Efficient mechanochemical synthesis of regioselective persubstituted cyclodextrins

Article abstract of DOI:10.3762/bjoc.12.230

A number of per-6-substituted cyclodextrin derivative syntheses have been effectively carried out in a planetary ball mill under solvent-free conditions. The preparation of Bridion and important per-6-amino/thiocyclodextrin intermediates without polar aprotic solvents, a source of byproducts and persistent impurities, could be performed. Isolation and purification processes could also be simplified. Considerably lower alkylthiol/halide ratio were necessary to reach the complete reaction in comparison with thiourea or azide reactions. While the presented mechanochemical syntheses were carried out on the millimolar scale, they are easily scalable.

Full text of DOI:10.3762/bjoc.12.230

Efficient mechanochemical synthesis of regioselective
persubstituted cyclodextrins
Laszlo Jicsinszky*, Marina Caporaso, Katia Martina, Emanuela Calcio Gaudino
and Giancarlo Cravotto*
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2016, 12, 2364–2371.
Department of Drug Science and Technology and NIS - Centre for
Nanostructured Interfaces and Surfaces, University of Turin, Via P.
Giuria 9, 10125 Turin (Italy)
doi:10.3762/bjoc.12.230
Received: 30 July 2016
Accepted: 26 October 2016
Published: 10 November 2016
Email:
Laszlo Jicsinszky* - ljicsinszky@gmail.com; Giancarlo Cravotto* -
giancarlo.cravotto@unito.it
This article is part of the Thematic Series "Green chemistry".
Guest Editor: L. Vaccaro
* Corresponding author
Keywords:
© 2016 Jicsinszky et al.; licensee Beilstein-Institut.
License and terms: see end of document.
green chemistry; nucleophilic substitution; planetary ball mill; siRNA
delivery intermediate; sugammadex
Abstract
A number of per-6-substituted cyclodextrin derivative syntheses have been effectively carried out in a planetary ball mill under sol-
vent-free conditions. The preparation of Bridion® and important per-6-amino/thiocyclodextrin intermediates without polar aprotic
solvents, a source of byproducts and persistent impurities, could be performed. Isolation and purification processes could also be
simplified. Considerably lower alkylthiol/halide ratio were necessary to reach the complete reaction in comparison with thiourea or
azide reactions. While the presented mechanochemical syntheses were carried out on the millimolar scale, they are easily scalable.
Introduction
Cyclodextrins (CDs) are cyclic α(1→4)glucopyranosides and procedures. Ball mill assisted syntheses are good alternatives to
have been fully described in a number of publications [1-3]. overcome solubility difficulties in syntheses or isolation of
They are most noted for their ability to form non-covalent asso- natural compounds from vegetables using CDs [9,10]. Environ-
ciations called “inclusion complexes”. Natural CDs exhibit mentally benign synthetic methods of CD derivatives have been
many favourable properties, which advance their use in a wide recently reviewed [11].
range of applications. However, syntheses for many special ap-
plications, such as DNA sequencing [4,5], gene delivery [6], The key intermediate in the bulk preparation of selectively per-
and drug targeting hosts [7,8], can be problematic as they 6 substituted CDs is the per-6-deoxy-6-halide derivative, which
require sophisticated, efficient and yet simple methods which covers per-6-bromo [12] and per-6-iodo [13] compounds. Al-
lead to acceptable purity and impurity profiles. Furthermore, the though per-6-chloro-CD derivatives can also be easily prepared
special structural properties of selectively substituted CDs mean [14] but the lower reactivity of chloro compounds restricts the
that their syntheses are not always environmentally friendly use of per-6-chloro-CDs [15] in solution. The solubility of per-
2364

Beilstein J. Org. Chem. 2016, 12, 2364–2371.
6-halogeno-CDs is very limited in water and the majority of sequencing equipment [4]. p-Toluenesulfonyl (Ts) esters are
organic solvents, meaning that their preparation and purifica- still an irreplaceable leaving group for carbohydrates as well.
tion is far from being environmentally friendly.
The Ts → azide exchange in Ts-β-CD is effective in both solu-
tion and under high-energy ball milling (HEBM) conditions,
Per-6-S-(3-mercapto)propionyl-γ-CD (Sugammadex, Bridion®) whereas it is difficult in the per-halide analogues, because of so-
is not only the biggest success that CD derivatization has had, dium azide’s poor solubility in DMF, N,N-dimethylacetamide,
but its use in the removal of muscle relaxants may revolu- N-methylpyrrolidone and dimethyl sulfoxide. A solution to this
tionize surgery. It has very high affinity with curare analogues, problem might be found in the successive addition of sodium
especially rucoronium (K11 ≈ 1.8 × 107 M−1) [7], which are azide, considerably longer reaction times and/or higher temper-
widely used in surgery [6]. Its everyday use has led to increas- atures (≈100 °C) for the complete reaction. Per-6-bromo- and
ing demand and ever higher amounts being employed, meaning per-6-iodo-CDs are inevitably able to react with dimethylamine,
it will likely soon become a generic molecule in most hospitals. a common decomposition product of DMF. While the forma-
The solventless preparation technique can provide clear advan- tion of dimethylamine-moiety-containing CDs is virtually negli-
tages over classic methods since its use in humans requires very gible – thankfully, as it can cause serious problems in pharma-
high purity. The standard preparation of Sugammadex uses a ceutical or biological preparations – its physicochemical proper-
harsh base, such as sodium hydride, to activate the thiol group ties are very similar to those of the perazido derivatives, making
of 3-mercaptopropionic acid (MPA) and N,N-dimethylform- separation impossible. The similarity is even higher after the
amide (DMF) [16]. In the reaction solution and particularly on azide’s conversion to an amine. Solvent removal in the synthe-
larger scales (from 10 g to kilo lab scale), impurities that arise sis of per-6-azido-CDs is generally also challenging because of
from incomplete conversions, product and byproduct decompo- the high boiling polar aprotic solvents used; their complete
sition as well as solvent impurities increase synthesis and purifi- removal is a difficult task under laboratory conditions, even at
cation costs and time.
high vacuum.
Selectively per-6-thiolated CDs are also used in gold nanoparti- Ball milling is an effective method for the preparation of inclu-
cle chemistry, particularly in electrochemical sensors [17]. sion complexes which has recently begun to be appreciated by
Thiourea (TU) is one of the best precursors of those CDs organic chemists for its simplicity and flexibility [18-21]. While
because as the halogen is exchanged to the thiouronium salt of the mechanochemical manipulation of covalent bonds is hardly
CDs thiols can readily be obtained under aqueous basic condi- a brand new concept, its diffusion into carbohydrate chemistry,
tions. Thioureido-CDs are crystalline compounds that are and particularly into CD derivatization, has been rather slow
readily soluble in water, easy to purify and convert to thiols [8]. [22,23]. The ability of HEBM to favour the nucleophilic
These intermediates can be efficiently prepared via the reaction substitution reaction of 6I-monotosyl-β-CD has been demon-
of a TU excess (2–3 mol TU/halogen) and per-6-halogenated strated in a previous article [22]. Most interestingly, the side
CDs. The preparation is also carried out in a polar aprotic sol- reactions that appear to be unavoidable in the classical solution
vent, as in case of the azide exchange, while byproducts may syntheses can be eliminated in the solvent-free method de-
also be similar despite the higher nucleophilicity of sulfur. The scribed for the preparation of various CD derivatives. HEBM
reaction under classic conditions usually requires large TU was found to be particularly efficient when good nucleophiles,
excess.
such as sulfur-containing reactants or inorganic azides, were
used.
Various per-6-alkylthio-β-CD derivatives are used in siRNA
delivery and gene therapy [6]. The alkyl chain, usually The aim of our study is to highlight the use of HEBM in
C10–C16, makes the product very amphiphilic, which means the preparation of a number of practically important CD
that it is difficult to purify not only from possible byproducts derivatives, as seen in Scheme 1. Although in our work
but also from the reagents. Additionally, sulfur-containing neither the reaction conditions nor the purifications were opti-
organic compounds can form very strong complexes with not mized in any terms the presented results can serve as starting
only the native but with the substituted CDs, too. Although a point to develop more environmentally benign synthetic
solvent-free synthetic method does not solve the problem of methods of important CD derivatives. The well-established
complexation but the reduced amount of reagents can simplify engineering of HEBM reactions makes them easy to scale-up,
the purification.
while simplified work-up procedures can further reduce the
presence of unwanted byproducts which explains our interest
Heptakis(6-azido-6-deoxy)-β-CD is the precursor to per-6- in the preparation of compounds such as 3a, 5b, and 6
amino-β-CD, which is an important component of DNA (Scheme 1).
2365

Beilstein J. Org. Chem. 2016, 12, 2364–2371.
Scheme 1: Synthesis of per-6-derivatized CDs. Ball milling conditions: 1500 steel balls of 1 mm diameter and 50 steel balls of 5 mm diameter, sun
wheel speed 650 min−1, 2 h grinding.
Results and Discussion
satisfactorily by the complexation of the leaving group which
Per-6-iodinated and, in some cases, per-6-brominated CD deriv- might have high affinity to the CD cavity [3,25,26] preventing
atives are the most common activated per-6-CDs. Their synthe- its departure from the reaction centres resulting in steric
sis can be performed on large scales under safe conditions. blocking. While the halogen → azide exchange required consid-
However, there can be some difficulties (compounds 3 and 4, erably larger halogen/reagent ratios than the solution reactions,
entries 2–4, 7–11 in Table 1) in reaction and during purification, sulfur nucleophiles showed a more favourable tendency and the
meaning that we therefore also decided to test per-6-chloro-β- reagent/halogen ratio was either roughly equal or slightly lower
CD from which the more ionic TU salt can be formed and the than for the monosubstitution. In the solution synthesis [6] of
chloride has less affinity to the macrocycle. It was not possible β-CD-per-6-dodecane thioether (6) the residual DMF increases
to reproduce the literature method [14] for per-6-chloro-β-CD the product solubility in methylene chloride and precipitation
synthesis, but a protocol using p-toluenesulfonyl chloride under with MeOH removes the unreacted 1-dodecanethiol (DDS),
reaction conditions that were analogous to the iodination/bromi- whereas the absence of DMF, together with the strong complex
nation reaction resulted in the targeted heptakis(6-chloro-6- formed between DDS and the product, meant that the product
deoxy)-β-CD being produced in good yields [20]. Not unex- isolation was more difficult. The resulting crude mixture was
pectedly, per-6-chlorinated-β-CD showed very poor reactivity, only partially soluble in methylene chloride and MeOH precipi-
not only under classic solvent reaction conditions (entry 14 in tation gave a difficult-to-filter product, which still contained at
Table 1), but also under ball milling (entries 5 and 15 in least one mole of complexed DDS. This complexation resulted
Table 1).
in not only technical difficulties, but also confounded the
removal of impurities. The strong complex between the reagent
Mechanochemical syntheses from 6I-O-monotosyl-β-CD and product also resulted in an impure product of the β-CD
usually require moderate reactant excesses and similar molar version of Sugammadex (5b). The reagent/halogen ratio was
ratios to the solution method [24]. Despite the expectations, practically identical to the one used in the solution method of
based on the monosubstituted case [24], the persubstitutions the originator [16] when mercaptopropionic acid reacted with
usually needed higher reagent/CD molar ratio. However, except the halogenated CD (entries 22, 23, and 26–28 in Table 1)
the azide cases (4a and 4b), the reagent/halogen molar ratios despite the use of the considerably milder and safer base, potas-
did not change such extent (Table 1). This can be explained sium tert-butoxide (KOt-Bu).
2366

Beilstein J. Org. Chem. 2016, 12, 2364–2371.
Table 1: Comparison of classic and green methods for the preparation of CD derivatives.
Entry Compound
Reagent/solvent
Reagent/
halogen
Method
Batch size
[mmol]
Yielda
[%]
Final temp. React. time
[°C]
[h]
molar ratio
1
2
3a
3b
thiourea/DMF
thiourea
2
1.5
3.5
3.5
3.5
2
soln.
BM
7.5
0.1
0.1
1
75
12
80
89
3 [27]
2
3
thiourea
BM
25
89
2
4
thiourea
BM
61
88
2
5
thioureab
BM
0.1
5
tracesc
90
92
2
6
thiourea/DMF
thiourea
soln.
BM
80
16
7
1.5
1.5
3.5
3.5
3.5
1.25
1.25
1.25
5
0.1
0.1
0.05
0.05
0.5
5
14c,d
9c,d
33d
39d
57d
90
85
2
8
thioureae
BM
82
2
9
thiourea
BM
85
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
thioureae
BM
82
2
thioureae
BM
86
2
4a
NaN3/DMF
NaN3/DMF
NaN3/DMFb
NaN3b
soln.
soln.
soln.
BM
100–105
100–105
100–105
86
~4.5
0.5
0.5
0.1
0.1
0.5
5
76
5
tracesc
tracesc
69
5 [24]
2
NaN3
5
BM
88
2
NaN3
5
BM
72
90
2
4b
5a
5b
NaN3/DMF
NaN3
1.25
10
soln.
BM
84
100–105
82
≈4.5
0.05
0.5
5
67
2
NaN3
10
BM
71
84
2
MPA/Cs2CO3/DMF
MPA/KOt-Buf
MPA/KOt-Bu
MPA/NaH/DMF
MPA/TEA/DMFf
MPA/KOt-Bu
MPA/KOt-Bue
MPA/KOt-Bue
DDS/K t-Bu/DMF
DDS/KOt-Bu/DMFg
DDS/KOt-Bug
1.5
1.43
1.5
1.25
3
soln.
BM
60
50
25 [28]
0.1
1
86
72
2
BM
71
75
2
soln.
soln.
BM
1.4
2.5
0.05
0.05
0.5
1
60
70
12 [16]
60c
81
60
24
1.25
1.5
1.5
no data
3
73
2
BM
86
71
2
2
BM
72d
>90
83
76
6
soln.
soln.
BM
80
96 [29]
120
2
0.5
0.1
80
1.6
95
62
aIsolated yields but due to small batch-sizes and not optimized purifications the yields of BM reactions are informative only. bFrom per-6-chloro-β-CD.
cContains also incompletely substituted structures, further isolation/purification were not performed. dThe mother liquor contained considerable
amounts of product by TLC. eFrom per-6-bromo-γ-CD. fMPA: 3-mercaptopropionic acid; TEA: N,N,N-triethylamine; KOt-Bu: potassium tert-butoxide;
KOt-Bu/MPA molar ratio ≈2.1:1.;gDDS: 1-dodecanethiol; DDS/KOt-Bu molar ratio 1:1.
Our first target was to investigate the reaction between a good to disulfide under basic conditions. The TU reactions (3a and
nucleophile (sulfur) and per-halogenated CDs. The TU method 3b) in the ball mill gave yields that were usually lower that
is widely used in the synthesis of various thiols and thioester those given by the monosubstitution reactions [22]. The very
compounds [8,30] and the CD thiouronium salt intermediates low aqueous solubility of the CD halides led to the intact
are not isolated [8] despite their good crystallization properties, starting materials being completely lost during the work-up.
while the excess/residual TU and used solvents are removed The low TU/halogen ratio (entries 2, 7 and 8 in Table 1) gave
upon the conversion to thiols only. However, the easy crystalli- incomplete reactions. Incompletely substituted derivatives were
zation of these salts is a nice feature of these compounds as it in the majority in the product, as can be seen in the multiple
can help to remove incompletely substituted compounds anomeric proton peaks in the 1H NMR spectra and the CH2-
(defected structures) and unused reagent as the thiolate is less halogen signals in the 2D NMR spectra. The larger excess of
stable after their basic decomposition; thiols are easy to oxidize TU gave the complete substitution of the halogen (entries 3, 4,
2367

Beilstein J. Org. Chem. 2016, 12, 2364–2371.
9–11 in Table 1), but the TU removal also caused difficulties on because the products are very poorly soluble in water and NaI
the hundred-milligram scale despite the crude products only and NaN3 can be readily separated. In order to accelerate the
contained the targeted derivative. The higher yields of the decomposition of the assumed NaI/CD complex first water then
10-times larger scale (entries 4 and 11in Table 1) clearly 50% aq EtOH was used as wetting substance. Usually 50%
demonstrate the technical difficulty of purification and the EtOH is able to decompose most of the CD complexes and
effect of batch sizes on yields. It was found that a considerable owing to the hydrogen bonding destruction increases the solu-
portion of TU-CD remained in the mother liquor because of the bility of poorly soluble CD derivatives. Wetting the solid
relatively high amounts of used solvents and the solubility of dispersion with water and 50% EtOH resulted in not only lower
thiouronium-CDs in EtOH. Solubility difficulties led to chro- temperature but practically no reaction was experienced as con-
matographic purification attempts failing even in RP-18 firmed by the lack of azide in the IR spectrum of the isolated
silicagel columns, too. The yield increased considerably when solid. As it is previously found [22] applying a wetting
bromine derivatives (entries 10 and 11 in Table 1) were used substance the grinding temperature is always lower than the dry
instead of their iodine analogues (entries 7 and 9 in Table 1). milling. It was assumed that the very low solubility of both per-
An identical TU/Br to TU/I ratio was used in case of 3b and halogeno and per-azido-CDs in water or 50% EtOH and com-
further reduced the yield (entry 8 in Table 1), demonstrating plete dissolution of the sodium azide the milling energy was not
that the similar halogen/tosyl ratio used in the monotosylated enough to warm the reaction mixture to an appropriate value.
case is not sufficient to lead to complete substitution. Owing to The very low solubility of the CD azides created an additional
the lower reactivity of bromo-CD a higher TU/per-6-bromo-γ- disadvantage and so we discarded the use of these wetting sol-
CD ratio had to be used. The more ionic character of formed vents. Alternatively, a wetting substance, which is an equally
thiouronium bromide was seen not only in the higher yields, but poor solvent for the CD derivatives and NaI, 1-pentanol [22]
also in the lower product content of the ethanolic mother liquor. was also tested. In this case, the final grinding temperature was
Chloride salts are even more ionic compounds whose character also found lower, approximately 20 °C lower, than the dry
can reduce the organic solvent solubility of thiouronium chlo- milling and no azido-CD was found by IR spectroscopy in the
rides. It, however, seemed reasonable to test a per-6-chlori- isolated solid.
nated CD. Unfortunately, practically no reaction was found to
occur and the reaction mixture contained only traces of incom- In the TU reactions, we found that the isolated yields depend on
pletely substituted TUβ-CD (entry 5 in Table 1). A further batch-size and we studied the effect of downsizing a solution
increase in the TU/chlorine ratio seemed to be unreasonable, as reaction for the preparation of compound 4a (entry 13 in
removing the higher amount of TU brings back the purification Table 1). In the downsized reaction due to the low solubility of
problems. The higher TU ratio somehow also increased the the NaN3 in DMF, an unreasonably high solvent/reagent ratio
yield in the β-CD version. Ten-fold scaling up of the experi- was necessary in the solution reaction and the yield was de-
ments showed increasing yields (entries 4 and 11 in Table 1) creased due to the technical difficulties in the purification, e.g.,
but the small scale still prevented to reach the solution reaction relatively larger loss upon mass transfers or filtrations. No
outcome but proved our concept. Gram-scale preparations essential differences in reaction time were found despite all the
easily overcome the technical difficulties of TU removal found necessary NaN3 being dissolved at the beginning of the reac-
in the small scale, as it becomes clear in the scale-up experi- tion and only a slightly longer reaction time was required to the
ments (entries 4 and 11 in Table 1).
complete reaction despite the higher dilution. Although in this
case the crude contained relatively less residual DMF but its
Although the preparation of 6I-monoazido-6I-monodeoxy-β-CD complete removal was still impossible. The scaled up ball mill
[22] was very effective, the analogue reactions with per-6- azidations (entries 17 and 20 in Table 1) resulted in little higher
substituted CDs (4a and 4b) were less efficient. Either the reac- yields but the relatively small amounts still caused technical
tion did not proceed at all or only partial substitution was difficulties which is shown in the yield-increasing ratios of the
achieved at low NaN3/halogen ratios, while only an increased β- and γ-CDs.
NaN3 ratio afforded the complete substitution of the CH2–I
groups, possibly because of the steric hindrance of the bulky so- Finally, the smaller solution reaction scale resulted in consider-
dium iodide. Iodine and metal iodides are preferred salts in ably reduced yields, which were however close to those of the
various CD complexes [26,31]. Lack of solvents the diffusion/ scaled-up mechanochemical method. The mechanochemical
decomplexation of NaI from the cavity is slow. The higher syntheses of 4a and 4b were successfully scaled up 10-fold
amounts of sodium azide can exclude the formed NaI from the (entries 17 and 20 in Table 1) with acceptable yields which
CD cavity in solid state. However, on a larger multigram scale provide the proof of concept for the ball-mill-assisted synthesis
the higher amount of used NaN3 can be easily regenerated of per-6-azido-CDs.
2368

Beilstein J. Org. Chem. 2016, 12, 2364–2371.
Chloride salts have considerably less affinity to the β-CD cavity tate the product from the acidic solution, which could be isolat-
(≈1/6 of iodide [26]). Per-6-chlorinated CD were also tested in ed by filtration, and then it was immediately converted to the
this reaction despite the fact that only traces of incompletely ammonium salt. Although, some cases the further simplified
substituted azido-CDs and decomposition products were found work-up resulted in somehow lower yields because of the redis-
in solution and that the reaction mixture composition was simi- solution of the protonated product during the filtration, the com-
lar to the solution composition (entries 14 and 15 in Table 1).
plete elimination of an organic solvent could be achieved.
Yields can be improved further at even larger scales.
In conclusion, it appears that the advantage of mechanochem-
istry is restricted to the elimination of high boiling solvents in The encouraging results with MPA led us to an attempt to
the azide cases (entries 16, 17, 19, and 20 in Table 1) which simplify the synthesis of an intermediate of a promising candi-
provides significant, if not dramatic, improvements in the syn- date for siRNA delivery. The solution phase synthesis of
thesis. However, it is also true that the price of sodium azide is heptakis(6-deoxy-6-S-alkyl)-β-CDs is typical for the syntheses
considerable lower than the costs of the utilization and regener- of intermediates of such compounds, however, their high
ation of polar aprotic organic solvents, including environmental lipophilicity means that isolation is not a technically trivial task.
impact.
Scale up of the 1-dodecanethiol mechanochemical reaction
The TU reactions (compounds 3a and 3b) demonstrated that a (synthesis of compound 6, entry 31 in Table 1) was not per-
good nucleophile, such as sulfur, could be effectively used in formed because the batch size was in the range of our solution
the preparation of useful per-6-thio-CD derivatives. While reaction. Mechanochemical synthesis was much more efficient
simple CD-6-thiols are still at the scientific stage, the sodium than conventional solution methods. The final temperature was
salt of octakis(6-deoxy-6-S-(3-mercapto)propionyl)-γ-CD has considerably lower (entry 31 vs 30 in Table 1) and the reaction
been slowly becoming an important surgical aid. The solution time was dramatically shorter. The lack of residual DMF
phase reaction is a dangerous process due to the use of sodium somehow changed the solubility behaviour of the product and
hydride. The relatively large number of possible side-reactions affected the purification to some extent. This green approach
in solution brings further challenges to production. The effi- resulted in better yields and purity despite the lower molar ratio
cient and green synthesis of Sugammadex is an exciting task. of the reagent DDS.
The comparable masses of reaction components meant that
reaction assembly had less influence on the yields in the prepa- The persubstituted species showed less variability in their
ration of 3-mercaptopropionyl derivatives (5a and 5b). While susceptibility to the heat effect of the ball milling than their
an inert wetting component (1-pentanol) was needed in the monosubstituted analogues. The temperatures inside the jar
monosubstituted analogue [22], no such component was neces- were considerably lower when one of the reactants was liquid or
sary in the per-6-substitution and the order in which the became liquid during the reaction (entries 22, 23, 26–28 and 31
reagents were added had no effect on the yield. Purification and in Table 1). In all cases, the temperature-time curves showed a
conversion to the pharmaceutically active form became very saturation-like trend (see Supporting Information File 1 for
simple, as the protonated form is poorly soluble in water: the details) and processes were generally in the 70–90 °C range at
precipitate formed upon acidification by HCl, was filtered and the end of the reaction, which is not essentially different from
the solid was re-dissolved in the equivalent amount of base. the monosubstituted case [18].
But, it is also true, that complex formation between the MPA
and the product, particularly in case of the β-CD version, Although limits caused by the intrinsic complexation properties
pointed out that pharmaceutical grade preparations need more of CDs sometimes affected the reaction rate, a solvent-free syn-
fine tuning not only in the reaction but in the purification thetic method may simplify the purification of compounds 5
method, as well. Although, Bridion® is a sodium salt of the and 6.
MPAγ-CD but in order to avoid the overdosing of base on the
small preparation scale in our experiments aqueous ammonia Conclusion
was chosen as base and acetone was needed to get less deficit in Syntheses of per-6-substituted CD derivatives can be effec-
the precipitates. Considerable losses during filtrations were tively carried out in a ball mill under solvent-free conditions. In
found in the small-scale cases, which significantly reduced the many cases, ball mill preparations display a positive balance in
yields because of the inevitably used diluted solutions.
cost-benefit analyses. Wet grinding for the preparation of per-6-
azido-CDs, using solubilizing and non-solubilizing solvents,
The ten-fold scale up (entry 28 in Table 1) further simplified the showed practically no reaction in the planetary ball mill. Impor-
product isolation because no acetone was necessary to precipi- tant intermediates and final products of per-6-amino- and per-6-
2369

Beilstein J. Org. Chem. 2016, 12, 2364–2371.
thio-CDs can be prepared without polar aprotic solvents, by
which the byproduct formation and difficult-to-remove impuri-
ties can be eliminated. The lack of solvents in the examples de-
scribed herein simplified the isolation and purification pro-
cesses. Our basic aim was to proof the concept and although the
purifications were not optimized the prepared compounds were
enough pure to record correct NMR spectra to identify the sub-
stitution location and completeness.
Supporting Information
Supporting Information File 1
Details of synthetic procedures and characterization of
prepared compounds.
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-12-230-S1.pdf]
Although in the monosubstituted case usually less reagent/
leaving group molar ratios were found [22], in the majority of
per-substitutions higher reagent/CD molar ratio was needed but
the reagent/halogen ratio not always changed dramatically. As
may be expected, the sulfur nucleophiles resulted in consider-
ably better or almost equal yields as compared to the conven-
tional solution methods. A potential drawback of the method
lies in the fact that the lack of highly solubilizing organic sol-
vents can cause difficulties in the primary stage of purification.
Acknowledgements
This work was funded by the University of Turin (Fondi
Ricerca Locale 2014). Part of this work was carried out by
Gabriele Caudera during his master thesis in Pharmacy.
References
1. Bender, M. L.; Komiyama, M. Cyclodextrin Chemistry. Reactivity and
Structure: Concepts in Organic Chemistry; Springer: Berlin-Heidelberg,
1978; Vol. 6. doi:10.1007/978-3-642-66842-5
2. Szejtli, J. Cyclodextrins and Their Inclusion Complexes; Akadémiai
Kiadó: Budapest, 1982.
Experimental
Full synthetic details and spectroscopic data are reported in
3. Szejtli, J.; Oda, T., Eds. Cyclodextrins. Comprehensive Supramolecular
Chemistry; Pergamon: New York, 1996; Vol. 3.
4. Astier, Y.; Braha, O.; Bayley, H. J. Am. Chem. Soc. 2006, 128,
1705–1710. doi:10.1021/ja057123+
Supporting Information File 1.
5. Sanderson, K. Nature 2008, 456, 23–25. doi:10.1038/456023a
6. Gooding, M.; Malhotra, M.; McCarthy, D. J.; Godinho, B. M. D. C.;
Cryan, J. F.; Darcy, R.; O’Driscoll, C. M. Eur. J. Pharm. Sci. 2015, 71,
80–92. doi:10.1016/j.ejps.2015.02.007
The syntheses of per-6-iodo-β- and -γ-CD, I7β-CD (2a) and I8γ-
CD (2b), were performed using a small modification to the
known method [13], from freshly dried CDs on a 0.01 mol scale
with triphenylphosphine and iodine in DMF. Per-6-bromo-γ-CD
(2b’) was prepared in N-methylpyrrolidone by the same method
using bromine. Per-6-chloro-β-CD (2a') was synthesized in a
similar manner to per-6-iodo-CDs using p-toluenesulfonyl chlo-
ride.
7. Zhang, M.-Q. Drug Discovery Today: Technol. 2010, 7, e131–e137.
doi:10.1016/j.ddtec.2010.07.002
8. Rojas, M. T.; Koeniger, R.; Stoddart, J. F.; Kaifer, A. E.
J. Am. Chem. Soc. 1995, 117, 336–343. doi:10.1021/ja00106a036
9. Martina, K.; Rinaldi, L.; Baricco, F.; Boffa, L.; Cravotto, G. Synlett 2015,
26, 2789–2794. doi:10.1055/s-0035-1560173
10.Rinaldi, L.; Binello, A.; Stolle, A.; Curini, M.; Cravotto, G. Steroids 2015,
98, 58–62. doi:10.1016/j.steroids.2015.02.016
General conditions for the solution reactions
Syntheses of compounds 3a, 3b, 4a, 4b, 5b and 6 were carried
out in DMF at 60–100 °C. For 5b, triethylamine was used as
base, while KOt-Bu was used for 6.
11.Cravotto, G.; Caporaso, M.; Jicsinszky, L.; Martina, K.
Beilstein J. Org. Chem. 2016, 12, 278–294. doi:10.3762/bjoc.12.30
12.Vizitiu, D.; Walkinshaw, C. S.; Gorin, B. I.; Thatcher, G. R. J.
J. Org. Chem. 1997, 62, 8760–8766. doi:10.1021/jo9711549
13.Gadelle, A.; Defaye, J. Angew. Chem., Int. Ed. Engl. 1991, 30, 78–80.
doi:10.1002/anie.199100781
General procedures for the high-energy ball
14.Khan, A. R.; D’Souza, V. T. J. Org. Chem. 1994, 59, 7492–7495.
doi:10.1021/jo00103a051
milling reactions
Syntheses of compounds 3a, 3b, 4a, 4b, 5b and 6 were carried
out in a Retsch PM100 High Speed Planetary Ball Mill.
1500 steel balls of 1 mm diameter (44.94 g) and 50 steel balls of
5 mm diameter (25.54 g, total weight of balls = 70.5 g,
V = 15 mL), were placed in a stainless steel jar of 50 mL, with a
sun wheel speed of 650 min–1 for 120 min, weight = 780 g (jar,
cap, and balls). Temperatures were measured using a Lafayette
TRI-88 no-contact thermometer, built-in laser pointer, with
±2 °C reading accuracy, distance to spot size = 8:1, measuring
distance 18–23 cm. The measurement matrix formed "a five on
a die", two measurements were made at each point and the
values were averaged.
15.No hits until 01/05/2016 in Science direct
(http://www.sciencedirect.com), Scifinder® (http://scifinder.cas.org), and
Web of ScienceTM (http://apps.webofknowledge.com).
16.Bom, A.; Bradley, M.; Cameron, K.; Clark, J. K.; van Egmond, J.;
Feilden, H.; MacLean, E. J.; Muir, A. W.; Palin, R.; Rees, D. C.;
Zhang, M.-Q. Angew. Chem. 2002, 114, 275–280.
doi:10.1002/1521-3757(20020118)114:2<275::AID-ANGE275>3.0.CO;
2-A
17.Wu, T.; Liu, Z.; Guo, Y.; Dong, C. J. Electroanal. Chem. 2015, 759,
137–143. doi:10.1016/j.jelechem.2015.11.005
18.Stolle, A.; Szuppa, T.; Leonhardt, S. E. S.; Ondruschka, B.
Chem. Soc. Rev. 2011, 40, 2317–2329. doi:10.1039/C0CS00195C
2370

Beilstein J. Org. Chem. 2016, 12, 2364–2371.
19.Beyer, M. K.; Clausen-Schaumann, H. Chem. Rev. 2005, 105,
2921–2948. doi:10.1021/cr030697h
20.Hernández, J. G.; Avila-Ortiz, C. G.; Juaristi, E. Useful Chemical
Activation Alternatives in Solvent-Free Organic Reactions. In
Comprehensive Organic Synthesis, 2nd ed.; Molander, G. A.;
Knochel, P., Eds.; Elsevier BV, 2014; pp 287–314.
doi:10.1016/B978-0-08-097742-3.00935-6
21.Ranu, B. C.; Chatterjee, T.; Mukherjee, N. Carbon–Heteroatom Bond
Forming Reactions and Heterocycle Synthesis under Ball Milling. In
Ball Milling Towards Green Chemistry; Stolle, A.; Ranu, B., Eds.; RSC
Green Chemistry; Royal Society of Chemistry: Cambridge, U.K., 2014;
pp 1–33. doi:10.1039/9781782621980-00001
22.Jicsinszky, L.; Caporaso, M.; Tuza, K.; Martina, K.; Calcio Gaudino, E.;
Cravotto, G. ACS Sustainable Chem. Eng. 2016, 4, 919–929.
doi:10.1021/acssuschemeng.5b01006
23.Menuel, S.; Doumert, B.; Saitzek, S.; Ponchel, A.; Delevoye, L.;
Monflier, E.; Hapiot, F. J. Org. Chem. 2015, 80, 6259–6266.
doi:10.1021/acs.joc.5b00697
24.Tuza, K. Synthesis of Cyclodextrin based C. Perfringens Antidotes.
(Hungarian), Master's Thesis, Univ. L. Eötvös, Budapest, Hungary,
2011.
25.Matsui, Y.; Ono, M.; Tokunaga, S. Bull. Chem. Soc. Jpn. 1997, 70,
535–541. doi:10.1246/bcsj.70.535
26.Rohrbach, R. P.; Rodriguez, L. J.; Eyring, E. M.; Wojcik, J. F.
J. Phys. Chem. 1977, 81, 944–948. doi:10.1021/j100525a003
27.Uccello-Barretta, G.; Evangelisti, C.; Balzano, F.; Vanni, L.; Aiello, F.;
Jicsinszky, L. Carbohydr. Res. 2011, 346, 753–758.
doi:10.1016/j.carres.2011.02.001
28.Adam, J. M.; Bennett, D. J.; Bom, A.; Clark, J. K.; Feilden, H.;
Hutchinson, E. J.; Palin, R.; Prosser, A.; Rees, D. C.; Rosair, G. M.;
Stevenson, D.; Tarver, G. J.; Zhang, M.-Q. J. Med. Chem. 2002, 45,
1806–1816. doi:10.1021/jm011107f
29.Mazzaglia, A.; Donohue, R.; Ravoo, B. J.; Darcy, R. Eur. J. Org. Chem.
2001, 1715–1721.
doi:10.1002/1099-0690(200105)2001:9<1715::AID-EJOC1715>3.0.CO
;2-A
30.Eccles, K. S.; Elcoate, C. J.; Lawrence, S. E.; Maguire, A. R. ARKIVOC
2010, No. ix, 216–228.
31.Mochida, K.; Kagita, A.; Matsui, Y.; Date, Y. Bull. Chem. Soc. Jpn.
1973, 46, 3703–3707. doi:10.1246/bcsj.46.3703
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(http://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
doi:10.3762/bjoc.12.230
2371