A highly reusable rhodium catalyst-organic framework for the intramolecular cycloisomerization of 1,6-enynes

Article abstract of DOI:10.1021/ol201333s

The intramolecular cycloisomerization of 1,6-enynes in 95-99% ee is reported using an immobilized Rh catalyst-organic framework synthesized from alternating ring-opening metathesis polymerization (altROMP) assembly. The framework was reused up to seven times, and it was used in high turnover number (TON) batch reactions. The catalyst provided the highest TONs to date (up to 890) for the cycloisomerizations, with catalyst loadings ranging from 0.2 to 0.06 mol %.

Full text of DOI:10.1021/ol201333s

ORGANIC
LETTERS
2011
Vol. 13, No. 13
3522–3525
A Highly Reusable Rhodium Catalyst-
Organic Framework for the Intramolecular
Cycloisomerization of 1,6-Enynes
Elizabeth G. Corkum, Michael J. Hass, Andrew D. Sullivan, and Steven H. Bergens*
Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada T6G 2G2
steve.bergens@ualberta.ca
Received May 17, 2011
ABSTRACT
The intramolecular cycloisomerization of 1,6-enynes in 95ꢀ99% ee is reported using an immobilized Rh catalyst-organic framework synthesized
from alternating ring-opening metathesis polymerization (altROMP) assembly. The framework was reused up to seven times, and it was used in
high turnover number (TON) batch reactions. The catalyst provided the highest TONs to date (up to 890) for the cycloisomerizations, with catalyst
loadings ranging from 0.2 to 0.06 mol %.
We report a highly reusable, immobilized, catalyst-or-
ganic framework fortheintramolecular cycloisomerization
of 1,6-enynes. The Rh-catalyzed cycloisomerization of 1,
6-enynes, discovered by Zhang et al. in 2000,1 is an
extremely versatile CꢀC bond-forming reaction that has
produced a wide range of furans, lactones, lactams, cyclo-
pentanes, and cyclopentanones.2,3 These products have
numerous pharmaceutical applications and, remarkably,
are almost exclusively produced in >99% ee by cyclo-
isomerization.4 Most reports of Rh-catalyzed cycloisome-
rizations utilize 10 mol % [Rh(BINAP)]þ (BINAP = 2,
20-bis(diphenylphosphino)-1,10-binaphthyl) as the homo-
geneous catalyst, with20 mol % AgSbF6 asan activator, in
dichloroethane (DCE) solvent. In fact, most examples of
nontandem enantioselective cycloisomerizations occur
with 5ꢀ10 mol % of chiral catalyst regardless of the
(1) (a) Cao, P.; Wang, B.; Zhang, X. J. Am. Chem. Soc. 2000, 122,
6490. (b) Cao, P.; Zhang, X. Angew. Chem., Int. Ed. 2000, 39, 4104.
^
(2) For recent reviews, see: (a) Michelet, V.; Toullec, P. Y.; Genet,
J.-P. Angew. Chem., Int. Ed. 2008, 47, 4268. (b) Zhang, Z.; Zhu, G.;
Tong, X.; Wang, F.; Xie, X.; Wang, J.; Jiang, L. Curr. Org. Chem. 2006,
10, 1457. (c) Modern Rhodium-Catalyzed Organic Reactions; Evans,
P. A., Ed.; Wiley-VCH: Weinheim, Germany, 2005.
(3) (a) Lei, A.; Waldkirch, J. P.; He, M.; Zhang, X. Angew. Chem.,
Int. Ed. 2002, 41, 4526. (b) Lei, A.; He, M.; Wu, S.; Zhang, X. Angew.
Chem., Int. Ed. 2002, 41, 3457. (c) Hashmi, A. S. K.; Haufe, P.; Nass,
A. R. Adv. Synth. Catal. 2003, 345, 1237. (d) Lei, A.; He, M.; Zhang, X.
J. Am. Chem. Soc. 2003, 125, 11472. (e) Tong, X.; Zhang, Z.; Zhang, X.
J. Am. Chem. Soc. 2003, 125, 6370. (f) Tong, X.; Li, D.; Zhang, Z.;
Zhang, X. J. Am. Chem. Soc. 2004, 126, 7601. (g) Hashmi, A. S. K.;
Haufe, P.; Nass, A. R.; Bats, J. W. Adv. Synth. Catal. 2004, 346, 421. (h)
Mikami, K.; Kataoka, S.; Yusa, Y.; Aikawa, K. Org. Lett. 2004, 6, 3699.
(i) Mikami, K.; Kataoka, S.; Aikawa, K. Org. Lett. 2005, 7, 5777. (j)
Mikami, K.; Kataoka, S.; Wakabayashi, K.; Aikawa, K. Tetrahedron
Lett. 2006, 47, 6361. (k) Liu, F.; Liu, Q.; He, M.; Zhang, X.; Lei, A. Org.
Biomol. Chem. 2007, 5, 3531. (l) Nishimura, T.; Kawamoto, T.; Nagaosa,
M.; Kumamoto, H.; Hayashi, T. Angew. Chem., Int. Ed. 2010, 49, 1638.
(m) Wang, J.; Xie, X.; Ma, F.; Peng, Z.; Zhang, L.; Zhang, Z. Tetra-
hedron 2010, 66, 4212. (n) Kim, S. Y.; Chung, Y. K. J. Org. Chem. 2010,
75, 1281.
(4) (a) Lei, A.; He, M.; Zhang, X. J. Am. Chem. Soc. 2002, 124, 8198.
(b) He, M.; Lei, A.; Zhang, X. Tetrahedron Lett. 2005, 46, 1823. (c)
Nicolaou, K. C.; Edmonds, D. J.; Li, A.; Tria, G. S. Angew. Chem., Int.
Ed. 2007, 46, 3942. (d) Nicolaou, K. C.; Li, A.; Ellery, S. P.; Edmonds,
D. J. Angew. Chem., Int. Ed. 2009, 48, 6293.
(5) Zhang et al. have reported lower loadings of [Rh(BINAP)]þ (as
low as 0.4 mol%) in the nontandem cycloisomerization of the methyl
enyne (X = Me). For more information, see ref 3b.
r
10.1021/ol201333s
Published on Web 05/31/2011
2011 American Chemical Society

metal.2,5 Some tandem reactions, however, occur with
lower loadings of catalyst presumably because product
inhibition and isomerization are avoided.2a,4a These rela-
tively high loadings of catalyst and additives narrow the
industrial applications of these systems. We reasoned that
immobilization of the catalyst would provide higher turn-
over numbers (TONs) by allowing catalyst reuse, thereby
avoiding product inhibition and isomerization that may
occur during high TON homogeneous reactions. Further,
the use of a heterogeneous catalyst would allow for a wide
rangeof solvents asthe solvent wouldno longerbe required
to dissolve the catalyst.
polymerization (altROMP) between trans-Ru(2)Cl2Py2 (1,
where 2 = (R)-5,50-dinorimido-BINAP14) and cis-cyclooc-
tene (COE) using RuCl2(=CHPh)(PCy3)2 (3) as a catalyst to
assemble a catalyst-organic framework cross-linked with
1.15,16 This Ru-based catalyst-organic framework was reused
for an enantioselective ketone hydrogenation with high TONs
per run (g1000) for 25 runs before loss in activity occurred.
More than 35 reuses were carried out without loss in enantios-
electivity or detectable leaching of Ru. We now report the
preparation of a Rh-BINAP-containing catalyst-organic
framework and its reuse as a catalyst for the enantioselective
cycloisomerization of 1,6-enynes.
Two general approaches to separate the catalyst and reac-
tant phases in enantioselective catalysis are the use of multi-
phase solvent mixtures (e.g., aqueous/organic,6 supercritical
CO2,7 fluorous catalysts and solvents,8 and ionic liquids9) and
catalyst immobilization. Homogeneous catalysts are immobi-
lized through noncovalent and covalent interactions with a
support. The noncovalent interactions10 include electrostatic
attractions between ionic catalysts and supports, physisorp-
tion onto the support, hydrogen bonding, and encapsulation
within the support. Covalent immobilization techniques11
include the formation of metalꢀsupport bonds and formation
of bonds between a modified ligand and an inorganic or
organic support. Copolymerization of modified catalyst li-
gands or grafting modified ligands onto polymeric supports is
often carried out by incorporating the ligand into the support
first, followed by metalating the incorporated ligand. The
result is often incomplete metalation of the ligand-polymer
support and poor mass transport at the active sites, leading to
low catalyst activity and poor reusability. The direct polymer-
ization of metal-containing monomers (MCM) ensures that
ligand-containing sites on the polymer are metalated,12 but
mass transport at the active sites must still be addressed. We
recently reported13 an alternating ring-opening metathesis
We synthesized the chloro-bridged Rh complex
[RhCl(2)]2 (4) by reacting 2 with [RhCl(C2H4)2]2 in
CH2Cl2.17,18 4 was used immediately for altROMP with-
out isolation to avoid its decomposition.
We found that 4 readily underwent altROMP assembly
with COE and 3 (3/4/COE = 1:10:120, 40 °C, 24 h, CH2Cl2)
(14) 2 was prepared in one step from the known diamine precursor.
See: (a) Okano, T. K. H.; Akutagawa, S.; Kiji, J.; Konishi, H.;
Fukuyama, K.; Shimano, Y. U.S. Patent 4,705,895, 1987. (b) Berthod,
M.; Mignani, G.; Woodward, G.; Lemaire, M. Chem. Rev. 2005, 105, 1801.
(15) For examples of BINAP-containing polymers, see: (a) ter Halle,
R.; Colasson, B.; Schulz, E.; Spagnol, M.; Lemaire, M. Tetrahedron
Lett. 2000, 41, 643. (b) Saluzzo, C.; Lamouille, T.; Le Guyader, F.;
Lemaire, M. Tetrahedron: Asymmetry 2002, 13, 1141. (c) Fan, Q. H.;
Ren, C. Y.; Yeung, C. H.; Hu, W. H.; Chan, A. S. C. J. Am. Chem. Soc.
1999, 121, 7407. (d) Fan, Q. H.; Wang, R.; Chan, A. S. C. Bioorg. Med.
Chem. Lett. 2002, 12, 1867. (e) Fan, Q. H.; Deng, G. J.; Lin, C. C.; Chan,
A. S. C. Tetrahedron: Asymmetry 2001, 12, 1241. (f) Pu, L. Chem.;Eur.
J. 1999, 5, 2227. (g) Yu, H. B.; Hu, Q. S.; Pu, L. Tetrahedron Lett. 2000,
41, 1681. (h) Framery, E.; Andrioletti, B.; Lemaire, M. Tetrahedron:
Asymmetry 2010, 21, 1110. (i) Saluzzo, C.; Lemaire, M. Adv. Synth.
Catal. 2002, 344, 915. (j) Hu, A.; Ngo, H. L.; Lin, W. B. J. Am. Chem.
Soc. 2003, 125, 11490. (k) Org. Lett. 2004, 6, 2937. (l) Wang, P.; Liu, X.;
Yang, J.; Yang, Y.; Zhang, L.; Yang, Q.; Li, C. J. Mater. Chem. 2009, 19,
(6) (a) Dwars, T.; Oehme, G. Adv. Synth. Catal. 2002, 344, 239. (b)
ꢀ
Joo, F. Acc. Chem. Res. 2002, 35, 738. (c) Pinault, N.; Bruce, D. W.
Coord. Chem. Rev. 2003, 241, 1.
(7) Gordon, C. M.; Leitner, W. Homogeneous Catalysis in Super-
critical Solvents as a Special Unit Operation. In Multiphase Homoge-
neous Catalysis; Cornils, B., Herrmann, W. A., Vogt, D., Horvath, I., Olivier-
Bourbigon, H., Leitner, W., Mecking, S., Eds.; Wiley-VCH: Weinheim,
Germany, 2005.
(8) (a) Handbook of Fluorous Chemistry; Gladysz, J. A., Curran, D. P.,
ꢀ
Horvath, I., Eds.; Wiley-VCH: Weinheim, Germany, 2004. (b) de Wolf, E.;
ꢀ
van Koten, G.; Deelman, B.-J. Chem. Soc. Rev. 1999, 28, 37. (c) Horvath,
I. T. Acc. Chem. Res. 1998, 31, 641.
(9) (a) Welton, T. Coord. Chem. Rev. 2004, 248, 2459. (b) Sheldon, R.
Chem. Commun. 2001, 2399. (c) Wasserscheid, P.; Keim, W. Angew.
Chem., Int. Ed. 2000, 39, 3772.
(10) For recent reviews, see: (a) Fraile, J. M.; Garcıa, J. I.; Mayoral,
J. A. Chem. Rev. 2009, 109, 360. (b) Heitbaum, M.; Glorius, F.; Escher, I.
Angew. Chem., Int. Ed. 2006, 45, 4732. (c) McMorn, P.; Hutchings, G. J.
Chem. Soc. Rev. 2004, 33, 108. (d) Zhao, X. S.; Bao, X. Y.; Guo, W.; Lee,
F. Y. Mater. Today 2006, 9, 32.
^
8009. (m) Guerreiro, P.; Ratovelomanana-Vidal, V.; Genet, J.-P.; Dellis,
P. Tetrahedron Lett. 2001, 42, 3423.
(16) For examples of use of ROMP to prepare polymeric catalysts,
see: (a) Chen, B.; Sleiman, H. F. Macromolecules 2004, 37, 5866. (b)
Fuchter, M. J.; Hoffman, B. M.; Barrett, A. G. M. J. Org. Chem. 2006,
71, 724. (c) Sankaran, N. B.; Rys, A. Z.; Nassif, R.; Nayak, M. K.;
Metera, K.; Chen, B.; Bazzi, H. S.; Sleiman, H. F. Macromolecules 2010,
43, 5530. (d) Carlisle, J. R.; Wang, X.-Y.; Weck, M. Macromolecules
2005, 38, 9000. (e) Dragutan, V.; Dragutan, I.; Fischer, H. J. Inorg.
Organomet. Polym. 2008, 18, 18. (f) Dragutan, V.; Dragutan, I.; Fischer,
H. J. Inorg. Organomet. Polym. 2008, 18, 311.
(11) For recent reviews, see: (a) Dioos, B. M. L.; Vankelecom, I. F. J.;
Jacobs, P. A. Adv. Synth. Catal. 2006, 348, 1413. (b) Leadbeater, N. E.;
Marco, M. Chem. Rev. 2002, 102, 3217. (c) Fan, Q.-H.; Li, Y.-M.; Chan,
A. S. C. Chem. Rev. 2002, 102, 3385. (d) Wang, Z.; Chen, G.; Ding, K.
Chem. Rev. 2009, 109, 322. (e) Ding, K.; Wang, Z.; Wang, X.; Liang, Y.;
Wang, X. Chem.;Eur. J. 2006, 12, 5188. For MOFs, see:(f) Liu, Y.;
Xuan, W.; Cui, Y. Adv. Mater. 2010, 22, 4112. (g) Corma, A.; Garcıa, H.;
(17) Tokunaga, N.; Yoshida, K.; Hayashi, T. Proc. Natl. Acad. Sci.
U.S.A. 2004, 101, 5445.
(18) 2 is initially formed as a mixture of three NMR distinct diaster-
eomeric atropisomers that differ in the rotameric orientations of the
norimido groups about the areneꢀnitrogen bonds. However 2 precipi-
tated from 90 °C solutions of toluene as one C2 dissymmetric diastereo-
meric atropisomer, which was then used in the synthesis of 4. See ref 13
for more details.
ꢀ
Llabres i Xamena, F. X. Chem. Rev. 2010, 110, 4606. (h) Ma, L.; Abney,
C.; Lin, W. B. Chem. Soc. Rev. 2009, 38, 1248.
(12) Mastrorilli, P.; Nobile, C. F. Coord. Chem. Rev. 2004, 248, 377.
(13) Ralph, C. K.; Bergens, S. H. Organometallics 2007, 26, 1571.
Org. Lett., Vol. 13, No. 13, 2011
3523

to form the catalyst-organic framework 5. We found that
ROMP of 4 did not readily occur in the absence of COE,
likely because of steric crowding around the endo norimido
groups. We believe that altROMP with COE proceeds by
reaction between a norimido group in 4 and 3 to form a Ru-
alkylidene that is too crowded to react with another molecule
of 4. Instead, this alkylidene reacts with COE to insert a C8
spacer and form a less crowded Ru-alkylidene group that
reacts with 4and so on. The difference in ring strain and steric
crowding between the olefin groups in 4 and COE results in
AgSbF6 in DCE at room temperature (TON = 10).3c The
first cycloisomerization of 7 over 5/BaSO4 was carried out
with 5 mol % Rh, 10 mol % AgSbF6 in 1,4-dioxane at
60 °C and was complete in 99% yield and in 99% ee after
3 h. Significantly, the catalyst-organic framework 5/BaSO4
was isolated and reused for four additional runs (five total)
with loadings of 7/Rh = 100:1 for each run, without
additional AgSbF6 (Table 1). Although it was necessary
to increase time and temperature for each run, a total TON
of 380 was obtained in 99% ee. The cyclohexyl enyne 6 was
more active toward 5/BaSO4. After the first run over 5/
BaSO4 (5 mol % Rh, 10 mol % AgSbF6 in 1,4-dioxane at
60 °C for 3 h), six more reuses with loadings of 6/Rh =
100:1 (50ꢀ65 °C, and >98% yield/run) were carried out to
give a total TON of 620 in >95% ee (minor enantiomer
not detected by NMR shift reagent).
1
the altROMP assembly. H NMR spectra recorded in situ
showed that the ratio of COE to norimido unit consumed
during the altROMP of 4was ∼1.7:1. We previously reported
that the corresponding ratio for the altROMP of the Ru-
MCM 1 and COE was 1:1.13 We attribute this difference to
the dimeric nature of 4; specifically, the bent orientation of the
chloro bridges in 4 engenders more steric crowding upon
the norimido groups in 4 than in 1. The norimido groups in 4
are thereby of lower net reactivity toward altROMP than
those in 1, resulting in incorporation of a higher number of
Table 1. Reuse Results for the Cycloisomerization of 6a,b and
7a,c
1
COE spacers into the catalyst-organic framework 5. H
NMR spectroscopy showed that no free norimido groups
remained at the end of the altROMP and that a mixture of
cis- and trans-olefin groups was present in the framework.
a Run 1 was carried out in 0.1 M solutions of 6 and 7 respectively in
1,4-dioxane under the following conditions: Sub/Ag/Rh = 20/2/1, 60
°C. All subsequent runs were carried out in 0.2 M solutions of 6 and 7
respectively without any additional AgSbF6 added under the following
conditions: Sub/Rh = 100/1. b % ee determined by NMR shift reagent.
c % ee determined by chiral GC.
The 31P NMR spectrum of 5 consisted of a broad signal
centered on the chemical shift for the dimer 4, showing that
the electronic environment of the Rh centers was not
significantly altered by the altROMP assembly. We note
that, unlike the catalyst-organic framework containing 1,
framework 5 is doubly cross-linked by norimido-BINAP
groups and by the chloro-bridged Rh centers. The frame-
work 5 was deposited as a thin film onto BaSO4 to provide
mechanical stability and to aid mass transport during
catalysis.
For a comparison, we attempted the homogeneous
cycloisomerization of cyclohexyl enyne 6 with 2.5 mol %
[RhCl(BINAP)]2, 10 mol % AgSbF6 in 1,4-dioxane at
40 °C. The reaction was complete after 2 h (TON = 10)
to give a complex mixture that likely formed by isomeriza-
tion of the product. The same reaction carried out in DCE
(5 mol % [RhCl(COD)]2, 10 mol % (R)-BINAP, 20 mol %
AgSbF6) was complete after 5 min (TON = 10) at room
temperature to also give a complex mixture of isomerized
products. To our knowledge, these are the first enantiose-
lective catalytic reactionsinwhichthe immobilizedcatalyst
provides higher TONs and is more selective than the
parent, homogeneous catalyst.
The intramolecular cycloisomerization of 1,6-enynes 6
and 7 was used to evaluate 5. To our knowledge, the
cycloisomerization of 1,6-enyne 6 has not been reported.
Table 2 summarizes the results from high loading batch
reactions without reuse in 1,4-dioxane and in 2-MeTHF
with AgSbF6/Rh = 5:1 at 70 °C. These cycloisom-
erizations were best carried out by allowing AgSbF6 and
The reported homogeneous cycloisomerization of phe-
nyl enyne 7occurs with 10 mol % [Rh(BINAP)]þ, 20mol%
3524
Org. Lett., Vol. 13, No. 13, 2011

Table 2. Batch Reactivity Results of the Cycloisomerization of
6a,b and 7a,c
entry sub solvent loading (sub/Ag/Rh) time (h) TON % ee
2
45
4
200 >95
800 >95
630 >95
500 >95
480 >99
890 >99
1b
6
dioxane
1000/5/1
Table 3. Results of the Cycloisomerization of 8
2b
3b
4c
5c
6
6
7
7
MeTHF
MeTHF
dioxane
MeTHF
1000/5/1
500/5/1
500/5/1
1600/5/1
loading
2
entry
(sub/Ag/Rh)
time (h)
TON
% ee
23
20
1a
2a
100/5/1
300/5/1
2.5
48
99
>99
>99
a All runs were carried out at 70 °C. b % ee determined by NMR shift
reagent. c % ee determined by chiral GC.
285
a Both runs were carried out at 70 °C in 0.9 mL of 1,4-dioxanes. % ee
was determined by chiral GC.
5/BaSO4 to react for 30 min at 40 °C to aid framework
swelling and chloride abstraction prior to introduction of
substrate. The cycloisomerization of the cyclohexyl enyne
6 was faster in 2-MeTHF, giving a TON of 500 after 2 h
compared to 200 after 2 h in 1,4-dioxane (entries 1 and 3).
The total TON (800 vs 630, entries 1 and 2), however, was
greater in 1,4-dioxane. These differences likely result from
the lower coordinating ability of 2-MeTHF compared to
1,4-dioxane. The phenyl enyne 7 reacted to give a TON of
480 in 1,4-dioxane (entry 4) and 890 in 2-MeTHF (entry 5).
To our knowledge, these are thehighest reportedTONsfor
any cycloisomerization. The % ee’s were the same in both
solvents.
We have presented the altROMP assembly of the
BaSO4 supported catalyst-organic framework 5 and
its high activity toward the cycloisomerization of 1,6-
enynes. The immobilized catalyst provided higher TONs
and, in some cases, was more selective than the parent,
homogeneous systems. The higher TONs and selectivity
are rare, if not unique, in the field. Their possible origins
include support (BaSO4) effects, catalyst-framework in-
teractions, and encapsulation phenomena. Current re-
search is focused upon optimizing the structure of the
catalyst-organic framework 5 and optimizing reaction
conditions for cycloisomerization and other reactions, as
well as investigating the origins of its activity and selectiv-
ity through controlled structural modifications and ex-
tensive characterization.
Zhang reported the elegant tandem cycloisomerization
of 8 to form the R-methylene-γ-butyrolactone 9, which can
be converted in two steps into (þ)-pilocarpine.4a (þ)-
Pilocarpine is a leading pharmaceutical in the treatment
€
of narrow- or wide-angle glaucoma as well as Sjogren
Acknowledgment. This work was supported by the Na-
tional Sciences and Engineering Research Council of
Canada (NSERC) and the University of Alberta. We
greatly appreciate the assistance of Wayne Moffat and
the University of Alberta Analytical and Instrumentation
Laboratory for GC-MS and polarimetry analysis, as well
as undergraduate researcher Hye-lin Kim.
syndrome.19 Using 5/BaSO4 (1 mol % Rh) in one batch,
we obtained 100% conversion and 99% yield for the
cycloisomerization of 8 in 1,4-dioxane at 70 °C (Table 3,
entry 1). At 0.33 mol % Rh, the reaction was 90%
complete after 24 h and 100% complete after 48 h with
5% isomerization byproducts (entry 2). The catalyst 5/
BaSO4 was also active in 2-MeTHF but gave a mixture of
products likely resulting from isomerization.
Supporting Information Available. Text and figures
giving experimental procedures for substrates, ligand,
and catalyst-organic framework preparation and use.
This material is available free of charge via the Internet
at http://pubs.acs.org.
ꢀ
(19) (a) Ramos-Casals, M.; Tzioufas, A. G.; Stone, J. H.; Siso, A.;
Bosch, X. JAMA 2010, 304, 452. (b) Diestelhorst, M.; Nordmann, J.-P.;
Toris, C. B. Surv. Ophthalmol. (Suppl. 1) 2002, 47, S155. (c) Worthen,
D. M.; Zimmerman, T. J.; Wind, C. A. Invest. Ophthalmol. 1974, 13, 296.
(d) Nusair, S.; Rubinow, A. Semin. Arthritis Rheum. 1999, 28, 360.
Org. Lett., Vol. 13, No. 13, 2011
3525