A divergent route towards single-chemical entity triazine dendrimers with opportunities for structural diversity

Article abstract of DOI:10.1039/b617875h

This manuscript describes the successful synthesis and characterization of five generations of dendrimers based on melamine. Early generations of these materials appear to be single chemical entities: upon purification, no detectable impurities are observed using NMR spectroscopy, mass spectrometry, and HPLC and GPC analysis. The analysis of larger generation materials precludes an unambiguous statement of purity. The synthetic route to these targets is divergent, relying on dichlorotriazine monomers that react with a polyamine dendrimer core of generation n to produce a poly(monochlorotriazine) dendrimer of generation n + 1. Subsequently, the poly(monochlorotriazine) is derivatized through nucleophilic aromatic substitution before additional nucleophilic amines are unmasked and the process iterated. The Royal Society of Chemistry and the Centre National de la Recherche Scientifique.

Full text of DOI:10.1039/b617875h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
A divergent route towards single-chemical entity triazine dendrimers with
opportunities for structural diversitywz
Hannah Crampton, Emily Hollink, Lisa M. Perez and Eric E. Simanek*
Received (in Montpellier, France) 7th December 2006, Accepted 24th January 2007
First published as an Advance Article on the web 2nd March 2007
DOI: 10.1039/b617875h
This manuscript describes the successful synthesis and characterization of five generations of
dendrimers based on melamine. Early generations of these materials appear to be single chemical
entities: upon purification, no detectable impurities are observed using NMR spectroscopy, mass
spectrometry, and HPLC and GPC analysis. The analysis of larger generation materials precludes
an unambiguous statement of purity. The synthetic route to these targets is divergent, relying on
dichlorotriazine monomers that react with a polyamine dendrimer core of generation n to produce
a poly(monochlorotriazine) dendrimer of generation n + 1. Subsequently, the
poly(monochlorotriazine) is derivatized through nucleophilic aromatic substitution before
additional nucleophilic amines are unmasked and the process iterated.
The divergent approach employed here overcomes the
Introduction
limitations of our convergent route. This route relies on the
Dendrimers have received widespread attention for potential
applications ranging from medicine to materials science.1–4
iterative incorporation of dichlorotriazine 1 bearing two BOC-
protected nucleophilic amines onto a poly(amine) dendrimer
The field benefits greatly from the diversity of scientists and
of generation n (Scheme 1). Upon installation of the nucleo-
engineers engaged in these studies. This diversity is predicated
philic group (RNuH) and deprotection, the cycle can be
on the commercial availability of these materials, notably
repeated. Here, we report the successful synthesis and char-
PAMAM, PMMH and PPI.5–7 Availability relies on synthetic
acterization of these architectures through generation five. The
routes to these materials that are efficient in terms of yield,
structural diversity referred to is reflected in the surface BOC
costs of materials, and process conditions. To date, each
and piperidine groups of each dendrimer, though clearly, the
dendrimer displays only one type of surface group, although
latter cannot be manipulated further.
the identity of this group is dependent on architecture. Our
own interests include contributing a dendrimer scaffold with
groups on the surface that can be orthogonally manipulated.
Results and discussion
Recently, we reported on a divergent synthetic strategy that
allowed access to a variety of macromolecular architectures
including stars, dendrimers and hybrid molecules by exploiting
the differential reactivity of dichlorotriazines.8 To us, this
represents a significant advance over our previous convergent
routes. While both routes rely on the stepwise nucleophilic
aromatic substitution of chlorotriazines, the convergent ap-
proach had practical limitations.9 Of lesser issue, the macro-
molecule’s constitution was dictated at the commencement of
synthesis with the choice of surface groups. This limitation was
offset by carrying latent functionality (in the form of ortho-
gonal protecting groups) through the synthesis that could be
manipulated.10,11 Of greater issue, the repetitious coupling of
smaller fragments into larger ones required that most of the
mass be carried through the entire synthesis. Regardless,
reasonable quantities of materials of high purity up to gen-
eration three could be obtained for evaluation.
Synthesis
The dichlorotriazine monomer, 1, used iteratively throughout
this synthesis satisfies many demands: it is prepared in high
yields in two steps, it is readily purified, and it is a stable
crystalline solid. Indeed, methods for the kilogram-scale pre-
paration of this material will be described in due course.
Monomer 1 undergoes nucleophilic aromatic substitution
cleanly to the poly(monochlorotriazine). Scheme 2 outlines
the synthetic route adopted for these materials.
Here, we carry piperidine groups through the synthesis
under the assumption that these groups (i) promote greater
Department of Chemistry, Texas A&M University, College Station,
TX 77845, USA. E-mail: simanek@tamu.edu
w This paper was published as part of the special issue on Dendrimers
and Dendritic Polymers: Design, Properties and Applications.
z Electronic supplementary information (ESI) available: NMR and
mass spectra. See DOI: 10.1039/b617875h
Scheme 1 General synthetic route.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 1283–1290 | 1283

View Article Online
Scheme 2 Synthesis of G1–G5 dendrimers. Reagents and conditions: (a) 1, DIPEA, THF, 0–25 1C, 15 h; (b) piperidine, THF, 25 1C, 12–16 h;
(c) 5 M HCl–MeOH, 25 1C, 15 h; (d) 1, DIPEA, THF–water, 0–25 1C, 15 h; (e) 1, DIPEA, EtOAc–THF–water, 25 1C, 16 h; (f) 1, DIPEA, CHCl3,
25 1C, 5 d.
solubility than the aminoalcohols used in the preliminary
report, and (ii) undergo facile reactions with monochlorotria-
zines. The nomenclature adopted identifies the generation, Gn,
and the functional group of greatest interest as Cl, Pip
(piperidine) and NH2. The iterative nature of the synthesis
proceeds from the core, 2, through G1-Cl upon reaction with
monomer 1, through G1-Pip upon capping the poly(mono-
chlorotriazine) with piperidine, to G1-NH2 upon deprotection.
Iterative reaction with 1 yields G2-Cl, and the cycle of reac-
tions continues.
Characterization of these materials rests on analysis using
thin layer chromatography (TLC), NMR spectroscopy, mass
spectrometry, HPLC and GPC. While exact estimates of
purity for higher generation materials are imprecise, we are
confident that these materials are as close (and in many cases
closer) to single-chemical entity status as other dendrimers
described in the literature. We believe that lower generation
materials are pure (497%) as defined by the conventional
definition of synthetic organic chemistry.
Characterization
One advantage of this synthetic sequence is the set of
universal conditions employed for both the piperidine step
and deprotection. Throughout the course of the synthesis,
both of these reactions proceed quantitatively. These condi-
tions are described explicitly in the experimental section. The
problematic step of this relatively straightforward synthesis is
iteration with 1. Under the conditions currently employed, the
yields for this step decrease with each successive generation. In
addition, solubility of the Gn-NH2 intermediates require that
the solvent be reconsidered at every generation. We find that
moving from tetrahydrofuran to methylene chloride and
ultimately to chloroform provides the best results and the
cleanest reactions. Solvents that proved less useful include
dimethylformamide, dimethylacetamide, dimethyl sulfoxide,
methanol, water, tetrahydrofuran/water with and without
sodium dodecyl sulfate, and the ionic liquid octylmethylimi-
dazolium tetrafluoroborate.
Characterization of the Gn-NH2 species. The polarity of the
polyamine cores makes TLC analysis problematic. Both full
and partially deprotected dendrimers adhere to silica and
1
cannot be separated. H NMR spectroscopy is quite diagno-
stic, especially for lower generations where solubility is not a
problem: it reveals both the loss of the tert-butyl protons of
the BOC groups, as well as the significant upfield shift of the
CH2NHBOC peak upon deprotection (from 3.07 to 2.68
ppm). Mass spectrometry confirms the purity of these dendri-
mers through G4. While G5-NH2 has been isolated, character-
ization of this material by mass spectrometry has proven
unsuccessful to date.
Characterization of the Gn-Cl species. TLC reveals a sig-
nificant shift in polarity upon complete substitution of the
polyamine core. Unfortunately, and presumably attributed to
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
1284 | New J. Chem., 2007, 31, 1283–1290

View Article Online
the polarity of the amine intermediates, we cannot follow the
reaction progress by TLC. That is, intermediates derived from
a substoichiometric number of reactions with monomer 1
cannot be separated from starting polyamine. NMR spectro-
scopy is of limited use due to the broad and overlapping lines.
However, the CH2NHBOC peak stands alone at B3 ppm in
each generation, and can be integrated with respect to the area
of peaks from B3.8–3.2 ppm that correspond to methylenes a
to nitrogen derived from cyclic and acyclic groups (Fig. 1). By
defining the area of the former methylenes, the expected and
observed integrations can suggest the extent of reaction. In
Fig. 1, the expected value for the downfield region is shown in
parentheses for the latter protons, while the observed integra-
tion is reported above.
Mass spectrometry confirms at isotopic resolution that
substitution is complete in lower generations: a single parent
line is seen for each dendrimer, and no lines corresponding to
truncated intermediates are seen. The traces for G4-Cl and G5-
Cl lack this level of resolution. Fig. 2 shows the trace for G4-Cl
with a molecular ion peak that shows incremental losses of
BOC as a ‘‘tailing’’ to lower m/z values. A similar pattern is
evident from the doubly-charged M2+ ion. We believe that
these loss-of-BOC species arise during the MALDI-MS ioni-
Fig. 2 The MALDI-TOF MS spectrum of G4-Cl reveals the limita-
tions of characterization. The tailing towards lower m/z values reflects
loss of BOC (presumably during ionization) and obscures evidence of
incomplete reaction.
zation event: previous studies wherein the voltage was changed
affected the relative populations of these species.
The Gn-Cl species were subjected to HPLC and GPC
analysis. In either case, G5-Cl did not elute from the columns.
However, G1-Cl through G3-Cl showed sharp HPLC traces
and symmetric GPC traces (Fig. 3). Both the HPLC and GPC
traces of G4-Cl show tailing that could be attributed to
impurities of similar polarity and dimension. Accordingly,
we cannot confidently ascribe these materials as single-chemi-
cal entities using the accepted definitions of organic chemistry.
However, this material does represent a purity that is rarely
seen in synthetic macromolecules.
Characterization of the Gn-pip species. The substitution of
chloride by piperidine does not have an appreciable effect on
the polarity of these dendrimers, particularly at higher gen-
erations, making it impossible to follow reaction progress by
TLC. Analysis by NMR spectroscopy of dendrimers beyond
generation one is hindered by broad and overlapping lines, but
as with the Gn-Cl dendrimers, the CH2NHBOC peak can be
Fig. 1 1H NMR spectra of the aliphatic regions of the Gn-Cl targets.
The integration of the upfield peak is assigned the theoretically
expected value and compared to the observed and (expected) values
for the downfield signals. The integration region is indicated with
arrows.
Fig. 3 GPC traces of Gn-Cl (n = 1–4) and a blank.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 1283–1290 | 1285

View Article Online
integrated with respect to the group at 3.8–3.2 ppm, which
contains the piperidine a-CH2 peak. The mass spectra of
G1-Pip through G4-Pip confirm that substitution is complete:
the trace from G5-Pip is too broad to interpret.
reported in angstroms for these targets corroborate earlier
investigations.
Conclusions
Computation
We report the synthesis and characterization of five genera-
tions of dendrimers based on melamine. The divergent route
employed appears to yield single-chemical entities through
generation four. Establishing the purity of fourth and fifth
generation materials remains challenging due to the current
reliance on NMR spectroscopy: mass spectrometry provides a
broad peak centered appropriately and HPLC and GPC
analysis have not been particularly successful, but corroborate
the assertion that in many cases, the purity of these materials
competes favorably with what is commercially available.
These targets are available in excellent yields at lower genera-
tions and modest yield at higher generations. Purification has
proven straightforward. Extending these reactions to larger
generation dendrimers will require additional investigations of
reaction conditions, or the substitution of piperidine with
some other capping group.
Computational modeling of these architectures reveals the
onset of globular structure at generation three as indicated
by localization of the core, shown in maroon, to the interior of
the dendrimer (Fig. 4). The three arms of each dendrimer that
eminate from the core are colored red, green and blue. Distinct
surface domains emerge at generation four and are pro-
nounced in generation five. Smaller generations are not suffi-
ciently encumbered sterically to completely allow segregation
of end groups into specific regions of space. The dimensions
The value of these investigations are in establishing syn-
thetic routes that combine the purity usually associated with
convergent syntheses and the efficiency associated with diver-
gent routes. Of perhaps greatest importance is the potential for
diversity in this system. One can imagine adding a different
nucleophile at each generation during the capping step, where
here we use piperidine. This potential for diversity, along with
the purity of these materials, makes them particularly good
candidates for a variety of uses, including drug delivery.
Experimental
All solvents and reagents were purchased from Aldrich Che-
mical Co. or Acros Organics and were used without purifica-
tion. The reagent HN(CH2CH2CH2NHBoc)2 was prepared by
a modification of previously reported method.12 The core 2
was reported previously.13
Computation
Computational results were obtained using the software pack-
age Cerius2 4.9 by Accelrys Inc. Minimization and dynamics
calculations were performed with the Open Force Field (OFF)
program, using the pcff second-generation force field.14 The
dendrimer was initially drawn and minimized in a fully
extended conformation. Constant volume and temperature
(NVT) molecular dynamics (MD) calculations were then
performed on the minimized structure via simulated annealing.
The simulated annealing was carried out for 560.0 ps, over a
temperature range of 300–1000 K, with DT = 50 K, using the
T-Damping temperature thermostat, a relaxation time of
0.1 ps, and a time step of 0.001 ps. The dendrimer was
minimized after each annealing cycle, resulting in 200 struc-
tures.
Characterization
Thin-layer chromatography was performed using EMD silica
gel 60 F254 pre-coated glass plates (0.25 mm). Preparative
column chromatography was performed using EMD silica gel
Fig. 4 Computation reveals the onset of globular structure at gen-
eration 3. The core is shown in maroon. The dimensions are in
angstroms.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
1286 | New J. Chem., 2007, 31, 1283–1290

View Article Online
1
60 (0.040 mm particle size). H and 13C{1H} NMR data were
The solution was washed with water (3 ꢂ 300 mL), and then
dried with MgSO4. Following filtration, the solvent was
removed in vacuo. The product was obtained as a pure white
solid by reprecipitation with hexanes from a clear solution of
acquired on a Varian 300 MHz spectrometer at 25 1C unless
otherwise indicated. 1H and 13C{1H} NMR chemical shifts are
listed relative to tetramethylsilane in parts per million, and
were referenced to the residual proton or carbon peak of the
solvent. MS analyses were performed by the Laboratory for
Biological Mass Spectrometry at Texas A&M University.
MALDI-TOF mass spectra were obtained on an Applied
Biosystems voyager-DE STR Biospectrometry workstation.
Samples were diluted to 0.1 mg mlꢁ1 and mixed with 2,4,6-
trihydroxyacetophenone at 20 mg mlꢁ1 in a 1 : 5 analyte :
matrix ratio. HPLC analyses were performed on a Waters
Delta 600 system with a Waters 2487 dual wavelength absor-
bance detector at 240 nm. A Waters Symmetry C18 silica-based
RP-HPLC column (4.6 ꢂ 250 mm, 5 mm, 100 A) was used with
a mobile phase of 70 : 30 acetonitrile–THF at a flow rate of 1
mL minꢁ1. Injection volumes were 40 mL at a concentration of
0.5 mg mLꢁ1. Analyses were performed using Empower Pro
software. GPC analyses were performed using a Viscotek
VE3210 UV/Vis detector at 260 nm and 30 1C. A Visco-
GEL mixed bed I-MBMMW-3078 GPC column (7.8 mm ꢂ
30 cm) was used with a mobile phase of THF at a flow rate of 1
mL minꢁ1. Injection volumes were 100 mL at a concentration
1
EtOAc. Yield: 11.64 g (87%). H NMR (300 MHz, CDCl3) d
3
5.02 (br, 2H, NH), 3.61 (t, JHꢁH = 7 Hz, 4H, NCH2), 3.12
3
(m, 4H, CH2NHBoc), 1.77 (p, JHꢁH
= 7 Hz, 4H,
NCH2CH2), 1.43 (s, 18H, C(CH3)3). 13C{1H} NMR (75.5
MHz, CDCl3) d 170.1 (s, C3N3), 164.7 (s, C3N3), 156.0 (s,
C(O)), 79.3 (s, C(CH3)3), 44.9 (s, NCH2), 37.3 (s,
CH2NHBoc), 28.3 (s, C(CH3)3), 27.7 (s, NCH2CH2). MS
(ESI): calc. 478.1862 (M+); found 479.1992 (M + H+).
G1-Cl. Solutions of 2 (1.21 g, 3.63 mmol) and 1 (6.09 g, 12.7
mmol) were dissolved separately in THF (100 mL each) to give
a slurry and a clear solution, respectively. DIPEA (6.5 mL, 36
mmol) was added to the solution of C3N3(piperazine)3, and
then both solutions were cooled to 0 1C. The mixture was
added dropwise to the solution of 1, and the slurry was stirred
at 0 1C for 1 h. The solution was warmed gradually to 25 1C
and stirred for an additional 15 h. The mixture was filtered
through Celite, and then the solvent was removed in vacuo.
The residue was dissolved in CH2Cl2 (150 mL), and this
solution was washed with water (3 ꢂ 200 mL). The organic
phase was dried with MgSO4. Excess 1 was removed by
filtration of the organic phase through a silica plug. The
product was reprecipitated from CH2Cl2 with hexanes to
afford a white solid. Yield: 5.63 g (93%). 1H NMR (300
MHz, CDCl3) d 5.59 (br, 3H, NH), 4.83 (br, 3H, NH), 3.82
(br, 24H, CH2, piperazine), 3.56 (m, 12H, pr-NCH2), 3.08 (m,
12H, CH2NHBoc), 1.73 (m, 12H, NCH2CH2), 1.424 (s, 27H,
C(CH3)3), 1.420 (s, 27H, C(CH3)3). 13C{1H} NMR (75.5 MHz,
CDCl3) d 169.2 (s, C3N3), 165.2 (s, C3N3), 164.9 (s, C3N3),
164.2 (s, C3N3), 156.1 (s, C(O)), 155.8 (s, C(O)), 79.2 (s,
C(CH3)3), 78.8 (s, C(CH3)3), 43.8 (s, CH2), 43.3 (s, CH2,
piperazine), 42.9 (s, CH2), 42.6 (s, CH2), 37.7 (s, CH2NHBoc),
36.7 (s, CH2NHBoc), 28.4 (s, C(CH3)3), 28.3 (s, C(CH3)3), 27.8
(s, NCH2CH2), 27.7 (s, NCH2CH2). MS (MALDI): calc.
1662.2559 (M+); found 1663.0484 (M + H+).
of 1 mg mLꢁ1
.
Syntheses
Monomer 1. Monomer 1 was prepared in two steps from a
commercially available triamine. Protection of the triamine to
yield HN(CH2CH2CH2NHBoc)2 commences with a solution
of HN(CH2CH2CH2NH2)2 (5.35 mL, 37.6 mmol) and DIPEA
(17.9 mL, 105 mmol) dissolved in 80 mL of THF. BOC-ON
(18.5 g, 75 mmol) was dissolved separately in 200 mL of THF.
Both solutions were cooled to 0 1C and then combined. The
mixture was warmed gradually to 25 1C and stirred for 4 h.
The solvent was removed in vacuo, and then the residue was
dissolved in 200 mL of CH2Cl2. This solution was washed with
water, 1 M NaOH (aq.) solution and brine (3 ꢂ 150 mL each).
The organic phase was dried with MgSO4. Following filtra-
tion, the solvent was removed in vacuo. The product was
purified by reprecipitation from a solution of hexanes and
trace methanol to yield a white solid. Yield: 10.44 g (84%). 1H
G1-Pip. A clear solution of G1-Cl (2.53 g, 1.52 mmol) and
piperidine (1.52 mL, 15.2 mmol) was prepared in THF (50
mL). The mixture was stirred at 25 1C for 16 h, and the
resulting slurry was filtered through silica to give a clear
solution. The solvent was removed in vacuo. The residue was
dissolved in CH2Cl2 (20 mL), and this solution was washed
with water (3 ꢂ 50 mL). The organic phase was dried with
Na2SO4. Following filtration, the solvent was removed in
vacuo to afford the product as a white powder (yield: 2.79 g,
3
NMR (300 MHz, CDCl3) d 5.19 (br, 2H, NH), 3.18 (t, JH–H
= 6 Hz, 2H, CH2NHBoc), 3.16 (t, JHꢁH = 6 Hz, 2H,
CH2NHBoc), 2.61 (t, JHꢁH = 6 Hz, 4H, NCH2), 1.61 (qd,
3
3
4JHꢁH = 6 Hz, 2JHꢁH = 6 Hz, 4H, NCH2CH2), 1.41 (s, 18H,
C(CH3)3). 13C{1H} NMR (75.5 MHz, CDCl3) d 156.1 (s,
C(O)), 78.9 (s, C(CH3)3), 47.2 (s, NCH2), 38.7 (s,
CH2NHBoc), 29.6 (s, NCH2CH2), 28.3 (s, C(CH3)3). MS
(ESI): calc. 331.2471 (M+); found 332.2490 (M + H+).
To obtain 1, a solution of cyanuric chloride (5.38 g, 29.2
mmol) in THF (200 mL) was cooled to 0 1C. A clear solution
of HN(CH2CH2CH2NHBoc)2 (9.21 g, 27.8 mmol) in THF
(100 mL) was added dropwise to the cyanuric chloride solu-
tion, followed by dropwise addition of a solution of DIPEA
(4.49 mL, 27.8 mmol) in THF (150 mL). The solution was
stirred at 0 1C for 1 h, then warmed gradually to 25 1C and
stirred for an additional 12 h. The solvent was removed in
vacuo, and then the residue was taken up in CH2Cl2 (200 mL).
1
499%). H NMR (300 MHz, CDCl3) d 5.27 (br, 6H, NH),
3.80 (br, 24H, CH2, piperazine), 3.72 (br, 12H, C5H10N, a-H),
3.96 (br, 12H, CH2, pr-NCH2), 3.07 (br, 12H, CH2NHBoc),
1.70 (br, 12H, NCH2CH2), 1.62 (br, 6H, C5H10N, g-H), 1.55
(br, 12H, C5H10N, b-H), 1.42 (s, 54H, C(CH3)3). 13C{1H}
NMR (75.5 MHz, CDCl3) d 166.1 (s, C3N3), 165.5 (s, C3N3),
165.1 (s, C3N3), 156.2 (s, C(O)), 79.2 (s, C(CH3)3), 44.4 (s,
CH2, piperazine), 43.4 (s, pr-NCH2), 43.3 (s, pr-NCH2), 42.0
(C5H10N, a-C), 37.4 (s, CH2NHBoc), 37.3 (s, CH2NHBoc),
28.7 (s, C(CH3)3), 27.9 (s, NCH2CH2), 26.0 (C5H10N, b-C),
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 1283–1290 | 1287

View Article Online
25.1 (C5H10N, g-C). MS (MALDI): calc. 1807.2050 (M+);
as a white powder. Yield: 3.49 g (499%). 1H NMR (300
MHz, CDCl3) d 6.70–5.49 (br, 18H, NH), 3.79 (br, 24H, CH2,
piperazine), 3.68 (br, 36H, C5H10N, a-H), 3.53 (br, 36H, CH2,
NCH2), 3.34 (br, 12H, CH2, CH2NH–C3N3), 3.02 (br, 24H,
CH2NHBoc), 1.80 (br, 12H, G1-NCH2CH2), 1.66 (br, 24H,
G2-NCH2CH2), 1.58–1.52 (br, 54H, C5H10N, g-C5H10N, b-
H), 1.39 (s, 108H, C(CH3)3). 13C{1H} NMR (75.5 MHz,
CDCl3) d 165.9 (s, C3N3), 165.6 (s, C3N3), 165.2 (s, C3N3),
165.0 (s, C3N3), 156.0 (s, C(O)), 79.0 (s, C(CH3)3), 44.3 (br),
43.3 (br), 42.2 (br), 37.3 (s, CH2NHBoc), 28.7 (s, C(CH3)3),
27.9 (br, NCH2CH2), 26.0 (br, C5H10N, b-C), 25.1 (br,
C5H10N, g-C). MS (MALDI): calc. 4157.3818 (M+); found
4158.6648 (M + H+).
found 1808.3271 (M + H+).
G1-NH2. A 5 M HCl solution (35 mL) was added to a clear
solution of G1-Pip (2.75 g, 1.52 mmol) in CH2Cl2 (3 mL) and
MeOH (70 mL), and the solution was stirred at 25 1C for 15 h.
The volatile components were concentrated in vacuo until only
ca. 15 mL of water remained. The residue was made basic (pH
= 14) with 40 mL of 5 M NaOH (aq.) solution. The resulting
milky suspension was extracted with CHCl3 (5 ꢂ 250 mL). The
combined organic phases were dried with Na2SO4. Following
filtration, the solvent was removed in vacuo to afford the
product as a white solid. Yield: 1.845 g, 499%. 1H NMR
(300 MHz, CDCl3) d 3.79 (br, 24H, CH2, piperazine), 3.71 (br,
3
12H, C5H10N, a-H), 3.63 (br t, JHꢁH = 7 Hz, 12H, pr-
G2-NH2. A concentrated aqueous HCl solution (15 mL)
was added to a solution of G2-Pip (1.35 g, 0.32 mmol) in
MeOH (40 mL), and the clear solution was stirred at 25 1C for
16 h. The volatile components were concentrated in vacuo until
only ca. 5 mL of water remained. The residue was made basic
(pH = 14) with 25 mL of 1 M NaOH (aq.) solution, and the
resulting milky suspension was extracted with CHCl3 (5 ꢂ 250
mL). The organic extractions were combined, and then the
solvent was removed in vacuo to afford the product as a white
solid. Yield: 959 mg, 499%. 1H NMR (300 MHz, CDCl3 with
trace CD3OD) d 5.24 (br, 6H, NH), 3.80 (br, 24H, CH2,
piperazine), 3.73 (br, 12H, G1-C5H10N, a-H), 3.65 (br, 36H,
G2-C5H10N, a-H, NCH2), 3.57 (br, 24H, CH2, NH2NCH2),
3.32 (br, 12H, CH2NH–C3N3), 2.64 (br, 24H, CH2NH2), 1.81
(br, 12H, G1-NCH2CH2), 1.70 (br, 24H, G2-NCH2CH2),
1.58–1.51 (br, 54H, C5H10N, g-H, C5H10N, b-H, and NH2).
13C{1H} NMR (75.5 MHz, CDCl3 with trace CD3OD) d 165.8
(s, C3N3), 165.4 (s, C3N3), 165.2 (s, C3N3), 165.1 (s, C3N3),
165.0 (s, C3N3), 164.6 (s, C3N3), 164.3 (s, C3N3), 43.8 (s,
NCH2), 42.8 (s, NCH2), 41.7 (s, NCH2), 38.1 (s, G2-
CH2NH2), 37.3 (br, G1-CH2NH–C3N3), 30.3 (s, G2-
NCH2CH2), 27.7 (br, G1-NCH2CH2), 25.5 (C5H10N, b-C),
24.6 (C5H10N, g-C). MS (MALDI): calc. 2954.1932 (M+);
found 2955.0237 (M + H+).
NCH2), 2.68 (br t, 3JHꢁH = 7 Hz, 12H, CH2NH2), 1.73 (br m,
12H, NCH2CH2), 1.62 (br, 6H, C5H10N, g-H), 1.54 (br, 24H,
C5H10N, b-H and NH2). 13C{1H} NMR (75.5 MHz, CDCl3) d
165.5 (s, C3N3), 165.2 (s, C3N3), 164.8 (s, C3N3), 44.0 (C5H10N,
a-C), 43.0 (s, CH2, piperazine), 42.4 (s, NH2–NCH2), 39.1 (s,
CH2NH2), 39.0 (s, CH2NH2), 31.3 (s, NCH2CH2), 25.7
(C5H10N, b-C), 24.9 (C5H10N, g-C). MS (MALDI): calc.
1206.8904 (M+); found 1207.9496 (M + H+).
G2-Cl. A solution of G1-NH2 (889 mg, 0.74 mmol), DIPEA
(2.1 mL, 12.5 mmol) and 1 (2.99 g, 6.21 mmol) was prepared in
a THF–water mixture (200 : 10 mL) to give a slurry. The
solution was stirred at 25 1C for 16 h, and then filtered through
Celite. The solvent was removed in vacuo. Purification was
achieved using column chromatography on silica gel (40 : 1
CH2Cl2–MeOH; Rf = 0.19 using 20 : 1 CH2Cl2–MeOH as the
developing solvent) to afford the product as a white solid.
Yield: 2.66 g (93%). The excess/unreacted 1 may also be
recovered from this purification (Rf = 0.50 using 20 : 1
CH2Cl2:MeOH as the developing solvent). 1H NMR (300
MHz, CDCl3) d 6.08 (br, 2H, NH) 5.61 (br, 8H, NH), 5.44
(br, 2H, NH), 4.98 (br, 6H, NH), 3.82 (br, 24H, CH2,
piperazine), 3.74 (br, 12H, C5H10N, a-H), 3.64–3.56 (br,
36H, CH2, NCH2), 3.39 (br, 12H, CH2, CH2NH–C3N3),
3.08 (br, 24H, CH2NHBoc), 1.85 (br, 12H, G1-NCH2CH2),
1.73 (br, 30H, G2-NCH2CH2, C5H10N, g-H), 1.58 (br, 12H,
C5H10N, b-H), 1.43 (s, 54H, C(CH3)3), 1.41 (s, 54H, C(CH3)3).
13C{1H} NMR (75.5 MHz, CDCl3) d 169.2 (s, C3N3), 168.4 (s,
C3N3), 165.4 (s, C3N3), 165.1 (s, C3N3), 164.9 (s, C3N3), 164.6
(s, C3N3), 156.0 (s, C(O)), 155.7 (s, C(O)), 78.9 (s, C(CH3)3),
78.6 (s, C(CH3)3), 44.0 (br), 43.1 (br), 37.6 (s, CH2NHBoc),
36.6 (s, CH2NHBoc), 28.2 (s, C(CH3)3), 27.7 (s, NCH2CH2),
25.6 (C5H10N, b-C), 24.8 (C5H10N, g-C). MS (MALDI): calc.
3865.2622 (M+); found 3868.2945.
G3-Cl. A solution of G3-NH2 (0.234 g, 0.0794 mmol),
DIPEA (0.43 mL, 2.86 mmol) and 1 (0.914 g, 1.9 mmol) was
prepared in a CH2Cl2 (50 mL)–EtOAc (15 mL)–water (5 mL)
mixture to give a slurry. The solution was stirred at 25 1C for
16 h, the solvents removed in vacuo and the residue taken up in
50 mL CH2Cl2. The solution was washed with water (3 ꢂ 100
mL) and brine (1 ꢂ 100 mL). The organic phases were filtered
through silica to remove excess starting material. The solvent
was removed in vacuo and any residual starting materials were
removed by reprecipitation from CH2Cl2 with hexanes to
afford the product as a white solid. Yield: 0.5118 g (78%).
1H NMR (300 MHz, CDCl3) d 9.33, 5.62, 4.98 (br, 42H, NH),
3.75 (br, 24H, CH2, piperazine), 3.65–3.52 (br, 120H, NCH2
and C5H10N, a-H), 3.36–3.29 (br, 36H, CH2NH–C3N3), 3.03
(br, 48H, CH2NHBoc), 1.89 (br, 12H, G1-NCH2CH2), 1.79
(br, 24H, G2-NCH2CH2), 1.68 (br, 48H, G3-NCH2CH2), 1.49
(br, 54H, C5H10N, b-H and g-H), 1.39 (s, 108H, C(CH3)3),
1.37 (s, 108H, C(CH3)3). 13C{1H} NMR (75.5 MHz, CDCl3)
d 169.7 (s, C3N3), 168.8 (s, C3N3), 168.2 (s, C3N3), 165.9
(s, C3N3), 165.6 (s, C3N3), 165.3 (s, C3N3), 165.1 (s, C3N3),
G2-Pip. A clear solution of G2-Cl (3.18 g, 0.823 mmol) and
piperidine (1.5 mL, 14.8 mmol) was prepared in THF (50 mL).
The mixture was stirred at 25 1C for 12 h, and the resulting
slurry was filtered through silica to give a clear solution. The
solvent was removed in vacuo. The solid was washed with
hexanes and dissolved in 50 mL CH2Cl2. The solution was
washed with water (2 ꢂ 50 mL) and brine (1 ꢂ 50 mL). The
organic phases were dried with Na2SO4 and the solvent was
removed in vacuo. The residue was reprecipitated from CH2Cl2
with hexanes, then EtOAc with hexanes to afford the product
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
1288 | New J. Chem., 2007, 31, 1283–1290

View Article Online
164.8 (s, C3N3), 156.4 (s, C(O)), 156.1 (s, C(O)), 79.2 (s,
C(CH3)3), 79.0 (s, C(CH3)3), 44.4 (br), 43.7 (br), 43.6 (br),
43.3 (br), 39.0 (br), 38.8 (br), 38.1 (br), 37.1 (br), 28.6 (s,
C(CH3)3), 28.1 (br), 27.9 (br), 26.0 (br), 25.2 (br). MS (MAL-
DI): calc. 8271.27 (average M+); found 8272.27 (average M +
H+).
with water (3 ꢂ 100 mL) and brine (1 ꢂ 100 mL). The organic
phases were filtered through silica to remove excess starting
material. The solvent was removed in vacuo and any residual
starting materials were removed by reprecipitation from
CH2Cl2 with hexanes to afford the product as a white solid.
Yield: 1.718 g, 69%. 1H NMR (300 MHz, CDCl3) d 9.39, 5.65,
5.03 (br, 90H, NH), 3.79–3.53 (br, 288H, CH2, piperazine,
NCH2 and C5H10N, a-H), 3.35–3.30 (br, 84H,
CH2NH–C3N3), 3.04 (br, 96H, CH2NHBoc), 1.88–1.69 (br,
180H, NCH2CH2), 1.58–1.47 (br, 126H, C5H10N, b-H and g-
H), 1.38 (bs, 432H, C(CH3)3). 13C{1H} NMR (75.5 MHz,
CDCl3) d 168.8 (br, C3N3), 168.2 (br, C3N3), 166.4 (br, C3N3),
165.9–165.8 (br, C3N3), 165.6 (s, C3N3), 165.4–165.3 (br,
C3N3), 165.1 (br, C3N3), 164.9 (br, C3N3), 156.4 (s, C(O)),
156.1 (s, C(O)), 79.3 (s, C(CH3)3), 79.0 (s, C(CH3)3), 44.5–44.3
(br), 43.7–43.5 (br), 38.9 (br), 38.0–37.8 (br), 37.2–37.0 (br),
28.6 (s, C(CH3)3), 28.1 (br), 26.1 (br), 25.2 (br). MS (MALDI):
calc. 17 083.3001 (average M+); found 17 088.03 (Baverage
M + H+).
G3-Pip. A clear solution of G3-Cl (0.66 g, 0.080 mmol) and
piperidine (0.316 mL, 3.19 mmol) was prepared in THF (30
mL). The mixture was stirred at 25 1C for 12 h, and the
resulting slurry was filtered through silica to give a clear
solution. The solvent was removed in vacuo. The solid was
washed with hexanes and dissolved in 50 mL CH2Cl2. The
solution was washed with water (2 ꢂ 50 mL) and brine (1 ꢂ 50
mL). The organic phases were dried with Na2SO4 and the
solvent was removed in vacuo. The residue was reprecipitated
from CH2Cl2 with hexanes to afford the product as a white
powder. Yield: 721 mg (499%). 1H NMR (300 MHz, CDCl3)
d 7.00–5.00 (br, 42H, NH), 3.78 (br, 24H, CH2, piperazine),
3.68 (br, 84H, C5H10N, a-H), 3.53 (br, 84H, NCH2), 3.32 (br,
36H, CH2NH–C3N3), 3.02 (br, 48H, CH2NHBoc), 1.78 (br,
36H, NCH2CH2), 1.65 (br, 48H, NCH2CH2), 1.50 (br, 126H,
C5H10N, g-H and b-H), 1.40 (s, 108H, C(CH3)3), 1.39 (s,
108H, C(CH3)3). 13C{1H} NMR (75.5 MHz, CDCl3) d 165.9
(s, C3N3), 165.6 (s, C3N3), 165.2 (s, C3N3), 164.6 (s, C3N3),
155.9 (s, C(O)), 78.7 (s, C(CH3)3), 43.9 (br), 42.9 (br), 41.7 (br),
33.9 (s, CH2NHBoc), 28.3 (s, C(CH3)3), 27.5 (br, NCH2CH2),
25.7 (br, C5H10N, b-C), 24.8 (br, C5H10N, g-C). MS
(MALDI): calc. 8855.5141 (average M+); found 8856.89
(average M + H+).
G4-Pip. A clear solution of G4-Cl (284 mg, 0.0166 mmol)
and piperidine (0.14 mL, 1.394 mmol) was prepared in THF
(16 mL). The mixture was stirred at 25 1C for 16 h, and the
resulting slurry was filtered through silica to give a clear
solution. The solvent was removed in vacuo. The solid was
washed with hexanes and dissolved in 50 mL CH2Cl2. The
solution was washed with water (2 ꢂ 50 mL) and brine (1 ꢂ 50
mL). The organic phases were dried with Na2SO4 and the
solvent was removed in vacuo. The residue was reprecipitated
from CH2Cl2 with hexanes to afford the product as a white
powder. Yield: 265 mg (87%). 1H NMR (300 MHz, CDCl3) d
7.12–5.08 (br, 90H, NH), 3.78–3.69 (br, 204H, CH2, piper-
azine, C5H10N, a-H), 3.56 (br, 180H, CH2, NCH2), 3.35 (br,
84H, CH2, CH2NH–C3N3), 3.02 (br, 96H, CH2NHBoc), 1.83
(br, 84H, NCH2CH2), 1.65 (br, 96H, NCH2CH2), 1.50 (br,
270H, C5H10N, g-H and b-H), 1.41 (s, 432H, C(CH3)3).
13C{1H} NMR (75.5 MHz, CDCl3) d 166.0–165.7 (br,
C3N3), 164.9 (br, C3N3), 156.2 (s, C(O)), 79.1 (s, C(CH3)3),
44.3 (br), 43.3 (br), 42.2 (br), 38.9 (br), 37.3 (s, CH2NHBoc),
28.7 (s, C(CH3)3), 27.9 (br, NCH2CH2), 26.0 (s, C5H10N, b-C),
25.1 (s, C5H10N, g-C). MS (MALDI): calc. 18 251.78 (average
M+); found 18 279.01 (Baverage M + Na+).
G3-NH2. 5 M HCl (25 mL) was added to a solution of G3-
Pip. (706 mg, 0.0797 mmol) in MeOH (70 mL), and the clear
solution was stirred at 25 1C for 16 h. The volatile components
were concentrated in vacuo until only ca. 5 mL of water
remained. The residue was made basic (pH = 14) with 26
mL of 5 M NaOH (aq.) solution, and the resulting milky
suspension was extracted with CHCl3 (5 ꢂ 150 mL). The
organic extractions were combined, dried with Na2SO4, and
then the solvent was removed in vacuo to afford the product as
a white solid. Yield: 520 mg, 499%. 1H NMR (300 MHz,
CDCl3) d 7.00–5.00 (br, 18H, NH), 3.76 (br, 24H, CH2,
piperazine), 3.62 (br, 84H, NCH2), 3.53 (br, 84H, NCH2),
3.27 (br, 36H, CH2NH–C3N3), 2.61 (br, 48H, CH2NH2), 1.99
(br, 48H, NH2), 1.66 (br, 84H, NCH2CH2), 1.54 (br, 42H,
C5H10N, g-H), 1.47 (br, 84H, C5H10N, b-H). 13C{1H} NMR
(75.5 MHz, CDCl3) d 166.0 (s, C3N3), 165.9 (s, C3N3), 165.3 (s,
C3N3), 165.1 (s, C3N3), 164.8 (s, C3N3), 164.6 (s, C3N3), 164.5
(s, C3N3), 43.8 (s, NCH2), 42.9 (s, NCH2), 42.2 (s, NCH2), 38.9
(s, CH2NH2), 37.4 (br, CH2NH–C3N3), 31.1 (s, NCH2CH2),
27.9 (br, NCH2CH2), 25.6 (C5H10N, b-C), 24.8 (C5H10N, g-C).
MS (MALDI): calc. 6448.7990 (M+); found 6471.9424
(M + Na+).
G4-NH2. A clear solution of G4-Pip (480 mg, 0.0263 mmol)
in CH2Cl2 (15 mL) and MeOH (90 mL) was prepared. A 5 M
HCl solution (24 mL) was added and the reaction was stirred
at room temperature for 16 h. The volatile components were
concentrated in vacuo until only ca. 5 mL of water remained.
The residue was made basic (pH = 14) with 26 mL of 5 M
NaOH (aq.) solution, and the resulting milky suspension was
extracted with CHCl3 (5 ꢂ 100 mL). The organic extractions
were combined, dried with Na2SO4, and the solvent was
removed in vacuo to afford the product as a white solid. Yield:
1
429 mg (499%). H NMR (300 MHz, CD3OD) d 3.65, 3.04,
2.04, 1.56. MS (MALDI): calc. 13 446.22 (average M+); found
13 453.87.
G4-Cl. A clear solution of G3-NH2 (0.94 g, 0.146 mmol) in
H2O (15 mL) and EtOAc (80 mL) was added to a clear
solution of 1 (5.03 g, 10.49 mmol) in CH2Cl2 (90 mL) with
DIPEA (4.79 mL, 31.47 mmol) and stirred at room tempera-
ture for 16 h. The solvents were removed in vacuo and the
residue taken up in 20 mL CH2Cl2. The solution was washed
G5-Cl. A clear solution of G4-NH2 (0.175 g, 0.013 mmol) in
35 mL CHCl3 with DIPEA (0.57 mL, 3.74 mmol) was
prepared and concentrated in vacuo to ca. 10 mL. Monomer
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
New J. Chem., 2007, 31, 1283–1290 | 1289

View Article Online
1 was added and the solution was stirred at room temperature
for 5 days. CHCl3 (50 mL) was added and the solution was
washed with water (3 ꢂ 75 mL) and brine (1 ꢂ 75 mL). The
organic layers were filtered through silica to remove excess 1.
Portions containing the product and an impurity were filtered
through silica again and the product was retrieved as a white
were combined, dried with Na2SO4, and the solvent was
removed in vacuo to afford the product as a white solid. Yield:
mg (499%). 1H NMR (300 MHz, CD3OD) d 3.73, 3.13, 2.08,
1.61.
Acknowledgements
1
solid. Yield: 0.111 g (25%). H NMR (300 MHz, CDCl3) d
7.70, 7.23, 5.67, 5.03 (NH), 3.65–3.36 (CH2, piperazine, NCH2,
C5H10N, a-H, and CH2NH–C3N3), 3.06 (CH2NHBoc),
1.84–1.47 (br, 642H, NCH2CH2, C5H10N, b-H and g-H),
1.40 (br, 864H, C(CH3)3). 13C{1H} NMR (75.5 MHz, CDCl3)
d 165.9–164.6 (br, C3N3), 163.5 (br, C3N3), 156.4 (s, C(O)),
156.2 (s, C(O)), 79.2 (s, C(CH3)3), 79.0 (s, C(CH3)3), 44.3 (br),
43.7 (br), 38.9 (br), 37.9 (br), 37.1 (s, CH2NHBoc), 28.7 (s,
C(CH3)3), 28.0, 27.9 (br, NCH2CH2), 26.0 (s, C5H10N, b-C),
25.6 (s, C5H10N, g-C). MS (MALDI): calc. 34 707.36 (average
M+); found 32 932.41, 15 823.00.
This work is supported by the NIH (EES; R01 GM 064560).
H. C. is an NSF Predoctoral Fellow. E. H. was a NSERC
Postdoctoral Fellow. We thank Dr Shane Tichy and his staff
in the Laboratory for Biological Mass Spectrometry for
analytical services.
References
1 B. Helms and E. W. Meijer, Science, 2006, 313, 929–930.
2 Dendrimers in Medicine and Biotechnology: New Molecular Tools,
ed. U. Boas, J. B. Christensen and P. M. H. Heegaard, RSC
Publishing, Cambridge, 1st edn, 2006, p. 182.
3 Dendrimer Catalysis (Topics in Organometallic Chemistry), ed. L.
H. Gade, Springer, Berlin, 1st edn, 2006, p. 189.
4 Dendrimers V: Functional and Hyperbranched Building Blocks,
Photophysical Properties, Applications in Materials and Life
Sciences (Topics in Current Chemistry), ed. C. A. Schalley and
F. Vogtle, Springer, Berlin, 1st edn, 2003, p. 273.
¨
5 D. A. Tomalia and J. M. J. Frechet, J. Polym. Sci., Part A: Polym.
Chem., 2002, 40, 2719–2728.
6 A.-M. Caminade, C.-O. Turrin, R. Laurent, C. Rebout and J.-P.
Majoral, Polym. Int., 2006, 55, 1155–1160.
7 E. M. M. de Brabander-van den Berg and E. W. Meijer, Angew.
Chem., Int. Ed. Engl., 1993, 32, 1308–11.
8 E. Hollink and E. E. Simanek, Org. Lett., 2006, 8, 2293–2295.
9 M. B. Steffensen, E. Hollink, F. Kuschel, M. Bauer and E. E.
Simanek, J. Polym. Sci., Part A: Polym. Chem., 2006, 44,
3411–3433.
10 M. B. Steffensen and E. E. Simanek, Angew. Chem., Int. Ed., 2004,
43, 5178–5180.
G5-Pip. A clear solution of G5-Cl (0.0523 g, 0.0015 mmol)
in THF (5 mL) was prepared. Piperidine (0.022 g, 0.216 mmol)
was added and the solution was stirred at room temperature
for 16 h. The cloudy solution was filtered through silica to
remove excess piperidine. The solvents were removed in vacuo
1
to yield a white solid. Yield: 0.059 g (499%). H NMR (300
MHz, CDCl3) d 3.68 (CH2, piperazine, C5H10N, a-H), 3.54
(br, 372H, NCH2), 3.36 (CH2, CH2NH–C3N3), 3.04 (br, 192H,
CH2NHBoc), 1.68–1.52 (br, 930H, NCH2CH2, C5H10N, g-H
and b-H), 1.41 (s, 864H, C(CH3)3). MS (MALDI): calc.
37 044.33 (average M+); found 33 010.74.
G5-NH2. A clear solution of G5-Pip (0.0556 g, 0.0015
mmol) in CH2Cl2 (3 mL) and MeOH (9 mL) was prepared.
5 M HCl (aq.) (3 mL) was added and the solution was stirred
at room temperature for 2 days. The volatile components were
concentrated in vacuo until only ca. 1 mL of water remained.
The residue was made basic (pH = 14) with 5 mL of 5 M
NaOH (aq.) solution, and the resulting milky suspension was
extracted with CHCl3 (5 ꢂ 15 mL). The organic extractions
11 J. Lim and E. E. Simanek, Mol. Pharmacol., 2005, 2, 273–277.
12 S. P. Rannard and N. J. Davis, Org. Lett., 2000, 2, 2117–2120.
13 H. T. Chen, M. F. Neerman, M. F. A. R. Parrish and E. E.
Simanek, J. Am. Chem. Soc., 2004, 126, 10044–10048.
14 Cerius2 4.9 Forcefield-Based Simulations, January 2003, Accelrys
Inc., San Diego, CA, 2006.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2007
1290 | New J. Chem., 2007, 31, 1283–1290