) disappeared in mesylated PEG
9 spectrum and the signal of –OSO2CH3 (38 ppm) disap-
peared in 10 spectrum (Supporting Information Fig. S3).
confirmed by the fact that the products no longer exhibit
any metastable fragmentation, consistent with previous
observations, which confirm that the metastable decay char-
acteristic of polymer azides is lost when they are converted
to the relatively hearty triazole linkages.43 Notably, no signal
is observed in the molecular weight range of 3000 to 5000
that would correspond to the star with only two coupled
arms.
The assembly of star polymers could then be performed
using standard CuAAC ‘‘click’’ reaction condition (Scheme
2).42 A slight excess of linear polymer was used to ensure
the complete formation of star polymers, typically using
1:1.05 mole ratio of alkyne to azide. Because this approach
requires very carefully measured ratios of linear polymer
and dendritic core, the molecular weight of the polymer
arms must be determined with high accuracy. For the below
examples, the molecular weight used for stoichiometric cal-
culations was determined from MALDI-TOF MS. All GPC and
MALDI-TOF results suggest that the resultant crude star
polymer sample exhibited quantitative coupling of polymer
to each of the alkyne functionalities on the core molecules
(Tables 2 and 3).
Because of the difficulty in precisely determining the average
molecular weight of the azide functionalized arms, a slight
excess was required to ensure sufficient coupling to yield
quantitatively coupled cores. As a result, a method of purifi-
cation must be optimized to isolate the desired star polymer
product from trace amounts (typically ꢃ5%) of unreacted
polymer arms. Many different technologies have been devel-
oped to purified polymers, such as fractionation,37 dialysis,44
GPC,45 preparative GPC,46 liquid chromatography at critical
condition,47 modified high performance liquid chromatogra-
phy,48 silica gel chromatography,27 solid resin as scavenger,49
and so forth. In this study, many of these have been investi-
gated to ascertain the most effective and high-throughput
technique for isolating ‘‘pure’’ star polymers. GPC and prepa-
rative GPC have demonstrated their ability to isolate the
pure star polymers but are only amenable to purifications
on small scales, therefore other techniques were sought. Di-
alysis is attractive for separating compounds of disparate
sizes (e.g., removal of residual monomer from a polymer)
but is less effective when there is a less substantial size dif-
ference. Even though the molecular weight increases sub-
stantially for the six-arm and 12-arm stars relative to the
‘‘one-arm’’ linear precursors, the more compact confirmation
of the resulting star polymers reduces their size substantially
in solution, complicating their separation by dialysis. In
theory, scavenging resins functionalized with complimentary
‘‘click’’ functionalities offer a relatively simple and appealing
method for removing unattached arms.50,51 This method was
investigated for a number of the crude ‘‘click’’ star-formation
reaction mixtures, however, proved ineffective at removing
all unreacted linear arms. For example during the synthesis
For the synthesis of the polystyrene three-arm star polymers,
the GPC trace of the crude product exhibit a major distribu-
tion that exhibited a shift to a larger hydrodynamic volume,
and only a small trace of residual ‘‘one-arm’’ linear polymer.
The MALDI-TOF mass spectra of the product exhibited an ap-
proximate threefold increase in the molecular weight (Table
4).
In addition, end group analysis of the product agrees with
the expected mass increase for three couplings to the tris-
alkyne core, for example, the expected exact mass for a three
arm polymer with 24 repeat unit per arm (n ¼ 72) is
8096.09 (including silver counterion), and
a signal is
observed at 8096.08. Alternatively, using previously reported
techniques, the ‘‘residual mass’’ of the ‘‘end groups’’ (all por-
tions of the star that cannot be accounted for by styrene
repeat units) can be calculated from the MALDI-TOF MS as
72.9 Da, which matches the theoretical value of 72.9 Da.
Furthermore, the quantitative reaction of the azides can be
TABLE 3 MALDI-TOF and GPC Characterization of Purified PEG Stars with Calculated Value in Parenthesis
Compound
PEG Stars
Mn by MALDI
PDI
Mn by GPC
PDI
Mru
Mres
10
14
15
16
PEG-N3
3-arm
2,110
1.01
1.01
1.01
1.01
2,700
1.02
1.03
1.04
1.12
44.0 (44.1)
44.1 (44.1)
44.0 (44.1)
N/A
13.0 (13.0)
2.8 (2.8)
13.7 (13.8)
N/A
6,600 (6,690)
13,600 (13,600)
29,000 (27,400)
9,600
6-arm
14,100
14,200
12-arm
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101
1093

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 4 Repeat Unit (Mru) and Residual Mass (Mres) Data for
Purified 3-Arm Star Polystyrene, 11
Purification by fractionation was an attractive alternative, as
it should be insensitive to incomplete end group functionali-
zations, however, often requires multiple inefficient precipita-
tions, and therefore a substantial reduction in yields. Using a
separatory funnel, the crude-arm star PS reaction mixture
was dissolved in toluene, and then methanol was added
dropwise until the solvent turned cloudy. The solution was
set overnight resulting in the formation of an oily layer in
the bottom of the funnel. Characterization by GPC and
MALDI TOF MS verified that the bottom layer was enriched
with the high-molecular weight portions of the crude reac-
tion mixture. After two fractionations, the GPC trace verified
the removal of the small amount the linear starting material,
Theoretical
(11þAg)þ
Observed
(11þAg)þ
Mru
Mres
n
(11n-11n ꢁ 1
)
for 11
66
67
68
69
70
71
72
73
74
75
76
7471.2
7575.3
7679.5
7783.6
7887.8
7991.9
8096.1
8200.2
8304.4
8408.5
8512.7
7470.9
7575.1
7679.2
7783.5
7887.6
7992.0
8096.1
8200.3
8304.6
8408.8
8513.1
Average
Theoretical
72.6
72.6
72.6
72.8
72.7
72.9
72.9
73.0
73.1
73.2
73.3
72.9
72.9
104.1
104.2
104.3
104.1
104.4
104.1
104.2
104.3
104.2
104.2
104.2
104.1
and pure star polymers were obtained with low PDI (Mn
8100; PDI ¼ 1.07; Fig. 5).
¼
The removal of the linear polymer distribution around m/z ¼
2600 in the MALDI-TOF mass spectra was particularly con-
vincing, as it is well known that lower molecular weight ana-
lytes exhibit much stronger signal in MALDI-TOF MS analysis.
Finally, although the well-defined one-arm distribution (MW
ꢃ2000–4000 with a monomer spacing of 104) has been
removed, close examination of the baseline shows a weak,
broad distribution (with unresolved monomer spacing) that
is calculated to have a Mn of about 3100. These observations
are consistent with triply charged polymer (Fig. 6).52
of the three-arm star polystyrene, a solid resin bearing
alkyne functional groups was added to the crude star poly-
mer products and the ‘‘click’’ conditions (Cu(I)Br/PMDETA)
were reapplied (Scheme 3).
Using analogous synthetic and purification procedures, the
six-arm and 12-arm polystyrene stars were also prepared.
The crude products of both exhibited a large increase in size
as judged by GPC and the expected increases in molecular
weight as calculated by their MALDI-TOF mass spectra (Figs.
5 and 6). In the case of the six-arm polymer, the observed
Mn was within 10% of the expected value (Mn theo ¼ 15,900,
Mn obs ¼ 15,000) and a similar trend was noted for the 12-
arm polymer (Mn theo ¼ 32,100, Mn obs ¼ 30,000). MS end
group analysis again confirmed that the signal distributions
agreed with those calculated, as further confirmed by the
observed residual masses (Supporting Information Table S3).
After purification by fractional precipitation, the six-arm and
12-arm star polymers exhibit a monomodal product with a
The GPC traces of the scavenger product showed only a
slight reduction in the linear impurities (Fig. 4).
Additional scavenging attempts on the same sample did not
seem to further reduce the amount of linear impurity. Simi-
lar results were observed for other PS and PEG stars, sug-
gesting that a trace amount of the linear polymer had inac-
tive end groups, thus preventing it from coupling to both the
multiarm cores and the solid phase resin during subsequent
attempts to scavenge. This was not unexpected considering
the likelihood of losing a trace amount of end group func-
tionality during ATRP, or in the case of PEG, the likelihood of
a trace amount of a side reaction (e.g., hydrolysis), during
the two-step end group transformation.
SCHEME 3 Purification of three-arm star PS by scavenging with alkyne functionalized solid phase resin.
1094
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
generally low (around 20%) and the technique very time-
consuming. For the PEG stars, preparative size exclusion
chromatographic separation using a packed column of ‘‘Bio-
Beads’’ proved to be a relatively fast and effective technique,
also afforded substantially improved yields.27 These were
generally performed on a 20–100 mg scale using THF as the
mobile phase, while collecting fractions in which the lower
molecular weight linear impurities eluted last. Parallel analy-
sis of each fraction by GPC and MALDI provide comprehen-
sive characterization (Fig. 7). In the early fractions, Figure
7(c) oligomeric impurities were observed in the GPC, which
were not observed in the crude precursor. It is presumed
that trace amounts of higher molecular weight impurities
only became visible because they were concentrated during
chromatographic purification, though the possibility of con-
tamination cannot be ruled out. The middle fractions exhibit
a very pristine three-arm star PEG, as judged both MALDI-
TOF MS and GPC, with the only trace amounts of low molec-
ular weight signal in the mass spectra confidently assigned
FIGURE 4 GPC spectrum of scavenged three-arm star polysty-
rene 7 and 11.
very narrow polydispersity both by GPC (six-arm PDI ¼
1.04, 12-arm PDI ¼ 1.05) and by MALDI-TOF MS (six-arm
PDI ¼ 1.01, 12-arm PDI ¼ 1.02). MALDI-MS analysis con-
firms the loss of the ‘‘one-arm’’ PS precursor, (Mn ꢃ2000–
4000) but again shows trace amounts of triply charged spe-
cies in a distinctly different mass range (e.g., for the six-arm
star MW ¼ 4500–6500 in Fig. 6). It should be noted that a
slight increase in the Mn of the purified product was
observed, most likely a result of the fractionation procedure
favoring the precipitation of higher molecular weight star
polymers. In addition, low MW analysis of the 12-arm prod-
uct was not possible, because the laser power required for
ionization of the high MW product resulted in substantial
matrix noise below m/z ¼ 15,000.
To access well-defined hydrophilic stars, the attachment of
mono-azide functionalized PEG to the same dendritic cores
was also explored. Using analogous click coupling conditions,
the PEG-N3 linear arms and the trialkyne core, were coupled
to generate three-arm star polymers. Before purification, the
GPC trace revealed a nearly monomodal product with the
expected increase in hydrodynamic volume, though the cal-
culated Mn values were inaccurate because they were cali-
brated against linear polystyrene standards (Mn ¼ 9660, PDI
¼ 1.03; Fig. 7 and Table 3).
The MALDI-TOF mass spectra exhibited a single distribution
with a Mn 6570, close to the calculated value of the expected
three arm star, 6680. Again, rigorous calibration using den-
dritic standards enables the accurate verification of each sig-
nal within the observed mass spectra. For example, the cal-
culated mass of the three-arm star polymer with exactly 138
ethylene glycol repeat units (corresponding to an average of
46 repeating units for each arm) when complexed to sodium
is 6633.8 Da, while the observed signal was 6633.6 Da (Ta-
ble 5).
Further verification across the entire distribution was
achieved by comparing the calculated residual mass of 2.8
with the theoretical value of 2.8. Despite the increased signal
strength for lower molecular weight analytes, no MS signals
were observed that corresponded to star polymers with one
or more failed click coupling reactions.
Although the fractionation approach enabled successful isola-
tion of the pure polystyrene star polymers, the yields were
FIGURE 5 GPC spectra of three-arm star PS (11), six-arms star
PS (12), and 12-arms star PS (13).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101
1095

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 6 MALDI-TOF mass spectrum of three-arm PS 11 (a), six-arm star PS 12 (b), and 12-arm star PS 13 (c).
to double and triple charged species. As expected, the late
fractions show a mixture of three-arm star and linear pre-
cursor, but again the linear precursor can be distinguished
from multiple charged species because of the clear 44.0 Da
(monomer mass) spacing between each signal.
Impurity of one, two-arm products was synthesized pur-
posely using equivalent of azido-PEG to tris-alkyne.
MALDI-TOF MS showed the residual mass matched the
expected by products. To confirm the ability to identify all
likely impurities during the coupling reaction, a coupling
2
FIGURE 7 MALDI-TOF and GPC characterization of three-arm star PEG fractions from ‘‘Bio-Bead’’ chromatography.
1096
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 5 Repeat Unit (Mru) and Residual Mass (Mres) Data for
Purified 3-Arm PEG, 14
Theoretical
(14þNa)þ
Observed
(14þNa)þ
Mru
Mres
n
(14n-14n ꢁ 1
)
for 14
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
6325.4
6369.5
6413.5
6457.6
6501.7
6545.7
6589.8
6633.8
6677.9
6721.9
6766.0
6810.0
6854.1
6898.1
6942.2
6986.2
7030.3
7074.3
6325.3
6369.2
6413.3
6457.3
6501.4
6545.4
6589.5
6633.6
6677.7
6721.7
6765.6
6810.0
6853.9
6897.9
6941.9
6986.1
7030.0
7074.2
Average
2.8
2.6
2.7
2.6
2.6
2.7
2.7
2.7
2.8
2.7
2.6
2.9
2.7
2.8
2.6
2.8
2.7
2.8
2.7
43.9
44.1
44.0
44.1
44.1
44.1
44.1
44.1
44.0
43.9
44.4
43.8
44.1
43.9
44.2
43.9
44.1
44.1
with an insufficient amount of linear ‘‘arms’’ (ꢃ2 equivalents
per three-arm core) was purposely performed. The resulting
mixture of one-, two- and three-armed stars was character-
ized by MALDI TOF MS to confirm the unambiguous identifi-
cation of each major component (Fig. 8 and Supporting In-
formation Tables S4 and S5).
FIGURE 9 GPC of three-arm star PEG 14, six-arms star PEG 15,
and 12 arms star PEG 16.
For one-arm impurity, the Mres was 20.9, close to the theo-
retical value of 21.0. The Mres of two-arm impurity was 33.9,
while the expected value is 34.0.
Using analogous techniques, the PEG six-arm and 12-arm
stars were prepared and purified. Characterization by GPC
and MALDI-TOF provided strong evidence for a quantitative
coupling of six-arm PEG, with the products after Biobead
FIGURE 8 MALDI-TOF spectrum of incompletely coupled stars
with only one arm (left), two arms (middle) and all three arms
(right) of PEG on the trisalkyne core, 2.
FIGURE 10 MALDI-TOF mass spectra of three-arm PEG 14 (a),
six-arm star PEG 15 (b), and 12-arm star PEG 16 (c).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101
1097

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 4 Synthesis of three- and six-arm star PCL-b-PEG 19 and 23.
FIGURE 11 MALDI-TOF MS spectrum of three-arm star PCL 17 and its alkynylated product 18.
purification exhibiting a narrow monomodal distribution in
its GPC trace (PDI ¼ 1.04) as well as its MALDI mass spec-
trum (PDI ¼ 1.01) [Figs. 9 and 10(b) and Supporting Infor-
mation Table S6].
hydroxyl end groups of 17 were esterified with pentynoic
acid anhydride to quantitatively functionalize each end with
the requisite alkyne for click coupling. The GPC trace of the
TABLE 6 MALDI-TOF Data for 3-Arm PCL-OH (17) and 3-Arm
PCL-Alkyne (18)
As predicted, the Mn calculated from the MALDI-TOF mass
spectra provides a more accurate value (Mn
¼ 13,600)
obs
than GPC (Mn obs ¼ 14,100) for the molecular weight of the
PEG six-arm star (Mn
¼ 13,600). The purified 12-arm
Theoretical
(17þNa)þ
Observed
(17þNa)þ
Theoretical
(18þNa)þ
Observed
(18þNa)þ
theo
n
star PEG also exhibited monomodal products based on the
GPC trace (Mn ¼ 14,200, PDI ¼ 1.12) [Figs. 9 and 10(c)].
The MALDI-TOF mass spectrum enabled the verification of
the quantitative formation of pure 12-arm star with Mn of
29,100 (Mn theo ¼ 27400) and PDI of 1.01 without evidence
of 11-arm or 10-arm impurities.53 Similar with the 12-arm
star PS, the MALDI-TOF mass spectra of 12-arm PEG stars
could not enable visualization of low-molecular weight range
due to matrix noise.
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
1511.885
1625.953
1740.022
1854.09
1511.914
1626.008
1740.087
1854.165
1968.236
2082.309
2196.379
2310.449
2424.515
2538.579
2652.646
2766.715
2880.779
2994.847
3108.909
3222.986
3337.049
3451.132
1751.964
1866.032
1980.1
1751.948
1866.037
1980.129
2094.195
2208.27
2094.168
2208.236
2322.304
2436.373
2550.441
2664.509
2778.577
2892.645
3006.713
3120.781
3234.849
3348.917
3462.985
3577.053
3691.121
1968.158
2082.226
2196.294
2310.362
2424.43
2322.338
2436.404
2550.472
2664.536
2778.601
2892.671
3006.739
3120.806
3234.866
3348.934
3463.006
3577.083
3691.154
Well-defined three-arm star block copolymers could be syn-
thesized in at least three different ways: diblock arms could
be grown divergently from a core molecule, preformed linear
diblocks could be coupled to a functionalized core, or homo-
polymer arms could be coupled to the ends of a functional-
ized three-arm star homopolymer.54 The last of these three
approaches was tested by coupling linear PEG to a PCL
three-arm star core (Scheme 4)
2538.498
2652.566
2766.634
2880.702
2994.77
3108.838
3222.907
3336.975
3451.043
The PCL three-arm star was prepared via polymerization of
e-caprolactone in bulk from 1,1,1-tris(hydroxymethyl)ethane
using Sn(Oct)2 catalysts at 130 ꢂC. Subsequently, the
1098
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
FIGURE 12 MALDI-TOF MS spectrum of six-arm star PCL 21 and its alkynylated product 22.
hydroxyl terminated star 17, showed a monomodal distribu-
tion with a narrow polydispersity and a Mn of 2230 when
corrected for PCL (4650 when calibrated against PS stand-
ards) The MALDI-TOF mass spectra of the hydroxyl termi-
nated polymer confirmed the synthesis of well-defined three-
arm star PCL without any measurable amount of linear poly-
mer byproducts that might occur from initiation from water
or other small molecule impurities [Fig. 11(a)].
cule. The six-arm star PCL was prepared via polymerization
of e-caprolactone in bulk using Sn(Oct)2 catalysts at 130 ꢂC.
1
From H NMR data, we can see all peak of –CH2OH group in
core initiator shifted from 3.70 ppm double peak to 4.03
ppm confirming that every hydroxyl group from the core ini-
tiate the polymerization of a PCL arm (Supporting Informa-
tion Figs. S5 and S9). The GPC shows a monomodal distribu-
tion with a narrow polydispersity (Mn ¼ 7760, PDI ¼ 1.11).
The MALDI-TOF mass spectra also confirm the synthesis of
well-defined six-arms PCL stars [Fig. 12(a)] and their quanti-
tative functionalization with pentynoate esters (480.48
increase in mass; Fig. 12 and Supporting Information Table
S7).
The observed repeat unit molecular weight (114.07) agreed
closely with the theoretical value (114.07) and the Mn calcu-
lated by MALDI-TOF MS (2220) agreed closely with the GPC
values. The identity of the core could also be confirmed by
MS calculations of the residual mass (6.1 Da), which agrees
closely with the theoretical value (6.0 Da). The attachment of
the three pentynoate esters via esterification with 4-penty-
noic acid anhydride could be confirmed by monitoring the
mass shifts of a particular signal in the MALDI-TOF MS, for
example, the 19-mer. The ion corresponding to the sodium
adduct of the hydroxyl functionalized star with a total of 19
repeat units of caprolactone has a theoretical exact molecu-
lar mass of 2310.4 Da (observed 2310.4 Da). On functionali-
zation with 3 pentynoate esters, the molecular weight for 18
is expected to increase by 240.04 Da (addition of 3 penty-
noic groups each weighing 80.01 Da), which yields a theoret-
ical molecular mass of 2550.4 (observed 2550.5) [Fig. 11(b)
and Table 6].
The alkynylated PCL star 18 was then reacted with 2 kDa
PEG-N3 10 under standard click conditions to yield the
three-arm star PCL-b-PEG 19. After purification, the GPC
trace revealed the higher molecular weight star block copoly-
mers were formed with narrow PDI (Mn ¼ 13,600, PDI ¼
1.03) (Fig. 13). The MALDI-TOF mass spectra exhibited a Mn
of 8970, close to calculated value of 9720. Detailed end
group analysis of the block copolymer was not possible,
however, because the non-equivalent masses of the different
monomers lead to a complex distribution in which individual
signals are no longer resolved (Fig. 14).
The six-arm PCL-PEG block copolymer was synthesized by
FIGURE 13 GPC of three-arm star and six-arm star PCL-b-PEG.
following the same procedure but using 20 as the core mole-
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101
1099

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
REFERENCES AND NOTES
1 Roovers, J. In Star and Hyperbranched Polymers; Marcel
Dekker: New York, 1998.
2 Quaglia, F.; Ostacolo, L.; Nese, G.; Canciello, M.; De, R.;
Ungaro, F.; Palumbo, R.; La, R.; Maglio, G. J. Biomed. Mater.
Res. Part A 2008, 87A, 563–574.
3 Aryal, S.; Prabaharan, M.; Pilla, S.; Gong, S. Int. J. Biol. Mac-
romol. 2009, 44, 346–352.
4 Poree, D. E.; Giles, M. D.; Lawson, L. B.; He, J.; Grayson, S.
M. Biomacromolecules 2011, 12, 898–906.
FIGURE 14 MALDI-TOF MS spectrum of three-arm (19) and
six-arms (23) star PCL-b-PEG.
5 Hadjichristidis, N.; Pitsikalis, M.; Iatrou, H. Macromol. Eng.;
Wiley-VCH Verlag GmbH & Co. KGaA, 2007; Vol. 2, pp 909–972.
6 Gao, H.; Matyjaszewski, K. Prog. Polym. Sci. 2009, 34, 317–350.
7 Wu, Z.; Liang, H.; Lu, J. Macromolecules 2010, 43, 5699–5705.
The alkynylated PCL star 22 was then reacted with 2 kDa
PEG-N3 10 under standard click conditions to yield the six-
arm star PCL-b-PEG. The GPC trace of the crude product con-
firmed a substantial increase in the molecular weight, as
well as a trace of unreacted PEG. After using Bio-bead
method to remove the unreacted PEG, a high pure six-arms
star polymer with narrow PDI (GPC: Mn ¼ 23,700, PDI ¼
1.07) was isolated (Fig. 13)
8 Wu, Z.; Liang, H.; Lu, J.; Deng, W. J. Polym. Sci. Part A:
Polym. Chem. 2010, 48, 3323–3330.
9 Hadjichristidis, N.; Pitsikalis, M.; Pispas, S.; Iatrou, H. Chem.
Rev. 2001, 101, 3747–3792.
10 Matyjaszewski, K.; Davis, T. P. Handbook of Radical Poly-
merization; Wiley: Hoboken, NJ, 2002.
11 Wang, J.-S.; Matyjaszewski, K. J. Am. Chem. Soc. 1995, 117,
5614–5615.
The MALDI-TOF mass spectra exhibited a Mn of 18,700, close
to theoretical value of 18,980 (Fig. 14).
12 Kato, M.; Kamigaito, M.; Sawamoto, M.; Higashimura, T.
Macromolecules 1995, 28, 1721–1723.
13 Grayson, S. M.; Frechet, J. M. Chem. Rev. 2001, 101,
3819–3868.
14 Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem. Int.
Ed. Engl. 2001, 40, 2004–2021.
CONCLUSIONS
15 Tornoe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem.
2002, 67, 3057–3064.
With the evolution of optimized dendronization methods and
living polymerization techniques, well-defined polymeric
components can be prepared with an exact number of func-
tional groups. The recent focus on highly efficient couplings,
chief among them the CuAAC coupling reaction, enables com-
plex architectures to be assembled from these components
yielding complex macromolecular architectures with unprec-
edented structural purity. In the case of star polymers, this
has been demonstrated with both hydrophobic and hydro-
philic arms attached via the ‘‘graft onto’’ method with dendri-
tic cores. Although many purification methods were investi-
gated, fractional precipitation and chromatographic
purification seem the most general, scalable, and successful
in removing residual linear starting materials. The excep-
tional purity of the stars were confirmed through detailed
characterization by both GPC and MALDI-TOF MS. In addi-
tion, this approach to making multiarm star polymers shows
promise toward making well-defined block copolymer stars
via the same click route. The ability to synthesize well-
defined star polymers and confirm their purity is critical to
advancing their biomedical applications.
16 Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B.
Angew. Chem. Int. Ed. Engl. 2002, 41, 2596–2599.
17 Lahann, J. Click Chemistry for Biotechnology and Material
Science; Wiley: Chickester, UK, 2009.
18 Golas, P. L.; Matyjaszewski, K. Chem. Soc. Rev. 2010, 39,
1338–1354.
19 Iha, R. K.; Wooley, K. L.; Nystrom, A. M.; Burke, D. J.; Kade,
¨
M. J.; Hawker, C. J. Chem. Rev. 2009, 109, 5620–5686.
20 Fournier, D.; Hoogenboom, R.; Schubert, U. S. Chem. Soc.
Rev. 2007, 36, 1369–1380.
21 Helms, B.; Mynar, J. L.; Hawker, C. J.; Frechet, J. M. J. J.
Am. Chem. Soc. 2004, 126, 15020–15021.
22 Laurent, B. A.; Grayson, S. M. J. Am. Chem. Soc. 2011, 133,
13421–13429.
23 Dirks, A. J.; Van, B.; Hatzakis, N. S.; Opsteen, J. A.; Van, D.;
Cornelissen, J. J. L. M.; Rowan, A. E.; Van, H.; Rutjes, F. P. J.
T.; Nolte, R. J. M. Chem. Commun. (Cambridge, U. K.) 2005,
4172–4174.
24 Li, M.; De, P.; Gondi, S. R.; Sumerlin, B. S. Macromol. Rapid
Commun. 2008, 29, 1172–1176.
25 Gao, H.; Matyjaszewski, K. Macromolecules 2006, 39,
The authors acknowledge the National Institute of Health
(EB006493), for financial support of this research, and the
National Science Foundation (MRI-0619770) is acknowledged
for enabling MALDI-TOF MS characterization. In addition, this
work was supported by the U.S. Army Medical Research and
Materiel Command under Award No. W81XWH-10-1-0377.
Opinions, interpretations, conclusions, and recommendations
are those of the author and are not necessarily endorsed by the
U.S. Army.
4960–4965.
26 Altintas, O.; Yankul, B.; Hizal, G.; Tunca, U. J. Polym. Sci.
Part A: Polym. Chem. 2006, 44, 6458–6465.
27 Hoogenboom, R.; Moore, B. C.; Schubert, U. S. Chem. Com-
mun. 2006, 4010–4012.
28 Xu, J.; Liu, S. J. Polym. Sci. Part A: Polym. Chem. 2009, 47,
404–419.
29 Gungor, E.; Bilir, C.; Durmaz, H.; Hizal, G.; Tunca, U.
J. Polym. Sci. Part A: Polym. Chem. 2009, 47, 5947–5953.
1100
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
30 Gungor, E.; Bilir, C.; Hizal, G.; Tunca, U. J. Polym. Sci. Part
A: Polym. Chem. 2010, 48, 4835–4841.
43 Hoskins, J. N.; Grayson, S. M. Macromolecules 2009, 42,
6406–6413.
31 Chan, J. W.; Yu, B.; Hoyle, C. E.; Lowe, A. B. Chem. Com-
mun. 2008, 4959–4961.
44 Parrish, B.; Breitenkamp, R. B.; Emrick, T. J. Am. Chem.
Soc. 2005, 127, 7404–7410.
32 Dag, A.; Durmaz, H.; Tunca, U.; Hizal, G. J. Polym. Sci. Part
A: Polym. Chem. 2009, 47, 178–187.
45 Narasimhan, V.; Burns, C. M.; Huang, R. Y. M.; Lloyd, D. R.
Adv. Chem. Ser. 1984, 206, 3–15.
33 Meier, M. A. R.; Gohy, J.-F.; Fustin, C.-A.; Schubert, U. S.
J. Am. Chem. Soc. 2004, 126, 11517–11521.
46 Ekmanis, J. L. ACS Symp. Ser. 1987, 352, 47–58.
47 Hunkeler, D.; Macko, T.; Berek, D. ACS Symp. Ser. 1993,
521, 90–102.
34 Kul, D.; Van Renterghem, L. M.; Meier, M. A. R.; Strandman,
S.; Tenhu, H.; YIlmaz, S. S.; Schubert, U. S.; Du Prez, F. E.
J. Polym. Sci. Part A: Polym. Chem. 2008, 46, 650–660.
48 Park, S.; Park, I.; Chang, T.; Ryu, C. Y. J. Am. Chem. Soc.
2004, 126, 8906–8907.
ꢀ
35 Ihre, H.; Padilla De Jesu´ s, O. L.; Frechet, J. M. J. J. Am.
Chem. Soc. 2001, 123, 5908–5917.
49 Leeb, L.; Gmeiner, P.; Loeber, S. QSAR Comb. Sci. 2007, 26,
1145–1150.
36 Malkoch, M.; Schleicher, K.; Drockenmuller, E.; Hawker, C.
J.; Russell, T. P.; Wu, P.; Fokin, V. V. Macromolecules 2005, 38,
3663–3678.
50 Urbani, C. N.; Bell, C. A.; Lonsdale, D. E.; Whittaker, M. R.;
Monteiro, M. J. Macromolecules 2007, 40, 7056–7059.
37 Knauss, D. M.; Huang, T. Macromolecules 2002, 35, 2055–2062.
51 Clark, P. G.; Guidry, E. N.; Chan, W. Y.; Steinmetz, W. E.;
Grubbs, R. H. J. Am. Chem. Soc. 2010, 132, 3405–3412.
38 Ihre, H.; Johansson, M.; Malmstroem, E.; Hult, A. Advances
in Dendritic Macromolecules; JAI Press: Stamford, CT, 1996,
pp 1–25.
52 Although the presence of the double charged species is
also expected, the overlap with the low MW portions of the
singly charged polymer makes this distribution difficult to
resolve.
39 Li, Y.; Hoskins, J. N.; Sreerama, S. G.; Grayson, M. A.; Gray-
son, S. M. J. Mass Spectrom. 2010, 587–611.
40 The use of excess sodium azide requires care during work up,
and disposal of waste, in addition to avoiding use of halogenated
solvents which may react with azide to form explosive byproducts.
53 Because of the low ionization potential in this molecular
weight range, increased laser power is required to provide suf-
ficient signal, and substantial low molecular weight noise is
observed preventing accurate examination of lower mass
ranges in the same spectra.
41 Li, Y.; Hoskins, J. N.; Sreerama, S. G.; Grayson, S. M. Mac-
romolecules 2010, 43, 6225–6228.
42 Li, Y.; Giles, M. D.; Liu, S.; Laurent, B. A.; Hoskins, J. N.;
Cortez, M. A.; Sreerama, S. G.; Gibb, B. C.; Grayson, S. M.
Chem. Commun. 2011, 47, 9036–9038.
54 Lammens, M.; Fournier, D.; Fijten, M. W. M.; Hoogenboom,
R.; Prez, F. D. Macromol. Rapid Commun. 2009, 30, 2049–2055.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE PART A: POLYMER CHEMISTRY 2012, 50, 1086–1101
1101