Elucidating factors leading to acidolytic degradation of sterically strained oligoether dendrons

Article abstract of DOI:10.1039/c4ob00540f

While steric hindrance can be particularly large in dendritic molecules, it is usually implicated with difficulties in the synthesis of higher generation structures and restricted access of reagents, including bond-cleaving agents, to the dendritic interior. A different situation, where the steric hindrance is translated into a steric strain within the dendritic molecule and, consequently, causes enhanced decomposition of the dendron-containing structure, has only occasionally been reported and exclusively for dendronized polymers. In this work we describe post-synthetic cleavage of sterically congested third-generation oligoether dendrons from solid supports, followed by their disassembly into monomeric building blocks under acidic conditions. This disassembly was monitored by H NMR, revealing the intermediate fragment structures and the exact order of bond cleavage within the dendritic molecule. These conclusions were supported by MS analysis of the cleavage mixtures. Though distinguishing between steric and electronic reasons of molecular disassembly can be a challenging task, we were able to analyze these factors separately by monitoring the dendron disassembly and comparing the rates of decay of parent dendrons and their fragments. This comparison reveals that while the electron donation of the steric congestion-inducing alkyl substituents is a prerequisite for the disassembly of the structures, this disassembly is very strongly accelerated in the sterically crowded dendrons and intermediates, where both steric and electronic factors contribute in a synergistic way to the disassembly phenomenon. the Partner Organisations 2014.

Full text of DOI:10.1039/c4ob00540f

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Elucidating factors leading to acidolytic
degradation of sterically strained oligoether
dendrons†
Cite this: Org. Biomol. Chem., 2014,
12, 5621
J. Karabline-Kuks, A. Fallek and M. Portnoy*
While steric hindrance can be particularly large in dendritic molecules, it is usually implicated with difficul-
ties in the synthesis of higher generation structures and restricted access of reagents, including bond-
cleaving agents, to the dendritic interior. A different situation, where the steric hindrance is translated into
a steric strain within the dendritic molecule and, consequently, causes enhanced decomposition of the
dendron-containing structure, has only occasionally been reported and exclusively for dendronized poly-
mers. In this work we describe post-synthetic cleavage of sterically congested third-generation oligoether
dendrons from solid supports, followed by their disassembly into monomeric building blocks under acidic
conditions. This disassembly was monitored by H NMR, revealing the intermediate fragment structures
and the exact order of bond cleavage within the dendritic molecule. These conclusions were supported
by MS analysis of the cleavage mixtures. Though distinguishing between steric and electronic reasons of
molecular disassembly can be a challenging task, we were able to analyze these factors separately by
monitoring the dendron disassembly and comparing the rates of “decay” of parent dendrons and their
fragments. This comparison reveals that while the electron donation of the steric congestion-inducing
alkyl substituents is a prerequisite for the disassembly of the structures, this disassembly is very strongly
accelerated in the sterically crowded dendrons and intermediates, where both steric and electronic
factors contribute in a synergistic way to the disassembly phenomenon.
Received 11th March 2014,
Accepted 30th May 2014
DOI: 10.1039/c4ob00540f
www.rsc.org/obc
hand, there are very few examples of the disassembly of
dendritic structures facilitated by steric strain.5 Moreover, to
Introduction
Steric crowding induces steric strain in certain organic mole- the best of our knowledge, these few examples were reported
cules and, accordingly, can lead to the increased propensity of for dendronized polymers, rather than for isolated dendrimers
these molecules to undergo decomposition.1 Frequently, the or dendrons. Due to the numerous proposed uses of dendri-
substituents that induce steric congestion also have an elec- mers with sterically-demanding groups,6 it is of considerable
tronic influence, and hence the cause of decomposition can importance to understand the limitations of sterically crowded
also be attributed to electronic factors. Therefore, distinguish- dendritic molecules, particularly with regard to their stability.
ing between steric and electronic reasons of molecular dis-
assembly could become a challenging task.
We have recently reported that sterically crowded polyether
dendrons can be prepared on a solid polystyrene support up to
Steric hindrance can be particularly large in dendritic mole- the third generation (Fig. 1a).7–9 We also observed and
cules and macromolecules, due to their highly branched reported that upon exposure to acid these structures undergo a
nature.2 This hindrance is known to preclude synthesis of facilitated disassembly (as compared to their analogues that
larger dendritic structures or, vice versa, to inhibit disassembly lack strain-inducing substituents). Using model “hetero-den-
of degradable dendritic molecules by obstructing access of drons” prepared from both the sterically congested and strain-
reagents to the built-in labile functionalities.3,4 On the other free monomers (Fig. 1b,c), we demonstrated that it is primarily
the electronic influence of the substituents (stabilization of
the benzylic cations by the ortho- and para-positioned propyls)
School of Chemistry, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv
that leads to the diminished stability of the dendrons.
However, the differences in the decomposition rates of the
“homo-” and “hetero-dendrons” pointed to an additional
University, Tel Aviv 69978, Israel. E-mail: portnoy@post.tau.ac.il
†Electronic supplementary information (ESI) available: Spectra of the H NMR
monitoring of the acidolytic disassembly of the dendrons G3(E) and G3des1,3(E),
factor that may be electronic (the increase in the basicity of the
the synthesis of G3des1,3(E), as well as the relevant MS spectra. See DOI:
10.1039/c4ob00540f
phenol oxygen) or steric. In addition, substantial differences
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 5621–5628 | 5621

View Article Online
Paper
Organic & Biomolecular Chemistry
interconversion can be followed with reasonable ease and a
degree of confidence.
The identification of various species in the reaction mixture
was based primarily on the aromatic proton signals and the
signals of protons in O-substituted benzylic positions. The
assignment of the signals was based on chemical shifts,
expected when taking into account electronic and steric influ-
ence of the neighbour substituents, as well as the general NMR
data and the NMR data specific to the related dendritic
molecules.7–9,11 In addition, we based the assignment of the
signals on their size and their size changes during the monitoring.
The position of the aromatic protons of the various dendrons
and fragments is highly sensitive to the ortho and para substi-
tuents, as well as to the steric crowding around them. Thus, the
combined electron-withdrawing effect of two ortho carbo-
methoxy groups induces a strong downfield shift of the aro-
matic proton signal (appearing at 7.92 to 8.14 ppm), as
compared with the protons neighbouring two oxyalkyl groups
(the “inner” modules of dendrons and fragments). The posi-
tion of the aromatic protons is also influenced by the para sub-
stituent, with the hydroxy group (focal point modules and the
G1(E) building block) affecting an upfield shift of the proton
signal, as compared to the position of those with a benzyloxy
substituent. Finally, the steric crowding can substantially
shift downfield the position of aromatic protons, compared
to the position of protons with similar ortho and para substi-
tuents. In other words, the “deeper” inside the dendritic
structure is located a given proton, the more downfield its
signal will appear, when compared to its analogue positioned
closer to the periphery. A similar trend was observed pre-
viously for various aryl–benzyl ether dendrons.9,11 In the spec-
tral area characteristic to protons of O-substituted benzylic
methylenes, chemical shifts in the 4.86–5.00 ppm range were
characteristic to benzyl ethers, while those in the range of
5.38–5.45 ppm were characteristic to benzyl trifluoroacetate
esters.
Fig. 1 Second-generation “homo-” and “hetero-dendrons”.7
were observed in the decomposition rates of the dendrons vs.
their partially-disassembled successors, clearly pointing to the
involvement of the steric strain. In order to further explore the
observed phenomenon and, possibly, elucidate all the factors
leading to the acid-induced disassembly of the dendrons, we
decided to carefully monitor the decomposition of the third-
generation structures: the previously reported “homo-
dendron”, as well as a “hetero-dendron” especially prepared
for these experiments. We herein report the results of this
study. This investigation could shed light on the stability of
oligoether dendrons in a low pH environment, with impli-
cations on their potential use in drug delivery systems and
other biomedical devices.
Notably, the G3(E) dendron appears as a main component
of the mixture in the first spectra of the monitoring sequence
and exhibits characteristic aromatic (8.136, 7.803 and
7.708 ppm) and low-field benzylic signals (4.995 and 4.925).12 It
is important to mention that, according to the first spectra of
the series, the third-generation dendron was slightly contami-
nated by lower generation dendrons (G2(E) and G1(E)).
However, since these species are also part of the disassembly
chain of G3(E) (as we will show subsequently), these impurities
Results and discussion
Following its cleavage from the support, the third-generation did not complicate the analysis of the decomposition pattern.
dendron G3(E) can disintegrate into monomeric building The first spectra of the sequences clearly show two major
blocks by a number of parallel or intersecting routes, schemati- fragments being created from G3(E), nearly in parallel. The
cally depicted in Scheme 1. If all of them occur in parallel first was identified capitalizing on a prior art as G2(E), which
with the same probability during the NMR monitoring of the exhibits characteristic aromatic (8.128 and 7.520 ppm) and a
acidolytic disassembly of the dendron, one may expect a series low-field benzylic (4.893 ppm) signals.7 The second fragment
of very complex spectra, which will be practically impossible to exhibited distinct aromatic singlets at 8.141 and 7.339 ppm
decipher. Fortunately, monitoring of the dendron decompo- and distinct benzylic singlets at 5.436 and 4.943 ppm. Two
sition in a 9 : 1 CDCl3–TFA mixture (Fig. 2) exhibited the for- additional singlet signals of this fragment, an aromatic and a
mation of 6 to 7 major fragments only (Fig. 3), in such a way benzylic, at 7.797 and 4.922 ppm respectively, were concealed
that their formation from the third-generation dendron or in these spectra by the closely situated signals of G3(E), but
5622 | Org. Biomol. Chem., 2014, 12, 5621–5628
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 1 All possible routes for the disassembly of G3(E).
became clearly visible in the later spectra of the series. This characterized by the characteristic aromatic singlet at 7.924 ppm,
set of signals was attributed to fragment A (Fig. 3). Concurrent while compound D exhibits a distinct aromatic signal at
growth of the G2(E) portion in the cleavage mixture supports 7.241 ppm and distinct low-field benzylic signals at 5.416 and
this assignment, since cleavage of G2(E) from G3(E) will inevi- 4.858 ppm. In addition, one of the aromatic signals of D at
tably lead to the formation of A.
8.129 ppm merges with the signal of G2(E) at the same location.
The fate of fragment D, known from our previous study, is
Subsequent spectra of the decomposition monitoring reveal
the expected disassembly of the G2(E) dendron to G1(E) and the disassembly into another equivalent of G1(E) and a
fragment D, according to our past observations.7 G1(E) is reduced first-generation module G1(OCOCF3). Indeed, in the
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 5621–5628 | 5623

View Article Online
Paper
Organic & Biomolecular Chemistry
later spectra of the monitoring series, an aromatic signal at
7.032 ppm and a benzylic signal at 5.378 ppm, gradually
increasing in intensity, are attributed to the G1(OCOCF3)
fragment.
The mode of disassembly of the fragment A is more
difficult to track. While it can disintegrate into G2(E) and G1
(OCOCF3), we cannot be sure what the contribution of this
mode of decomposition to the disassembly of A is, since both
fragments are generated by alternative routes. An additional
mode of the disintegration of A leads to G1(E) and a fragment
B (Fig. 1). While G1(E) is also formed in a number of the above
mentioned fragmentations (disassembly of G2(E) and of D),
the generation of B in the cleavage solution is likely to proceed
mostly through this route (vide infra) and will constitute a
proof for the existence of this mode of decomposition of A.
Beginning a few hours from the start of the cleavage and dis-
assembly, a new set of signals gradually appears (in the earlier
spectra most of the signals are concealed by the larger signals
of other fragments) and becomes distinct in the later monitor-
ing spectra. The set includes the aromatic singlets at 8.141,
7.486 and 7.300 ppm as well as the low-field benzylic singlets
at 5.445, 5.428, 4.884 and 4.921 ppm. Such a set of signals best
fits the proposed fragment B. Moreover, the decline in the
amount of A, which begins after 11 hours of the monitoring, is
largely compensated by the increase in the amount of the frag-
ment responsible for the appearance of these peaks. This
observations postulates that B is indeed formed, and the con-
version of A into B and G1(E) is the preferred mode of A’s
disassembly.
Fig. 2 H NMR monitoring of the disassembly of G3(E) in a 9 : 1 CDCl3–
TFA mixture.10 Residual CHCl3 and benzene reference appear at 7.260
and 7.362 ppm, respectively.
All the fragments observed in the cleavage mixture and
described above can be explained by the primary disassembly
event of G3(E) breaking into A and G2(E). This, however, leaves
open the question whether another mode of G3(E) decompo-
sition, breakage into G1(E) and a fragment C, actually takes place
during the acidolytic disassembly (Scheme 1). Having a very
similar structure to G3(E), fragment C is likely to exhibit NMR
signals that are very similar to those of its parent compound.
Thus, if C is produced, most of its signals will be concealed by
those of G3(E). However, the aromatic and benzylic signals,
closest to the site of cleavage that formed this fragment, must be
very different from those of its parent structure. We expect that
the chemical shifts of these two signals will be similar to those
of the related hydrogens in B, i.e. 7.486 and 5.445 respectively.
Since B accumulates in the solution in a detectable amount
only after a few hours (ca. 4 h), while the hypothetical fragment
C must be formed (if at all) during the initial stages of the
monitoring whilst the concentration of G3(E) is the highest,
there is a time window when the two distinct signals of C can
be detected. A careful examination of the first spectra of the
monitoring series indeed reveals two very small signals at
7.489 and 5.446 ppm, which cannot be attributed to B, since
other characteristic signals of B are still missing in these
spectra. Once the presence of B becomes visible, these two
signals merge with the related signals of B, until ca. 11 h from
the start of monitoring the presumable signals of C are not
detectable any more.
Fig. 3 Dendrons and fragments, observable in
mixture, and their characteristic chemical shifts.
a 9 : 1 CDCl3–TFA
5624 | Org. Biomol. Chem., 2014, 12, 5621–5628
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Even if we accept the assumption that the signals at 7.489 A + D, since all these fragments are formed via competing pro-
and 5.446 ppm represent C, its amount during the monitoring, cesses of the main decomposition “chain”.
particularly at the early stages when it is supposed to be the
The proposed disassembly route was confirmed also by two
highest, is negligible. Moreover, during these early stages of ES-MS analyses of the mixture composition. In both positive
the disassembly, the potential products of its four possible and negative peak detection modes, the measurement carried
decomposition modes are either present in negligible out ca. 45 minutes after the beginning of the cleavage exhibi-
amounts (D and B, Scheme 1), or are not detected at all (E and ted the characteristic (M + Na)+ and (M + CF3CO2)− peaks of
F, Scheme 1). The latter observation rules out the possibility of G3(E) (m/z 1807.3 and 1897.0 respectively), as well as those of
slow formation/fast decomposition of C and, together with the A (m/z 1129.7 and 1219.5 respectively) and G2(E) (m/z 813.5
first observation, proves that the main route of decomposition and 903.4 respectively). Smaller peaks that can be attributed to
of G3(E) proceeds through A and G2(E), while the pathway pro- B (m/z 949.4 and 1039.4 in the positive and negative ion detec-
ceeding through C and G1(E) is very minor at most.
tion modes respectively) and D (m/z 633.2 in the positive mode
The last point to address, in regard to the NMR monitoring, only) were also observed.13 Small signals at m/z 1626.8 and
is the fate of B. Though it can decompose to a hypothetical 1716.8, in the positive and negative detection modes respect-
fragment G and G1(E), no evidence for G was found in the ively, can be considered as an evidence for the presence of a
latest spectra of the monitoring series, when the concentration trace amount of C and, hence, as an additional proof of the
of B in the mixture was the highest. Thus, it is logical to minor pathway involving this fragment. In both detection
assume that B mainly decomposes to G1(OCOCF3) and D.
modes there are no peaks that can be attributed to any other
Based on the above analysis of the NMR monitoring of the fragment depicted in Scheme 1.14 The measurement carried
acidolytic disassembly of the third-generation dendron, G3(E), out after the mixture was incubated for 27 hours demonstrated
we can reconstruct the actual disassembly route that is a complete disappearance of G3(E) and C, along with a
depicted in Scheme 2. Some of the pathways seem very reason- notable decrease in the intensity of the signals of G2(E), while
able (A to G1(OCOCF3) + G2(E), B to G1(OCOCF3) + D) but the signals of
A remained strong. This change was
could not be verified since their products are also formed by accompanied by a significant increase in the intensity of the
competing pathways. The “detour” via C is a minor pathway at signals attributed to B and D. Due to the high abundance of
most, and it is also not clear if C decays to B + G2(E) or to G1(E) and G1(OCOCF3) in the mixture, their (M − H)− peaks
are clearly visible in the negative ion detection mode (m/z
293.1 and 429.1 respectively), in spite of the low molecular
mass of these fragments. Very weak peaks in the positive and
negative detection modes may point to trace amounts of E/F
(m/z 1445.7 and 1535.7 for the (M + Na)+ and (M + CF3CO2)−
peaks respectively). A low intensity signal in the positive mode
(m/z 769.3 for the (M + Na)+ peak) and a stronger signal in the
negative mode (m/z 859.3 for the (M + CF3CO2)− peak) may
point to the presence of fragment G in the mixture. This evi-
dence, however, is not supported by the NMR findings.
Additional information shedding light on the factors
responsible for the degradation of this type of dendrons was
obtained from monitoring the decomposition of the mixed
dendron G3des1,3(E) (Scheme 3),15 which was prepared by a
route similar to that applied for other dendrons using the
simpler dimethyl 5-hydroxyisophthalate building block, instead
of its dipropylated analogue, for the assembly of the first and
third layers. It is important to mention that, according to the
first spectra of the monitoring series, the dendron G3des1,3(E)
was slightly contaminated by lower generation dendrons
(G1des(E) and G2des2(E)). However, since the first of these
impurities is a part of the disassembly chain of G3des1,3(E) (as
we will show below), while the second decomposes quickly
(within 1 day) into known fragments according to the pathway
described by us previously,7 these impurities did not compli-
cate the analysis of the decomposition pattern.
It is noteworthy that G3des1,3(E) is practically stable in a
9 : 1 CDCl3–TFA mixture, which led to complete decomposition
of G3(E) within 11 hours. H NMR follow-up of the decompo-
Scheme 2 The actual disassembly route of G3(E).
sition of G3des1,3(E) in a 1 : 1 CDCl3–TFA cleavage solution
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 5621–5628 | 5625

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 3 The disassembly route of G3des1,3(E) and the characteristic chemical shifts of the dendron and the fragments.
exhibited the slow conversion of the parent dendron, charac-
ES-MS measurements of the components of the cleavage
terized by aromatic signals (8.363(t), 7.952(d), 7.455, 7.397 and solution fully support the aforementioned assignment and
7.173 ppm) and low-field benzylic signals (5.226 and course of the events. Thus, in the solution analyzed with the
5.041 ppm), into fragment I, exhibiting characteristic peaks in positive ion detection mode ca. 45 minutes after the beginning
the aromatic (8.379(t), 7.964(d), 7.385, 7.363 and 7.152 ppm) of the cleavage, in the area above m/z = 800, (M + Na)+ and
and low-field benzylic (5.483, 5.224 and 5.007 ppm) regions, (M + K)+ peaks of G3des1,3(E) (m/z 1385.5 and 1401.5 respect-
and two equivalents of G1des(E), exhibiting aromatic signals ively) are the highest. The (M + Na)+ peak of intermediate H
at 8.348(t) and 7.863(d) (Scheme 3).16 This transformation is (m/z 1289.5) is clearly visible, while a weak signal with m/z =
completed in slightly over 24 h, so that after two days no traces 1193.4 points to a trace amount of compound I. Moreover in
of G3des1,3(E) or of the fragment H (an intermediate between the negative ion detection mode (M + CF3CO2)− peaks of
G3des1,3(E) and I) are visible. The fragment I decomposes G3des1,3(E), H and I (m/z 1475.5, 1379.4 and 1283.4 respect-
much more slowly into G2des1(OCOCF3) and two more equiva- ively) are the highest in the area above m/z = 800. The ES-MS
lents of G1des(E), in such a way that only after 2 weeks trace analysis of the cleavage mixture incubated for 24 hours exhibi-
signals of the fragments I and J (an intermediate between ted the characteristic (M + Na)+ and (M + CF3CO2)− peaks (in
I and G2des1(OCOCF3)) completely disappear from the spec- the positive and negative modes respectively) of I, J (m/z 1097.3
trum. The dendron G2des1(OCOCF3), the stable final product and 1187.3 respectively) and G2des1(OCOCF3) (m/z 1001.3 and
of the decomposition sequence, exhibits characteristic aro- 1091.3 respectively), without any signals of H or G3des1,3(E).
matic singlets at 7.333, 7.317 and 7.145 ppm as well as benzylic After a seven day incubation, only the characteristic peaks of
singlets at 5.474 and 4.980 ppm. The fragment H, being an G2des1(OCOCF3) and J (trace amount) were visible.
intermediate between G3des1,3(E) and I, is characterized by
We can now reconsider the factors that are responsible for
signals that are very close to those of the two fragments. the acidolytic disassembly of the two third-generation den-
However, even a slight difference (a few thousandth of ppm) drons, and their “daughter” fragments. First, the current
results in a broadening of the combined peaks, when at least results support the conclusion of the previous study that the
two of these three compounds are present. The same is true for main factor leading to the observed acid-induced degradation
the intermediate fragment J and the preceding and the succeed- of the dendrons is the weakening of the benzylic C–O bond,
ing compounds I and G2des1(OCOCF3), respectively.
caused by the electron donating effect of the two alkyl substi-
5626 | Org. Biomol. Chem., 2014, 12, 5621–5628
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
tuents on the aromatic ring positioned ortho and para to this hand, all full dendrons within the four pairs are more steri-
benzylic position. In the “hetero-dendron”, the corresponding cally crowded and, therefore, are likely to sustain a much
bonds of the module that lacks these substituents were not stronger steric strain. This must be the reason for their signifi-
cleaved by the acid.
cantly higher decomposition rate, as compared to their “one-
Another substituent that apparently affects the benzylic arm” counterparts within the pair.
C–O bond lability and, through it, the rate of the dendrons’
disassembly, is the meta-positioned hydroxy/benzyloxy group.
While both A and G2(E) are presumably generated by the same
event (Scheme 2), and A, contrary to G2(E), cannot be gener-
Experimental
The third-generation dendrons were prepared on a Wang
Bromo polystyrene resin (1% crosslinked divinylbenzene–
styrene copolymer, 100–200 mesh, loading 0.90 mmol g−1
ated by an alternative route, the decomposition of G2(E)
occurs notably faster than that of the fragment A. Further-
more, the decomposition of the “hetero-dendron” G2des2(E) in
a 1 : 1 CDCl3–TFA solution, described in our previous report,7
is notably faster than the conversion of G3des1,3(E) into H, or
of H into I. In both cases a cleavage of the benzylic C–O bond
takes place in a very similar environment with the only appar-
ent disparity, to which the differences in the cleavage rates can
be attributed, being the meta-to-benzyl substituent. It seems
that the hydroxyl substituent affects faster cleavage, as com-
pared to benzyloxy functionality.
,
Novabaiochem) according to the previously reported pro-
cedures.7,9 1H NMR monitoring measurements were carried out
on a Bruker AVANCE-400 spectrometer, in CDCl3–TFA 9 : 1 or
CDCl3–TFA 1 : 1, with residual CHCl3 (7.26 ppm), as calibration
standard, and with benzene (7.36 ppm) or 1,1,2,2-tetrachlor-
oethane (5.95 ppm), as internal standards. For these measure-
ments, the resin was incubated with the cleavage solution for
30 min, filtered off, and the filtrate used for monitoring.
For the MS analysis, following the cleavage of the dendrons
by CDCl3–TFA solution (9 : 1, v/v for G3(E) and 1 : 1, v/v for
G3des1,3(E)), the resins were filtered off. At the designated
times the solvents were evaporated. In order to ensure the
complete evaporation of TFA, a small volume of ethyl acetate
was added, the sample redissolved, and then the solvent was
evaporated once again. This step was repeated 3 times. The MS
spectra were recorded on a Waters Synapt-HDMS instrument.
It is clear now that, beyond the aforementioned electronic
factors, the steric strain plays an important role in facilitating
the disassembly in the described dendritic systems. Already in
the previous report it was observed that G2(E) decomposes
much faster than D, its successor in the disassembly chain,
and similarly G2des2(E) decomposes significantly faster than
its “daughter” fragment K (Scheme 4).7 In the current study,
monitoring of the G3(E) demonstrated that, while G3(E)
decomposes rapidly and disappears from the decomposition
solution within 11 hours, its immediate successor A, which is
generated in the mixture in a significant amount in the first
few hours of the monitoring, decays significantly slower
remaining in a countable amount in the mixture even after
almost 2 days. Furthermore, while it takes slightly more than
24 hours for G3des1,3(E) to be degraded, mostly to I, in the
1 : 1 CDCl3–TFA cleavage solution, the monitoring demon-
strates that only after 2 weeks I is fully degraded into G2des1-
(OCOCF3). For all four pairs of compounds described above,
the differences in the rates of decomposition, which proceeds
through benzylic C–O bond cleavage, cannot be traced to the
electronic influence of the presence or absence of the second
dendritic arm. Such an influence will have to propagate
through five bonds and this is highly unlikely. On the other
Conclusions
Although the propyl substituents can impart both electronic
and steric effects, sometimes inseparable, on the chemical be-
havior of organic molecules, in the case of the oligoether den-
drons described in this work we have been able to analyze
these factors separately by monitoring the dendron disassem-
bly and comparing the rates of “decay” of parent dendrons
and their fragments. Nevertheless, this analysis proved that
both the electron donation and the steric strain induced by the
propyls contribute in a synergistic way to the disassembly
phenomenon. Importantly, it was demonstrated, for the first
time, that steric strain significantly accelerates decomposition
of a single dendron to its building modules.
Acknowledgements
This research was supported by grant no. 955/10 from the
Israel Science Foundation and grant no. 2012193 from the
United States–Israel Binational Science Foundation (BSF).
Notes and references
1 H. C. Brown and R. S. Fletcher, J. Am. Chem. Soc., 1949, 71,
1845–1854.
Scheme 4 The disassembly of the “hetero-dendron” G2des2(E).
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 5621–5628 | 5627

View Article Online
Paper
Organic & Biomolecular Chemistry
2 (a) Dendrimers and other dendritic polymers, ed. J. M. J.
Fréchet and D. A. Tomalia, John Wiley & Sons, Chichester,
2001; (b) G. R. Newkome, C. N. Moorfield and F. Vögtle,
Dendrimers and dendrons. Concepts, synthesis, applications,
Wiley-VCH, Weinheim, 2001.
3 D. Ruiz-Molina, J. Veciana, F. Palacio and C. Rovira, J. Org.
Chem., 1997, 62, 9009–9017.
4 (a) J.-S. Lee, J. Huh, C.-H. Ahn, M. Lee and T. G. Park,
9 For synthesis of related less crowded dendrons on solid
support, see: (a) A. Dahan and M. Portnoy, Macromolecules,
2003, 36, 1034–1038; (b) A. Mansour, T. Kehat and
M. Portnoy, Org. Biomol. Chem., 2008, 6, 3382–3387.
10 For brevity only the aromatic region is shown. For the area
of low-field benzylic signals, see ESI.†
11 B. Helms, C. O. Liang, C. J. Hawker and J. M. J. Fréchet,
Macromolecules, 2005, 38, 5411–5415.
Macromol. Rapid Commun., 2006, 27, 1608–1614; 12 Low field signals of the O-substituted benzylic positions
(b) T. Etrych, L. Kovar, J. Strohalm, P. Chytil, B. Rihova and
K. Ulbrich, J. Controlled Release, 2011, 154, 241–248;
(c) T. Etrych, J. Strohalm, P. Chytil, P. Cernoch,
were used for the characterization of the mixture com-
ponents, while the high-field benzylic signals of the propyl
substituents were not.
L. Starovoytova, M. Pechar and K. Ulbrich, Eur. J. Pharm. 13 Fragments with lower molecular mass seem to have lower
Sci., 2011, 42, 527–539.
5 (a) H. Yu, A. D. Schluter and B. Z. Zhang, Helv. Chim. Acta,
2012, 95, 2399–2410; (b) G. Sanclimens, L. Crespo, M. Pons,
affinity toward the trifluoroacetate anion and, hence,
exhibit smaller peaks, or do not exhibit peaks at all, in the
negative ion detection mode.
E. Giralt, F. Albericio and M. Royo, Tetrahedron Lett., 2003, 14 The low molecular masses of G1(E) and G1(OCOCF3), both
44, 1751–1754.
of which must be present in the reaction mixture, is the
likely reason for the absence of their (M + Na)+, (M +
CF3CO2)− or (M − H)− peaks in the spectra.
6 (a) J. F. G. A. Jansen, E. M. M. de Brabander-van den Berg
and E. W. Meijer, Science, 1994, 266, 1226–1229;
(b) P. Gandhi, B. Huang, J. C. Gallucci and J. R. Parquette, 15 des stands for des-dipropyl, while des1,3 indicates that this
Org. Lett., 2001, 3, 3129–3132; (c) A. Kowalewska and
W. A. Stańczyk, ARKIVOC, 2006, 5, 110–115; (d) K. Chiad,
third-generation dendron lacks propyl substituents on the
first and third layer modules; for spectra see ESI.†
M. Grill, M. Baumgarten, M. Klappez and K. Müllen, Macro- 16 A minor gradual change in the chemical shifts of the frag-
molecules, 2013, 46, 3554–3560.
7 J. Karabline and M. Portnoy, Org. Biomol. Chem., 2012, 10,
4788–4794.
8 For reviews on polyether dendrons, see: (a) S. M. Grayson
and J. M. J. Frechet, Chem. Rev., 2001, 101, 3819–3867;
(b) J. G. Li, C. Meng, X. Q. Zhang, L. Zhang and
A. F. Zhang, Prog. Chem., 2006, 18, 1157–1180.
ment signals was observed over the 2 weeks, due to the
slow evaporation of CDCl3 from the cleavage mixture. In
order to compensate, small amount of CDCl3 was added
before the 14th day measurement, leading to an opposite
shift. Accordingly, chemical shifts reported in the manu-
script are those observed within the first 2 days of
monitoring.
5628 | Org. Biomol. Chem., 2014, 12, 5621–5628
This journal is © The Royal Society of Chemistry 2014