Interfacial electron transfer to the zeolite-encapsulated methylviologen acceptor from various carbonylmanganate donors. Shape selectivity of cations in mediating electron conduction through the zeolite framework

Article abstract of DOI:10.1021/ja962645y

The series of (one-electron) reductions of methylviologen (MV²⁺) intercalated into zeolite-Y by various carbonylmanganate donors [C⁺Mn(CO)₄L^-, L = CO, P(OPh)₃] are very selective and highly dependent on the size/shape of the counterion C⁺, although the same electron transfers carried out (homogeneously) in solution always occur spontaneously, irregardless of C⁺. For example, the complete reduction of MV²⁺ extensively doped into zeolite-Y proceeds rapidly and quantitatively when the Na⁺ salts of the carbonylmanganates are employed as the reductants, but only to a very limited extent (1%) when the large PPN⁺ [bis(triphenylphosphine)iminium] salts of the carbonylmanganates are employed. The medium-size tetraethylammonium (TEA⁺) salt of Mn(CO)₄P(OPh)₃^- slowly effects an intermediate conversion (80%). Based on the fact that the large phosphite-substituted Mn(CO)₄P(OPh)₃^- donor cannot enter the supercage of zeolite-Y, we propose interfacial electron transfer from the carbonylmanganate to the MV²⁺ acceptor to occur only at the zeolite periphery. Importantly, the strong dependence of the further progress of the redox reaction with decreasing size of the cation C⁺ (i.e., shape selectivity) predicts that electron conduction throughout the zeolite framework requires the simultaneous transport of these cations in order to effect the complete reduction of all the encapsulated MV²⁺, as presented in Chart 5.

Full text of DOI:10.1021/ja962645y

12710
J. Am. Chem. Soc. 1996, 118, 12710-12718
Interfacial Electron Transfer to the Zeolite-Encapsulated
Methylviologen Acceptor from Various Carbonylmanganate
Donors. Shape Selectivity of Cations in Mediating Electron
Conduction through the Zeolite Framework
K. B. Yoon,* Y. S. Park, and J. K. Kochi*
Contribution from the Departments of Chemistry, Sogang UniVersity, Seoul 121-742, Korea,
and UniVersity of Houston, Houston, Texas 77204-5641
ReceiVed July 30, 1996. ReVised Manuscript ReceiVed October 7, 1996X
Abstract: The series of (one-electron) reductions of methylviologen (MV2+) intercalated into zeolite-Y by various
carbonylmanganate donors [C+Mn(CO)4L-, L ) CO, P(OPh)3] are very selective and highly dependent on the size/
shape of the counterion C+, although the same electron transfers carried out (homogeneously) in solution always
occur spontaneously, irregardless of C+. For example, the complete reduction of MV2+ extensively doped into
zeolite-Y proceeds rapidly and quantitatively when the Na+ salts of the carbonylmanganates are employed as the
reductants, but only to a very limited extent (1%) when the large PPN+ [bis(triphenylphosphine)iminium] salts of
-
the carbonylmanganates are employed. The medium-size tetraethylammonium (TEA+) salt of Mn(CO)4P(OPh)3
slowly effects an intermediate conversion (80%). Based on the fact that the large phosphite-substituted
-
Mn(CO)4P(OPh)3 donor cannot enter the supercage of zeolite-Y, we propose interfacial electron transfer from the
carbonylmanganate to the MV2+ acceptor to occur only at the zeolite periphery. Importantly, the strong dependence
of the further progress of the redox reaction with decreasing size of the cation C+ (i.e., shape selectivity) predicts
that electron conduction throughout the zeolite framework requires the simultaneous transport of these cations in
order to effect the complete reduction of all the encapsulated MV2+, as presented in Chart 5.
Introduction
interesting question as to how interfacial electron transfer
between an electron donor (dissolved in solution) and an electron
acceptor (immobilized in an insoluble matrix) is brought about.
In a more general context, this mechanistic problem is also
encountered in the oxidation-reduction of such diverse systems
as an active center deeply embedded in a metalloenzyme or a
metal-doped microheterogeneous catalyst, in which one of the
partners of the redox couple does not enjoy the usual diffusive
mobility inherent to homogeneous electron transfers. As such,
we now inquire as to what structural factors control the
interfacial electron transfer in order to identify the sequence of
steps by which the electron is transported from the bulk solution
into the interior of the solid phase. To address these questions,
we rely on (a) crystalline zeolites to provide a well-defined steric
environment for the encapsulated (cationic) acceptor and (b)
anionic donors A- of varying shapes which coupled with
differently sized counterions C+ provide a series of graded salts
[C+A-] to modulate the interfacial electron transfer.
Our initial model system as described in this study is based
on zeolite-Y as the molecular compartment, and methylviologen
(MV2+ ) N,N′-dimethyl-4,4′-bipyridinium) is selected as the
prototypical electron acceptor for the following reasons. First,
the physicochemical properties of MV2+, including its redox
behavior in various solutions, have been extensively studied.13-17
For example, the colorless methylviologen (λmax ∼ 260 nm)13
turns intense blue (molar extinction coefficient at 605 nm is
Zeolites are microcrystalline hosts to a wide variety of
different guests, including both neutral molecules as well as
charged ions.1,2 Indeed, the ready cation exchanges of the
widely available sodium zeolites have led to a series of
interesting nanostructures extensively doped with (complexed)
inorganic and organic cations.3-12 Such an intercalation of
redox-active cations offers us the opportunity to examine the
X Abstract published in AdVance ACS Abstracts, December 1, 1996.
(1) (a) Breck, D. W. Zeolite Molecular SieVes; Wiley: New York, 1974.
(b) Barrer, R. M. Zeolites and Clay Minerals as Sorbents and Molecular
SieVes: Academic: London, 1978.
(2) Yoon, K. B. Chem. ReV. 1993, 93, 321.
(3) (a) Yonemoto, E. H.; Kim, Y. I.; Schmehl, R. H.; Wallin, J. O.;
Shoulders, B. A.; Richardson, B. R.; Haw, J. F.; Mallouk, T. E. J. Am.
Chem. Soc. 1994, 116, 10557. (b) Kim, Y. I.; Mallouk, T. E. J. Phys. Chem.
1992, 96, 2879. (c) Krueger, J. S.; Mayer, J. E.; Mallouk, T. E. J. Am.
Chem. Soc. 1988, 110, 8232. (d) Persaud, L.; Bard, A. J.; Campion, A.;
Fox, M. A.; Mallouk, T. E.; Webber, S. E.; White, J. M. J. Am. Chem. Soc.
1987, 109, 7309.
(4) (a) Borja, M.; Dutta, P. K. Nature 1993, 362, 43. (b) Dutta, P. K.;
Incavo, J. A. J. Phys. Chem. 1987, 91, 4443. (c) Dutta, P. K.; Turbeville,
W. J. Phys. Chem. 1992, 96, 9410. (d) Dutta, P. K.; Borja, M. J. Chem.
Soc., Chem. Commun. 1993, 1568. (e) Ledney, M.; Dutta, P. K. J. Am.
Chem. Soc. 1995, 117, 7687.
(5) Fukuzumi, S.; Urano, T.; Suenobu, T. Chem. Commun. 1996, 213.
(6) Corma, A.; Forne´s, V.; Garcia, H.; Miranda, M. A.; Primo, J.; Sabater,
M.-J. J. Am. Chem. Soc. 1994, 116, 2276.
(7) (a) Sankararaman, S.; Yoon, K. B.; Yabe, T.; Kochi, J. K. J. Am.
Chem. Soc. 1991, 113, 1419. (b) Yoon, K. B.; Hubig, S. M.; Kochi, J. K.
J. Phys. Chem. 1994, 98, 3865.
(8) (a) Rolison, D. R. Chem. ReV. 1990, 90, 867. (b) Rolison, D. R. Stud.
Surf. Sci. Catal. 1994, 85, 543.
(9) (a) Calzaferri, G.; Lanz, M.; Li, J. J. Chem. Soc., Chem. Commun.
1995, 1313. (b) Li, J.; Pfanner, K.; Calzaferri, G. J. Phys. Chem. 1995, 99,
2119. (c) Li, J.; Calzaferri, G. J. Electroanal. Chem. 1994, 377, 163. (d)
Li, J.; Calzaferri, G. J. Chem. Soc., Chem. Commun. 1993, 1430.
(10) (a) Walcarius, A.; Lamberts, L.; Derouane, E. G. Electrochim. Acta
1993, 38, 2257. (b) Walcarius, A.; Lamberts, L.; Derouane, E. G. Ibid.
1993, 38, 2267.
(11) Gemborys, H. A.; Shaw, B. R. J. Electroanal. Chem. Interfacial
Electrochem. 1986, 208, 95.
(12) Gra¨tzel, M.; Kalyanasundaram, K. Eds.; Kinetics and Catalysis in
Microheterogeneous Systems; Marcel Dekker: New York, 1991.
(13) Summers, R. A. The Bipyridinium Herbicides; Academic: New
York, 1980.
(14) (a) Kosower, E. M.; Cotter, L. J. J. Am. Chem. Soc. 1964, 86, 5524.
(b) Farrington, J. A.; Ebert, M.; Land, E. J. J. Chem. Soc., Faraday Trans.
1, 1978, 74, 665. (c) Farrington, J. A.; Ebert, M.; Land, E. J.; Fletcher, K.
Biochim. Biophys. Acta 1973, 314, 372.
S0002-7863(96)02645-5 CCC: $12.00 © 1996 American Chemical Society

Zeolite-Encapsulated MethylViologen
J. Am. Chem. Soc., Vol. 118, No. 50, 1996 12711
Chart 1. Perspective showing the snug fit of methylviologen
within the 13 Å supercage of zeolite-Y.
Chart 3. Increasing steric size of the counter cations C+
)
+
TMA+ < TEA+ < TBA+ < PPh4 < PPN+ used in this
study.
the vicinity of the zeolite-Y aperture.) Thus, by simple ligand
substitution, we can readily create two contrasting situations in
which one donor [Mn(CO)4P(OPh)3-] is size excluded from the
zeolite, while the other donor [Mn(CO)5-] has free access into
the supercage. Furthermore, one of the best advantages to be
gained in using metal carbonyl anions as reducing agents is that
the overall dimension of the salt is effectively controlled by
the size of the charge-compensating cations (C+), since the
anionic carbonylmanganates must enter the zeolitic pores in the
paired state with their counterions.18 Therefore, by varying the
size of the cations C+ listed in Chart 3, we can further finely
tune the effective size of the manganate donors.
-
Chart 2. Ready passage of Mn(CO)5 through the 7.4 Å
aperture of zeolite-Y which precludes the admission of the
phosphite-substituted Mn(CO)4P(OPh)3 shown to the right
-
on the same scale.
In this report, the factors that govern the interfacial electron
transfer between the methylviologen acceptor encapsulated in
zeolite-Y and the carbonylmanganate donors located externally
(in solution) will be examined with different substituents (L)
and various counter cations (C+).
∼14 000)16 upon one-electron reduction, so that methylviologen
cation radical (MV•+) is readily distinguished from its dicationic
state (MV2+). Thus, the progress of the redox reaction can be
easily monitored both visually (qualitatively) and spectro-
photometrically (quantitatively). Second, methylviologen can
be readily incorporated into various zeolites via simple aqueous
ion exchanges,2-4,8-11 and Chart 1 presents the perspective view
of MV2+ intercalated within the quasi-spherical (13 Å) super-
cage of zeolite-Y. Third, the incorporation of methylviologen
can be accurately controlled so that the occupancy in each
Results
For this investigation, we initially established that the simple
methylviologen salt was always subject to rapid and quantitative
(one-electron) reduction in solution, irrespective of the car-
bonylmanganate donor C+Mn(CO)4L-, with L ) CO and
P(OPh)3. Interfacial electron transfer to MV2+ extensively
intercalated into zeolite-Y particles was then examined with the
same series of carbonylmanganate salts comprised of different
sizes/shapes of the counter cation C+ as follows.
supercage is limited to an average of either one or two MV2+
,
hereafter designated as MV(1.0)Y and MV(2.0)Y, respec-
tively.18,19 Furthermore, methylviologen remains intact within
the dehydrated zeolites to allow the accommodation of various
other guest molecules into the remaining (void) space for
subsequent intermolecular interactions.2,18,19
I. Homogeneous Electron Transfer between Various
Carbonylmanganates and Methylviologen in Solution. When
the bis(triphenylphosphine)iminium (PPN+) salt of pentacar-
-
bonylmanganate Mn(CO)5 (1 equiv) was added to an aceto-
The anionic carbonylmanganates with A- ) Mn(CO)4L- [L
) CO or P(OPh)3] are chosen as the prototypical electron donors
based on the following factors. First, we will show that they
both behave as clean (one-electron) reducing agents toward
MV2+ both in a homogeneous solution of acetonitrile and as a
heterogeneous suspension in tetrahydrofuran. Second, while
Mn(CO)5- can freely pass through the 7.4 Å aperture of zeolite-
Y, the triphenyl phosphite-substituted analogue Mn(CO)4P(OPh)3-
is too big to enter the pore opening,20 and hence cannot react
directly with all the encapsulated MV2+. (See Chart 2 for the
computer-generated perspective views of these manganates in
nitrile solution of the trifluoromethanesulfonate (OTf-) salt of
methylviologen MV(OTf)2 (in the glovebox), the colorless
solution immediately turned deep blue. The FT-IR analysis of
the dark (inky) blue solution showed that dimanganese deca-
carbonyl Mn2(CO)10 was formed in quantitative amounts (Table
1, entry 1). The UV-vis spectrophotometric analysis of the
solution also revealed the formation of methylviologen cation
radical (MV•+) in the essentially quantitative amounts according
to the stoichiometry in eq 1.
MV2+ + Mn(CO)5- 98 MV•+ + 1/2Mn2(CO)10
(1)
(15) Bird, C. L.; Kuhn, A. T. Chem. Soc. ReV. 1981, 10, 49.
(16) Watanabe, T.; Honda, K. J. Phys. Chem. 1982, 86, 2617.
(17) Wolszczak, M.; Stradowski, C. Radiat. Phys. Chem. 1989, 33, 355.
(18) Yoon, K. B.; Kochi, J. K. J. Phys. Chem. 1991, 95, 1348.
(19) Yoon, K. B.; Kochi, J. K. J. Phys. Chem. 1991, 95, 3780.
(20) The kinetic diameter of Mn(CO)5- was estimated to be 7.1 Å based
on the crystal structure reported by: Corraine, M. S.; Lai, C. K.; Zhen, Y.;
Churchill, M. R.; Buttrey, L. A.; Ziller, J. W.; Atwood, J. D. Organo-
Similarly, the same deep blue solution was instantaneously
obtained when methylviologen was treated with 1 equiv of the
phosphite-substituted salt PPN+Mn(CO)4P(OPh)3-. The spec-
troscopic analyses of the blue solution revealed that the
corresponding manganese dimer Mn2(CO)8[P(OPh)3]2 and MV•+
were also formed in quantitative amounts (see Table 1). The
use of tetraethylammonium (TEA+) and sodium (Na+) salts of
-
metallics, 1992, 11, 35. The kinetic diameter of Mn(CO)4P(OPh)3 was
estimated to be J10 Å based on the computer minimization.

12712 J. Am. Chem. Soc., Vol. 118, No. 50, 1996
Yoon et al.
Table 1. Homogeneous Electron Transfer between
MV2+(CF3SO3-)2 and C+Mn(CO)4L- a
time Mn2(CO)8L2,b
(h)
solvent
C+
L
(conv)c
MV•+ d (conv)c
CH3CN PPN+ CO
0.2 0.083 (95)
0.079
0.085
0.080
0.081
(91)
(98)
(92)
(93)
CH3CN PPN+ P(OPh)3 0.2 0.085 (98)
CH3CN TEA+ P(OPh)3 0.2 0.088 (100)
CH3CN Na+
P(OPh)3 0.2 0.084 (97)
THFe
THFe
THFe
THFe
PPN+ CO
2
24
2
0.019 (22)
0.050 (58)
0.040 (46)
0.070 (80)
-
-
f
f
PPN+ CO
-
-
f
f
PPN+ P(OPh)3
-
-
f
f
PPN+ P(OPh)3 36
-
-
f
f
-
a Equimolar (0.087 mmol) amounts of MV2+(CF3SO3
)
and
2
C+Mn(CO)4L- were mixed in 50 mL of solvent. b In mmol (2×). c In
%, based on the amount (0.087 mmol) of each reactant. d In mmol,
based on ꢀ ) 13,900 at 605 nm for MV•+ in CH3CN. e MV2+(CF3SO3-
)
2
was insoluble in THF. f MV•+CF3SO3- was sparingly soluble in THF.
Figure 1. Comparison of the UV-vis spectrum of the methylviologen
cation radical (MV•+) encapsulated in zeolite-Y and dissolved in
acetonitrile (inset). The dashed line represents the background for the
undoped zeolite-Y.
-
Mn(CO)4P(OPh)3 also led to the same results, and indicated
that the nature of the counter cation did not affect the
homogeneous redox reaction. These results clearly showed that
electron transfer from the phosphite-substituted carbonyl-
manganates Mn(CO)4P(OPh)3- to MV2+ occurred in homoge-
neous (acetonitrile) solution according to the stoichiometry:
the supercage of zeolite-Y to form intermolecular charge-transfer
salts with MV2+
. In order to accommodate these rather
anomalous results, we carried out further quantitative studies
as follows.
MV2+ + Mn(CO)4P(OPh)3- 9
8
2. Quantitative Accounting of the Carbonylmanganate
Oxidations by MV2+ -Doped Zeolite-Y. A. Oxidation of the
Pentacarbonylmanganate Salts C+Mn(CO)5- with MV(1.0)Y.
MV•+ + 1/2Mn2(CO)8[P(OPh)3]2 (2)
-
When Na+ Mn(CO)5 (0.069 mmol, 0.12 equiv)23 was added
We thus conclude that the various carbonylmanganates can
effectively serve as clean one-electron reducing agents for
methylviologen.
to MV(1.0)Y24 (0.057 mmol of MV2+ in 0.1 g of zeolite)
suspended in THF,25 the colorless zeolite particles immediately
turned dark blue. The reaction mixture was stirred for 0.2 h,
and the supernatant solution was then quantitatively collected
by filtration. FT-IR analysis of the (supernatant) solution
revealed the presence of only Mn2(CO)10 in quantitative
amounts,23 i.e.
The redox transformations were also examined in tetrahy-
drofuran (THF) in which MV(OTf)2 was completely insoluble.
Nonetheless, the same redox reactions proceeded, albeit some-
what slowly21 as demonstrated in Table 1 by the carbonyl-
manganate conversions of ∼60 and 80% for L ) CO and
P(OPh)3 after 24 and 36 h, respectively.22 The latter indicated
that the redox reactions between MV2+ and Mn(CO)4L- as
described in eq 1 and 2 were also applicable to a THF medium,
despite the quasi-heterogeneous conditions.
(MV2+)Y + Na+Mn(CO)5 100%8
-
(Na+MV•+)Y + 1/2Mn2(CO)10 (3)
II. Heterogeneous Electron Transfer between the Car-
bonylmanganates and MV2+ Encapsulated in Zeolite-Y. 1.
Preliminary (Qualitative) Observations. When samples of
MV2+-doped zeolite-Y suspended in either THF or acetonitrile
were treated with small amounts of C+Mn(CO)5-, the zeolite
particles immediately turned varying degrees of blue, regardless
of the size of the cation C+ (see Chart 3). In striking contrast,
the supernatant solution either remained colorless or turned pale
yellow. Based on a qualitative (visual) inspection, the intensities
of the blue zeolites were significantly weaker (especially in the
dried state) when they were produced from the large TBA+,
PPh4+, and PPN+ salts of Mn(CO)5- compared to those obtained
from the corresponding salts with TEA+, TMA+, and Na+ as
the smaller counter cations (see Chart 3). Consistent with the
blue colorations, the diffuse reflectance UV-vis spectra of the
filtered and dried zeolites all confirmed the presence of MV•+
within the blue zeolites (compare Figure 1). Although these
tests were qualitative, the results were rather surprising to us in
view of our previous studies18 which showed that only the added
salts with cations smaller than TBA+ could successfully enter
Most importantly, the diffuse reflectance UV-vis spectrum of
the filtered dark blue zeolite showed the characteristic absorption
band of the methylviologen cation radical (MV•+) with the high
intensity shown in Figure 1. Unfortunately, the precise amount
of MV•+ produced could not be accurately measured owing to
the experimental difficulty of quantitatively extracting all of the
highly air-sensitive MV•+ out of zeolite. However, the high
intensity of MV•+ in the blue zeolite, together with the
quantitative oxidation of Mn(CO)5- to Mn2(CO)10, was indica-
tive of the complete conversion of MV2+ to MV•+. The
resulting spectrum also showed a weak broad absorption in the
near infrared (NIR) region with an absorption maximum at about
1100 nm.26
-
When the Mn(CO)5 salt with the large PPN+ cation was
added to the THF slurry of MV(1.0)Y, the zeolite particles also
(23) Based on the total amount of MV2+ intercalated into zeolite-Y.
(24) The number in the parentheses represents the (average) number of
MV2+ in the supercage.
(25) (a) For the quantitative study, we chose THF to mimic the
heterogeneous redox chemistry between the MV2+ ions that are intercalated
within zeolite-Y and the manganese carbonyl salts in solution. The ion
triplets of MV (i.e., MVX2 where X- ) Cl-, Br-, I-, PF6-, OTf-) are
insoluble in this solvent and eliminate the likelihood of MV2+ ions being
leached out of the zeolite by ion exchange with the added manganese salts.
(b) By a similar reasoning, we deliberately avoided the use of acetonitrile,
since this polar solvent is capable of leaching cations from the doped zeolite.
(26) The NIR band is also present in crystalline methylviologen cation
radical. See Bockman, T. M.; Kochi, J. K. J. Org. Chem. 1990, 55, 4130.
(21) The slow redox reaction may be attributed to the low solubility of
MV•+OTf-.
(22) During the course of reactions in THF, the supernatant solution
turned pale greenish blue due to the formation of yellow manganese dimers
and to the partial solubility of MV•+OTf-. The simultaneous formation of
MV•+OTf- was inferred from the isolation of a black precipitate which
afforded a dark blue solution of MV•+ upon dissolution in acetonitrile.

Zeolite-Encapsulated MethylViologen
J. Am. Chem. Soc., Vol. 118, No. 50, 1996 12713
(0.028 mmol, 1.0 equiv) in the manner previously observed with
-
Na+Mn(CO)5 in eq 3 (see entry 5, Table 2). The dark blue
-
zeolite obtained from the addition of 0.5 equiv of Na+ BPh4
-
to a mixture of MV(1.0)Y and PPN+Mn(CO)5 was visually
indistinguishable from that obtained by the addition of 1.0 equiv
of Na+ BPh4-. However, quantitative comparison of the diffuse
reflectance UV-vis spectra in Figure 2 showed a significant
difference in the intensities of the visible absorption band at
λmax ) 605 nm.27
B. Oxidation of the Phosphite-Substituted Carbonyl-
manganate Salt C+Mn(CO)4P(OPh)3- with MV(1.0)Y. The
sodium salt of the bulky carbonylmanganate substituted with
triphenyl phosphite [i.e., Na+Mn(CO)4P(OPh)3-] was also
oxidized by MV(1.0)Y, despite the large ligand which prevented
its passage through the 7.4-Å window of zeolite-Y (see Chart
2). As presented in Table 3, a stoichiometric amount of
Mn2(CO)8[P(OPh)3]2 was detected in the supernatant solution
immediately after the reaction of MV(1.0)Y with this carbon-
ylmanaganate, i.e.
Figure 2. Diffuse reflectance UV-vis-NIR spectra of the blue
zeolites-Y obtained from the treatment of a THF slurry of MV(1.0)Y
-
with PPN+Mn(CO)5 in the presence of (A) 1.0 equiv, (B) 0.5 equiv,
(MV2+)Y + Na+Mn(CO)4P(OPh)3 100%8
-
and (C) no added Na+BPh4
.
-
(Na+MV•+)Y + 1/2Mn2(CO)8[P(OPh)3]2 (5)
turned blue immediately. However, FT-IR analysis of the
supernatant solution revealed the presence of only a negligible
amount of Mn2(CO)10, despite the fact that zeolite particles
turned blue with significant intensity (especially when suspended
in the solvent). Since the amount of Mn2(CO)10 produced was
too small to be accurately measured, the reaction was repeated
with a large excess of MV(1.0)Y (1.0 g containing 0.571 mmol
of MV2+) in order to generate more product. Indeed, the careful
spectroscopic analysis of the reaction mixture after 0.2 h
revealed the presence of only 0.003 mmol of Mn2(CO)10 in the
supernatant solution. This redox reaction did not proceed any
further, even after a prolonged period of stirring (24 h). Since
Mn2(CO)10 contains two carbonylmanganese moieties, the
isolated dimer corresponded to the consumption of 0.006 mmol
of Mn(CO)5-, which in return indicated only 1% reduction of
the doped MV2+ within the zeolite, i.e.
-
By contrast, the large PPN+ salt of Mn(CO)4P(OPh)3 was
inactive, leading to only 1% conversion of the manganese salt
into the dimer even after 24 h (cf. eq 3). Moreover, the results
-
in Table 3 show that the addition of Na+ BPh4 sensitively
-
controlled the further oxidation of Mn(CO)4P(OPh)3 by
MV(1.0)Y in measure with the amount of added sodium ions,
in a manner similar to that observed with PPN+Mn(CO)5- (Vide
supra). Thus, the results unambiguously demonstrated that
electron transfer occurred readily to the zeolite-encapsulated
MV2+ acceptor from the large, size-excluded electron donor
Mn(CO)4P(OPh)3-, as long as Na+ ion was available.
C. Oxidation of the Medium-Size Tetraethylammonium
-
Salt TEA+Mn(CO)4P(OPh)3 with MV(1.0)Y. Heretofore,
we showed that the redox reactions between MV(1.0)Y and
C+Mn(CO)4L- [L ) CO or P(OPh)3] occurred instantaneously
and fell into two categories, i.e., either 100% (C+ ) Na+) or
1% (C+ ) PPN+), simply depending on the size of the counter
cation. Since a one-electron reduction must be accompanied
by the simultaneous transport of charge-compensating cations,
we thought that the rate of electron transfer from
Mn(CO)4P(OPh)3- in solution to the MV2+ ions in the interior
of zeolite-Y may be governed by the size of the migrating
cations. To test this hypothesis, we chose the appropriate
medium-size cation C+ ) TEA+ to effect a slower reduction
of the zeolite-encapsulated MV2+ with Mn(CO)4P(OPh)3-. As
shown in Table 4, the oxidation of the tetraethylammonium salt
(MV2+)Y + PPN+Mn(CO)5 1%8
-
(PPN+MV•+)Y + 1/2Mn2(CO)10 (4)
Consistent with this low conversion of MV2+ to MV•+, the blue
color of the zeolite was much weaker, especially in the dried
-
state, compared to the one obtained with Na+Mn(CO)5 (the
color appeared rather intense when suspended in the solvent).
Indeed, as demonstrated in Figure 2 (curve C), the diffuse
reflectance UV-vis analysis showed a much weaker spectrum
of MV•+ than that of the dark blue colored zeolite obtained
with Na+Mn(CO)5-. The unreacted PPN+Mn(CO)5- remained
intact in the supernatant solution even after 24 h stirring, as
demonstrated by the excellent material balance established in
Table 2 (entry 3).
-
TEA+Mn(CO)4P(OPh)3 indeed proceeded at an intermediate
-
rate compared to that of the sodium salt Na+Mn(CO)4P(OPh)3
.
Thus, FT-IR analysis of the reaction mixture showed a gradual
growth of the carbonyl bands of the dimer at νco ) 2001 and
1979 (cm-1), at the expense of the carbonyl bands of the
monomer (νco ) 1962, 1867, 1842, and 1829 cm-1), as
illustrated in Figure 3. Stirring the mixture for extended periods
(24 h) ultimately led to about 80% oxidative conversion of
It is important to note that the further progress of the redox
-
reaction of MV(1.0)Y with PPN+Mn(CO)5 could be deliber-
ately (and precisely) controlled by the addition of an exchange-
able sodium ion source to the reaction mixture. For example,
when half an equiv of Na+ BPh4- 23 was added to the mixture
-
Mn(CO)4P(OPh)3 to its dimer. The subsequent addition of a
-
of PPN+Mn(CO)5 and MV(1.0)Y, the zeolite immediately
large excess (∼5 equiv) of a sodium salt (Na+ BPh4-) into the
incomplete reaction mixture led to an immediate consumption
of all the residual manganese anion (see the last entry of Table
turned deep blue. Spectroscopic analysis of the supernatant
solution revealed the presence of Mn2(CO)10 in amounts (0.014
mmol, 0.5 equiv) corresponding to the amount of added sodium
salt (entry 4 in Table 2). Further increases in the amount of
(27) (a) Note, however, that the magnitude of the reflectance at λmax
)
605 nm of the half-reduced MV(1.0)Y in curve B is higher than one-half
-
added Na+ BPh4 [e.g., 0.057 mmol (1 equiv)] led to the
that of curve A. (b) Control experiments showed that MV2+ was not reduced
quantitative conversion of Mn(CO)5- to the dimeric Mn2(CO)10,
by Na+BPh4 alone under reaction conditions.
-

12714 J. Am. Chem. Soc., Vol. 118, No. 50, 1996
Table 2. Interfacial Electron Transfer between MV(1.0)Y and C+Mn(CO)5 Dissolved in THF.a Effect of the Cation Size
Yoon et al.
-
b
c
- e
C+
time (h)
MVY (g)
(MV2+
)
NaBPh4
Mn2(CO)10d (conv)e
unreacted Mn(CO)5
total Mn mmol (%)
Na+
0.2
0.2
24
0.2
0.2
0.1
1.0
1.0
0.1
0.1
(0.057)
(0.571)
(0.571)
(0.057)
(0.057)
0.057 (100)
0.006 (1)
0.006 (1)
0.029 (51)
0.056 (98)
0.006
0.064
0.063
0.033
0.010
0.063 (91)
0.070 (101)
0.069 (100)
0.062 (89)
0.066 (96)
PPN+
PPN+
PPN+
PPN+
0.029
0.057
-
a MV(1.0)Y was suspended in 8 mL of THF and 0.069 mmol of C+Mn(CO)5 was added to the slurry. b In mmol. c Additive, in mmol. d In
mmol (2×). e In %, based on the amount of MV2+ in zeolite.
- a
Table 3. Interfacial Electron Transfer between MV(1.0)Y and the Phosphite-Substituted Carbonylmanganate Salts C+Mn(CO)4P(OPh)3
.
Effect of the Cation Size
b
c
C+
time (h)
MVY (g)
(MV2+
)
NaBPh4
Mn2(CO)8P2d (conv)e
unreacted Mn(CO)4P- b,f
total Mn mmol (%)
Na+
0.2
0.2
24
0.2
0.2
0.1
1.0
1.0
0.1
0.1
(0.057)
(0.571)
(0.571)
(0.057)
(0.057)
0.057 (100)
0.005 (1)
0.005 (1)
0.029 (51)
0.057 (100)
0.000
0.052
0.051
0.028
0.000
0.057 (100)
0.057 (100)
0.056 (98)
0.057 (100)
0.057 100)
PPN+
PPN+
PPN+
PPN+
0.029
0.059
a MV(1.0)Y was suspended in 8 mL of THF and 0.057 mmol of C+Mn(CO)4P(OPh)3- was added to the slurry. b In mmol. c Additive, in mmol.
d In mmol (2×), P represents P(OPh)3. e In %, based on the amount of MV2+ in zeolite. f P ) P(OPh)3.
Table 4. Rate Profiles for the Reduction of MV(1.0)Y by the
Table 5. Progressive Course of the Interfacial Electron Transfer
- a
- a
Medium-Sized Donor Et4N+Mn(CO)4P(OPh)3
between MV(2.0)Y and C+Mn(CO)4P(OPh)3
. Effect of the Cation
Size and the MV2+ Occupancy
time
(h) NaBPh4 Mn2(CO)8P2 (conv)d
unreacted total mass
b
c
Mne
Mne
balf
time
(h)
unreacted total mass
b
C+
Mn2(CO)8P2
(conv)c
Mnd
Mnd
bale
0.2
0.5
1
3
5
0.010
0.016
0.029
0.036
0.038
0.041
0.045
0.057
18
28
51
63
67
72
79
100
0.047
0.043
0.027
0.020
0.019
0.016
0.012
0.000
0.057
0.059 102
98
Na+
0.2
0.2
1
3
5
0.112
0.004
0.012
0.014
0.014
0.014
0.016
0.001
0.001
(100)
(4)
0.000
0.105
0.099
0.095
0.095
0.095
0.094
0.110
0.111
0.112 100
Et4N+
Et4N+
Et4N+
Et4N+
Et4N+
0.109
0.111
0.109
0.109
0.109
0.110
0.111
97
99
97
97
97
98
99
0.056
0.056
0.057
0.057
0.057
0.057
97
97
98
98
98
98
(11)
(13)
(13)
(13)
(14)
(1)
7
7
24
Et4N+ 24
0.2g
0.292
PPN+
0.2
a 100 mg of MV(1.0)Y (MV2+ ) 0.057 mmol) was suspended in 8
PPN+ 24
(1)
0.112 100
-
mL of THF, and 35.1 mg (0.058 mmol) of Et4N+Mn(CO)4P(OPh)3
was added to the slurry. b Additive, in mmol. c In mmol (2×), P
a 100 mg of MV(2.0)Y (MV2+ ) 0.112 mmol) was suspended in 8
represents P(OPh)3. d Based on the amount of MV2+ in zeolite. e In
-
mL THF and 68.1 mg (0.112 mmol) of Et4N+Mn(CO)4P(OPh)3 was
-
.
f Material balance of the
added to the slurry. b In mmol (2×), P represents P(OPh)3. c In %, based
mmol; Mn represents Mn(CO)4P(OPh)3
-
Mn(CO)4P(OPh)3 moiety in %. g 0.2 h after the addition of 5 equiv
on the amount of MV2+ in zeolite. d In mmol, Mn represents
-
of Na+BPh4 to the reaction mixture.
-
Mn(CO)4P(OPh)3
%.
.
e Material balance of Mn(CO)4P(OPh)3 moiety, in
hour, but only a small additional conversion (30%) was achieved
in the following 23 h period. Figure 4A illustrates the
distinctively differentiated reaction (rate) profiles obtained when
cations with the progressively increasing size were used, namely,
C+ ) Na+ . TEA+ . PPN+. Thus, this figure clearly
demonstrated that the size of the counter cation sensitively
-
governed the overall redox rate between the Mn(CO)4P(OPh)3
in solution and the MV2+ encapsulated in zeolite-Y.
-
D. Oxidation of C+Mn(CO)4P(OPh)3 with the Highly
Doped MV(2.0)Y. The rate of ion migration from the exterior
surface into the internal pores of zeolites was also expected to
be affected by the number (occupancy) and the size of the
preexisting host molecules. Accordingly, we prepared a zeo-
lite-Y sample which contained an average of two MV2+ cations
per supercage [designated as MV(2.0)Y] in order to test the
effect of the occupancy of the MV2+ in the supercage on the
degree and the rate of oxidation of Mn(CO)4P(OPh)3-. When
TEA+ was employed as the counter cation, the oxidation of
the manganese salt with MV(2.0)Y only proceeded to less than
15% conversion even after 24 h stirring (Table 5). By contrast,
with Na+ as the counter cation, the production of a stoichio-
metric (quantitative) amount of manganese dimer was observed
within less than 10 min. [Despite the high probability of two
MV•+ existing within each supercage, the diffuse reflectance
spectrum of the blue zeolite was quite similar to the spectrum
shown in Figure 2A.26] As expected, the oxidative conversion
-
Figure 3. Progressive conversion of TEA+Mn(CO)4P(OPh)3 into
Mn2(CO)8[P(OPh)3]2 followed by FT-IR spectroscopy during the redox
reaction with MV(1.0)Y in THF with time: t ) 0 (a), 0.2 (b), 0.5 (c),
1 (d), 7 (e), 24 h (f). Spectrum (g) was measured at 0.2 h after the
-
addition of Na+BPh4 into the solution of (f).
4). The close examination of the reaction profile further
revealed that the redox reaction proceeded in two distinct
phases: an initial relatively fast transformation followed by a
slower oxidation. For example, roughly 50% of the anion was
converted into the dimeric Mn2(CO)4[P(OPh)3]2 during the first
-
of PPN+Mn(CO)4P(OPh)3 by MV(2.0)Y was limited to 1%.

Zeolite-Encapsulated MethylViologen
J. Am. Chem. Soc., Vol. 118, No. 50, 1996 12715
Table 7. Diagnostic IR bands of the Metal Carbonylsa
compound
IR stretching band (cm-1
)
-
Na+Mn(CO)5
2013(vw), 1897, 1872(sh), 1863, 1832(w)
2013(vw), 1898, 1866, ∼1851(sh)
2012(vw), 1898, 1865, ∼1851(sh)
2012(vvw), 1869, 1863
1894, 1862
-
TMA+Mn(CO)5
-
TEA+Mn(CO)5
TBA+Mn(CO)5
-
-
PPN+Mn(CO)5
PPh4+Mn(CO)5
1894, 1862
-
-
Na+Mn(CO)4P(OPh)3
1964, 1877, 1850, 1804
1963, 1867(sh), 1842, 1829
1962, 1871, 1840
-
TEA+Mn(CO)4P(OPh)3
-
PPN+Mn(CO)4P(OPh)3
Mn2(CO)8[P(OPh)3]2
2001(vw), 1979
a In THF solution.
Figure 4. Rate profiles for the oxidative conversion of Mn(CO)4P(OPh)3-
to Mn2(CO)8[P(OPh)3]2, showing the effect of the counter cation C+
(as indicated) and the zeolite occupancy with (A) MV(1.0)Y and (B)
MV(2.0)Y in THF.
N-methyl-2-quinolinium (Ered ) -1.08), and 4-carbomethoxy-
N-pyridinium (Ered ) -0.79 V) in acetonitrile solution.29a In
this regard, the electron transfer (Table 1, eq 1) between the
carbonylmanganate and MV2+ ion is expected to be enhanced
owing to the much stronger acceptor strength of the bipyridinium
ion (Ered ) -0.45 vs SCE in CH3CN).15 Previous studies with
pyridinium acceptors showed that the redox reaction in eq 1
derives from the stepwise mechanism,29b i.e.
Table 6. Kinetics of Interfacial Electron Transfer between MVY
- a
and Na+Mn(CO)4P(OPh)3
.
Effect of the MV2+ Loading
time
conv unreacted total mass
MV( )Y MV2+ b Mnc (h) Mn2 (%)e
Mn
Mnf balg
d
MV(0.1)Y 0.006 0.057 0.2 0.0056 100
0.050
0.045
0.027
0.024
0.007
0.004
0.000
0.027
0.018
0.005
0.000
0.056 98
0.055 96
0.051 89
0.054 95
0.053 93
0.055 96
0.057 100
0.111 99
0.114 102
0.111 99
0.112 100
MV(0.2)Y 0.012 0.057 0.2 0.010
MV(0.5)Y 0.028 0.057 0.2 0.024
MV(1.0)Y 0.057 0.057 0.03 0.030
MV(1.0)Y 0.057 0.057 0.1 0.046
MV(1.0)Y 0.057 0.057 0.2 0.051
83
86
53
81
89
kET
-
•
•
MV2+ + Mn(CO)5
8 MV•+ + Mn(CO)5
(6)
(7)
2Mn(CO)5 fast8 Mn2(CO)10
MV(1.0)Y 0.057 0.057 1.0 0.057 100
MV(2.0)Y 0.112 0.112 0.03 0.084
MV(2.0)Y 0.112 0.112 0.1 0.096
MV(2.0)Y 0.112 0.112 0.2 0.106
75
86
95
where kET denotes the rate-limiting electron transfer step. The
instantaneous redox reaction in eq 1 indicates that electron
MV(2.0)Y 0.112 0.112 1.0 0.112 100
-
transfer between MV2+ and Mn(CO)5 is rapid, since the
a 100 mg of MVY was suspended in 8 mL of THF. b MV2+ loading
in mmol. c Added Mn(CO)4P(OPh)3- in mmol. d Mn2(CO)8[P(OPh)3]2
formed in mmol (2×). e Conversion (%) based on the amount of MV2+
subsequent dimerization step of the manganese carbonyl radical
is known to be diffusion controlled.30 [The quantitative
formation of dimeric Mn2(CO)10 is consistent with the substitu-
tion inertness of the manganese carbonyl radical under the
reaction conditions.] Coupled with the high stability of MV•+,
let us now consider the merits of using carbonylmanaganate as
the electron donor paired with C+ (as described in the Introduc-
tion) for the study of electron transfer across the zeolite surface.
Our previous work established that only the donors that can
fit through the zeolite pores can effectively interact with the
preloaded acceptors to form electron donor-acceptor complexes
within the zeolite supercage.2,18,19,31 Likewise, for this electron
transfer study, only those carbonylmanganate reducing agents
that can be admitted into the zeolite supercage are expected to
react with the zeolite-encapsulated methylviologen. Thus, the
immediate as well as quantitative redox reaction that occurs
between the size-excluded Na+Mn(CO)4P(OPh)3- salt and the
zeolite-intercalated MV2+ ions (Table 3) appears to be highly
anomalous at first glance, since such a large anion as
-
in zeolite. f Of Mn(CO)4P(OPh)3 added in mmol. g Material balance
-
of Mn(CO)4P(OPh)3 in %.
The three different rate profiles obtained with MV(2.0)Y and
the carbonylmanganate salts of Na+, TEA+, and PPN+ are
compared in Figure 4B, and those observed with MV(1.0)Y
are shown in Figure 4A. The results show that steric congestion
within the zeolite-Y supercage further affected the degree and
the rate of the redox reaction, the differentiation of which was
especially apparent with the medium-sized TEA+Mn-
-
(CO)4P(OPh)3 in Figure 4, A and B.
In order to evaluate the factors controlling the rate of charge
propagation with C+ ) Na+, kinetic measurements were also
made on a series of MV2+ doped zeolite-Y with different loading
levels of one MV2+ per 2, 5, and 10 supercages.28a The
comparative results in Table 6 show very little difference in
the conversion with variations in the loading of MV2+ (compare
entries 1, 2, 3, 6, and 9). Furthermore, the rate of MV2+
reduction in MV(1.0)Y was too fast for us to experimentally
differentiate it from that in MV(2.0)Y, as shown by a direct
comparison of the results in Table 6, entries 4-7 with those in
entries 8-11.28b
-
Mn(CO)4P(OPh)3 is subject to zeolite shape selectivity.32
Furthermore, the occurrence of the fast redox reaction between
Na+Mn(CO)4P(OPh)3- and the highly congested MV(2.0)Y with
two large MV2+ ions (again instantaneously and quantitatively),
strongly suggests that some non-shape selectiVe pathway must
(29) (a) Bockman, T. M.; Kochi, J. K. J. Am. Chem. Soc. 1989, 111,
4669. (b) See: Lehman, R. E.; Bockman, T. M.; Kochi, J. K. J. Am. Chem.
Soc. 1990, 112, 458. Kochi, J. K; Bockman, T. M. AdV. Organometal. Chem.
1991, 33, 51.
(30) (a) Waltz, W. L.; Hackelberg, O.; Dofman, L. M.; Wojcicki, A. J.
Am. Chem. Soc. 1978, 100, 7259. (b) Walker, H. W.; Herrick, R. S.; Olsen,
R. J.; Brown, T. L. Inorg. Chem. 1984, 23, 3748.
(31) (a) Yoon, K. B.; Kochi, J. K. J. Am. Chem. Soc. 1989, 111, 1128.
(b) Yoon, K. B.; Huh, T. J.; Corbin, D. R.; Kochi, J. K. J. Phys. Chem.
1993, 97, 6492. (c) Yoon, K. B.; Huh, T. J.; Kochi, J. K. J. Phys. Chem.
1995, 99, 7042.
(32) Csicsery, S. M. in Zeolite Chemistry and Catalysis; Rabo, J. A.
Ed.; ACS Monograph 171; American Chemical Society: Washington, DC,
1976; pp 680 ff.
Discussion
-
Pentacarbonymanganate Mn(CO)5 is known to be a strong
-
one-electron donor.28 For example, Mn(CO)5 forms highly
colored intermolecular 1:1 charge-transfer ion pairs (λCT > 560
nm) even with a series of relatively weak pyridinium acceptors
such as 4-phenyl-N-methylpyridinium (Ered ) -1.27 V vs SCE),
(28) (a) We thank a referee for suggesting these experiments. (b) At short
reaction times, the measured conversions in Table 6 were subject to
considerable uncertainty ((10%) owing to the experimental difficulties in
efficient filtration when t < 10 min.

12716 J. Am. Chem. Soc., Vol. 118, No. 50, 1996
Yoon et al.
be operating. Among the various (possible) explanations, we
favor a mechanism in which electrons are initially transferred
from the external, size-excluded carbonylmanaganate donor to
a zeolite-encapsulated methylviologen acceptor,33 and then they
migrate through the zeolite framework. In other words, we
believe electron conduction is responsible for the non-shape
selective behavior. Alternatively, one might suspect that the
anomalous behavior could arise from the prior substitution of
the bulky P(OPh)3 ligand by the smaller THF solvent molecule
to generate Mn(CO)4THF- which could then be admitted into
the supercage of zeolite-Y. However, the instantaneous and
quantitative formation of the dimeric Mn2(CO)8[P(OPh)3]2 as
the sole product eliminates this possibility. Furthermore, the
results in Table 3 show that the unreacted PPN+Mn-
Chart 4. Size exclusion of the large cation PPN+ from the
supercage of zeolite-Y during electron-transfer reduction of
methylviologen.
-
(CO)4P(OPh)3 remains structurally intact even when it is
exposed to THF for prolonged periods (entry 3, as well as the
last entry in Table 5), which demonstrates that this manganese
anion is stable and substitution inert under the reaction condi-
tions.
so, this represents a unique example in which electron transfer
occurs solely in the zeolite supercages that lie on the crystal
surface. This formulation also predicts that charge separation
between the size-excluded cation and the zeolite-intercalated
reduced species may be achieved at the monolayer level.
Although the unsubstituted manganese carbonyl anion
-
Mn(CO)5 is able to freely enter the zeolite-Y supercage to
effect electron transfer with MV2+ (see Chart 2), the redox
Consistent with this picture, we propose that the migration
of the cation (C+) from the exterior into the internal supercages
is also responsible for the observed slow progress of the reaction
-
reaction between Na+Mn(CO)5 with MV(1.0)Y rapidly pro-
duces the dimeric Mn2(CO)10 in quantitative amounts only in
the supernatant solution (entry 1, Table 2). This important
observation also indicates that electron conduction through the
zeolite occurs after the carbonylmanganate donor is externally
oxidized by MV2+ at the surface. Such a mechanistic view is
especially attractive for the redox reaction between Na+Mn(CO)5-
and the highly loaded MV(2.0)Y, whose supercages are unlikely
to accommodate an additional (carbonylmanganate) guest since
they are already congested with two methylviologen moieties.
-
of the medium-sized TEA+ salt of Mn(CO)4P(OPh)3 with
MV(1.0)Y. The effect is even more apparent with the highly
loaded MV(2.0)Y, as illustrated in Figure 4. This conclusion
favors intrazeolite transport of the relatively small Na+ ions to
occur quite rapidlyseven through the highly congested MV(2.0)Y.
Furthermore, the incomplete conversions observed with
-
TEA+Mn(CO)4P(OPh)3 [i.e., 79% over MV(1.0)Y and 14%
over MV(2.0)Y, even after stirring for 24 h] suggest that the
remaining free space of the acceptor-doped supercage also
controls the progress of the electron transport. Likewise, the
rather slow reaction observed between the insoluble crystalline
salt MV(OTf)2 and PPN+Mn(CO)4L- [Table 1, L ) CO or
P(OPh)3] in THF suspension can be attributed to the absence
of an appropriate channel (i.e., void space) within the crystalline
solid for the simultaneous transport of the counter cation during
the redox reaction.
In order to ensure charge balance, however, electron conduc-
tion requires the transport of positive charge from the external
surface into the internal supercages. Unlike electrons, however,
the transport of cations from the zeolite exterior into the internal
framework is subject to zeolite shape selectivity,32 owing to their
much larger, definitive shapes. Accordingly, we attribute the
limited conversion (<1%) consistently observed with the PPN+
salts of carbonylmanganates (see Tables 2, 3, and 5) to the
inability of this cation to be transported into the supercage,
owing to its large size. Such a size restriction of PPN+ (relative
to Na+) during electron transfer to MV2+ is pictorially repre-
sented in Chart 4. In fact, the limited conversion of
PPN+Mn(CO)4L- to an extent of 1% is quite revealing, if we
recall that the number of outermost (peripheral) supercages in
zeolite-Y corresponds to about 1% of the total number of
supercages.36 As such, we attribute the very low (1%) conver-
sion of methylviologen with PPN+Mn(CO)4L- to the reduction
of only those MV2+ ions that exist within the peripheral
supercages of the zeolite microcrystals. Note that the large size-
excluded PPN+ cations are likely to remain in the vicinity of
the external surface in order to fulfill the charge balance. If
According to our formulation, both electrons and cations must
be propagated from the exterior surface into the zeolite interior
in order to effect complete electron transfer between the
dissolved manganese donors and all the intercalated MV2+
acceptors.37 Although cations must move from one supercage
to another during the redox reactions, our experimental results
do not provide a direct mechanism as to how the electrons
actually migrate from the exterior surface into the interior
supercages. An attractive possibility is for electron transfer to
occur from a methylviologen cation radical (MV•+) in one
supercage to a dicationic methylviologen (MV2+) in one of the
neighboring supercages by a direct contact through the inter-
connecting windows,38 i.e.
(33) The driving force for electron transfer is ∆G ∼5 kcal mol-1 based
on the oxidation potential of Mn(CO)5 (-0.25 V vs SCE)34 and the
(37) The related electron transport from the electrode to the zeolite-
encapsulated substrates accompanied by cation transport into the zeolite
interior has been proposed from studies of zeolite-modified electrodes.8-10
[In this regard, the carbonylmanganate donors Mn(CO)4L- (L ) CO and
P(OPh)3)] can be likened to a cathode.] However, the electrochemical
methods suffer from various problems8 associated with the fabrication of
zeolite modified electrodes and, in particular, the use of highly concentrated
electrolytic solutions which commonly leach out the electroactively charged
species from the zeolite.
-
reduction potential of MV2+ (-0.45 V)35 in acetonitrile.
(34) Tilset, M.; Parker, V. D. J. Am. Chem. Soc. 1989, 111, 6711.
(35) The oxidation potential of MV2+ has been shown to vary depending
on the polarity of medium. Thus the E°red values of MV2+ are -0.45 in
CH3CN, -0.43 in DMF, and -0.68 in water (vs SCE). However, from an
estimation of the zeolite polarity to be similar to the mixture of water and
acetonitrile (see: Yoon, K. B.; Kochi, J. K. J. Am. Chem. Soc. 1988, 110,
6586), the oxidation potential of MV2+ can be approximated to be -0.5 V.
(36) The total number of supercages (Nt) can be estimated by assuming
that each supercage occupies a 13 Å cube. The ratio of the number of
outermost supercages (No) to Nt depends on the zeolite-Y particle size, and
No/Nt is estimated to be 0.011 for the average crystal size (0.7 micron) of
the zeolite-Y samples used in this study.
(38) (a) The electron transfer through the window between the molecules
encapsulated in the neighboring supercages has also been proposed in other
systems. See refs 4b and 6. (b) The results in Table 6 show that Na+
migration through zeolite-Y was too fast for us to distinguish the kinetic
order for the reduction of MV2+
.

Zeolite-Encapsulated MethylViologen
J. Am. Chem. Soc., Vol. 118, No. 50, 1996 12717
(C+, MV•+)sc + (MV2+)sc
/
2
Chart 5. Electron-conduction mechanism for the propagation
of MV•+ through zeolite supercages, showing the
accompanying transport of the counter cation C+.
1
(MV2+)sc + (C+, MV•+)sc , etc. (8)
1
2
where the subscript sc1 represents the outermost supercage, and
sc2 the neighboring supercage, etc. The subsequent propagation
of the electron through the neighboring zeolite supercages based
on this scheme is schematically depicted in Chart 5. A less
attractive possibility involves the sequential migration of MV•+
out of sc1 into the neighboring supercage (sc2) that is ac-
companied by the reverse migration of MV2+ into sc1, etc.
However, the simultaneous movement of the relatively large
methylviologen mono- and dications through the narrow (7.4
Å aperture) windows of zeolite-Y depicted in Chart 1 is expected
to be prohibitively slow.
Summary and Conclusion
Interfacial electron transfer to a zeolite-intercalated cationic
acceptor proceeds equally well, even from size-excluded donors,
but it is adversely affected by the large sizes/shapes of charge-
balancing cations. Thus, the one-electron reduction of methyl-
viologen encapsulated in zeolite-Y particles (containing either
1 or 2 MV2+ per supercage) by the sodium carbonylmanganates
Na+Mn(CO)4L- with L ) CO or P(OPh)3 occurs rapidly and
quantitatively, i.e.
into the highly concentrated electrolytic solutions,40 our studies
establish electron transfer/cation transport as described in Chart
5 to be the highly efficient process for electron conduction.
Moreover, a similar consideration may be applicable in the use
of zeolites to achieve long-lived charge separation during
photoinduced electron transfer,41 and in the reduction of
transition-metal complexes exchanged into zeolites for hetero-
geneous catalysis.42
(MV2+)Y + Mn(CO)4L-
8 (MV•+)Y + Mn(CO)4L•
THF
despite the fact that the triphenyl phosphite-substituted donor
is too large to penetrate the zeolite aperture as depicted in Chart
2. In both cases, the reduced methylviologen cation radical is
observed as very dark blue zeolite particles, and the oxidized
(17-electron) carbonylmanaganese radical is quantitatively
isolated from the supernatant solution as the dimeric Mn2(CO)10
or Mn2(CO)8[P(OPh)3]2. In strong contrast, the same carbonyl-
manganates as the size-excluded PPN+ salts are ineffective
donors, and they can reduce only a small fraction of the
methylviologen, i.e., (1%) corresponding to the MV2+ located
in the outermost supercages of the zeolite particles.
The surprising efficiency of the size-excluded donor
Mn(CO)4P(OPh)3-, coupled with the absence of reduced
Mn(CO)4L• species within the zeolite particles, indicates that
reduction proceeds via an interfacial electron transfer from
Mn(CO)4L- in solution to only that MV2+ in the peripheral
supercages. As such, the complete reduction of every interca-
lated (MV2+) acceptor requires a mechanism for electron
conduction into the interior of the zeolite particle. Indeed, the
shape-selective behavior of the charge-compensating cations C+
(compare eq 3 with eq 4) accounts for their simultaneous
transport, as the reduced (MV•+) center migrates throughout
the zeolite framework in Chart 5. Such a cation-modulated
electron conduction must be rapid to account for the instanta-
neous and quantitative reduction of every methylviologen
acceptor in MV(1.0)Y according to eqs 3 and 5, as well as the
doubly occupied MV(2.0)Y in entry 1, Table 5 [provided, of
course, that C+ is sufficiently small (like Na+) to readily
penetrate the MV2+-doped zeolite-Y]. In a related vein, Shaw
and co-workers39 previously reported that electrolyte cations also
show shape-selective behavior in electron conduction through
zeolite-modified electrodes in cyclic voltammetric studies.
Although their mechanistic delineation was complicated by the
competitive leaching of the charged species from the zeolite
Experimental Section
Materials. Zeolite-Y (LZY-52, Lot No. 968087061020-S) from
Union Carbide was washed with aqueous 1 M NaCl solution and
subsequently treated with distilled deionized water until the silver ion
test for chloride was negative. The zeolite was then dried in air to
allow equilibration with atmospheric moisture. The average size of
the zeolite-Y particle was 0.7 µ as determined from the scanning
electron micrograph (SEM). Methylviologen dichloride (MVCl2) from
Hannong Chemical was recrystallized repeatedly from methanol until
colorless. The triflate salt of MV2+ was prepared from 4,4′-bipyridyl
with methyl trifluoromethanesulfonate in dichloromethane.
MV(OTf)2: 1H NMR (CD3CN), δ 8.84 (d, JHH ) 4 Hz, 4 H), 8.36 (d,
JHH ) 4.8 Hz, 4 H), 4.39 (s, 6 H). All air-sensitive metal carbonyl
compounds were handled and stored in a glovebox charged with high
purity argon or in Schlenk tubes under an argon atmosphere. The
characteristic FT-IR carbonyl stretching bands of the carbonylmanganate
compounds are listed in Table 7. Mn2(CO)10 from Strem was
recrystallized from toluene by cooling the saturated solution to -35
°C, and then stored in a glovebox. The sodium salt of Mn(CO)5- was
prepared from Mn2(CO)10 by treating with sodium amalgam (Na/Hg)
(40) For a review, see: Bard, A. J.; Mallouk, T. Techniques of Chemistry,
Molecular Design of Electrode Surfaces; Murray, R. W., Ed.; Wiley: New
York, 1992; Vol. XXII, Chapter 6.
(41) (a) Ramamurthy, V., Ed.; Photochemistry in Organized and
Constrained Media; VCH: New York, 1991. (b) Holt, S. L., Ed. Inorganic
Reactions in Organized Media; ACS Symposium Series 177; American
Chemical Society: Washington, DC, 1981.
(42) (a) Minachev, K. M.; Isakov, Y. I. In Zeolite Chemistry and
Catalysis: Rabo, J. A,, Ed.; ACS Monograph 171; American Chemical
Society: Washington, DC, 1976; p 552. (b) Chen, N. Y.; Garwood, W. E.;
Dwyer, F. G. Shape SelectiVe Catalysis in Industrial Applications; Marcel
Dekker: New York, 1989. (c) Rabo, J. A.; Angell, C. L.; Kasai, P. H.;
Schomaker, V. Discuss. Faraday Soc. 1966, 41, 328. (d) Park, Y. S.; Lee.
Y. S.; Yoon, K. B. J. Am. Chem. Soc. 1993, 115, 12220. (e) Kasai, P. H.;
Bishop, R. J., Jr. J. Phys. Chem. 1973, 77, 2308. (f) Westphal, U.; Geismar,
G. Z. Anorg. Allg. Chem. 1984, 508, 165. (g) Anderson, P. A.; Edwards, P.
P. J. Am. Chem. Soc. 1992, 114, 10608. (h) Haug, K.; Srdanov, V.; Stucky,
G. Metiu, H. J. Chem. Phys. 1992, 96, 3495. (i) Sun, T.; Seff, K.; Heo, N.
H.; Petranovskii, V. P. Science, 1993, 259, 495. (j) See also ref 1a.
(39) Shaw, B. R.; Creasy, K. E.; Lanczycki, C. J.; Sargeant, J. A.;
Tirhado, M. J. Electrochem. Soc. 1988, 135, 869.

12718 J. Am. Chem. Soc., Vol. 118, No. 50, 1996
Yoon et al.
-
in THF. The complete conversion of the dimer to Mn(CO)5 was
spectrally monitored by FT-IR analysis. The yellow solution was
separated from the mercury residue with the help of a disposable pipet
and filtered through a glass frit. The product was isolated by adding
excess n-hexane into the THF solution. The yellow powder was
recrystallized at -35 °C by the gradual addition of n-hexane into the
THF solution (until cloudy). The sodium salt was metathesized with
1 equiv of each chloride salt of tetramethylammonium (TMA+,
Matheson, Coleman and Bell), tetraethylammonium (TEA+, Aldrich),
tetra-n-butylammonium (TBA+, Aldrich), bis(triphenylphosphine)-
iminium (PPN+, Aldrich), and tetraphenylphosphonium (PPh4+, Ald-
rich), respectively, in THF by stirring for 2-3 h. NaCl was removed
by filtration and n-hexane was added to each clear yellow solution to
isolate yellow powders. The salts were similarly purified by recrys-
tallization. Mn2(CO)8[P(OPh)3]2 was prepared by irradiating a n-hexane
solution (200 mL) of Mn2(CO)10 (7 g, 0.018 mol) and triphenyl
phosphite P(OPh)3, (16.7 g 0.054 mol, Aldrich) in a Rayonet photo-
chemical reactor (using 350-nm tubes) under argon.43 The reaction
mixture was contained in a Pyrex Schlenk flask. Orange crystals slowly
formed over the period of 2 days, and the filtered orange crystals were
washed with n-hexane. This procedure not only resulted in a very pure
product, but it also afforded higher yields than the reported thermal
MVC12 (3.12 mmol) in 500 mL of water. Quantitative analysis showed
that almost all the MV2+ ions were exchanged into NaY (1.34 MV2+
per supercage) from the first exchange. The second exchange was
carried out by shaking the MV2+-doped zeolite and 0.55 g of MVC12
(2.26 mmol) in 500 mL of water. The degree of acceptor incorporation
reached slightly more than 2.0 MV2+ per supercage after the second
exchange. The MV2+-doped zeolites were first dried in air at ambient
temperatures. The air-dried zeolites were then evacuated at 25 °C for
2 h under vacuum (<10-5 torr). The temperature of MV(1.0)Y was
slowly increased to 150 °C over a period of 2 h, and they were
evacuated further at this temperature for an additional 10 h. The
maximum dehydration temperature for MV(2.0)Y was 100 °C since
the acceptor-doped zeolite turned pale blue at higher temperatures. All
the dried zeolites were transferred into an argon-filled glovebox and
kept in the glass-stoppered containers.
Oxidation of Carbonylmanganates with MV2+-Doped Zeolites-
Y. Typically, THF (1 mL) was added to a glass vial (15 mL capacity)
containing a weighed amount (0.1 or 1.0 g) of an acceptor-doped zeolite
in order to soak the powders. Generation of heat was usually noticed
when the relatively polar solvent was added to the zeolites. A required
amount of the carbonylmanganate was weighed into a separate vial
and 5 mL of THF was added. The metal carbonyl solution was then
introduced into the vial containing zeolite with the aid of a pipet. The
vial used for the preparation of metal carbonyl was washed twice with
an additional 1 mL of the solvent. Each of the washed solvent was
also transferred into the reaction vial. A small Teflon-coated magnetic
stirring bar was added into the reaction vessel and then the slurry was
stirred with the aid of a magnetic stirrer for the desired length of time.
-
methods.44 The sodium salt of Mn(CO)4P(OPh)3 was prepared by
treating Mn2(CO)8[P(OPh)3]2 with sodium amalgam (Na/Hg) in THF.
The salt was isolated by adding excess n-hexane into the filtered
solution, and purified by adapting the method applied for Na+Mn(CO)5-
.
-
The TEA+ and PPN+ salts of Mn(CO)4P(OPh)3 were prepared from
-
Na+Mn(CO)4P(OPh)3 by metathesizing sodium ion with the corre-
sponding cation using TEA+Cl- and PPN+Cl-, respectively. The TEA+
The reaction mixture was filtered by passing it through a cotton-
packed glass column. The cotton-packed column was previously
washed with solvent and dried. The packed cotton did not affect the
carbonylmanganates used in this study. The filtered solution was
collected into a 25- or 50-mL volumetric flask. The reaction vial was
washed repeatedly with small amounts of fresh solvent and each washed
solvent was passed through the (filtered) zeolite cake. Washing the
filtered zeolites was continued until the collected solution accumulated
to approximately 20 mL for dilute reaction mixtures (25-mL volumetric
flask) or about 30 mL for concentrated reaction mixtures (50-mL
volumetric flask). After filling the volumetric flask to the marked point
with an additional amount of fresh solvent, the collected solution was
quantitatively analyzed by FT-IR spectrophotometry. The products
were identified by spectral comparisons with the authentic samples.
They were quantified ((5%) by comparison with the calibration curves
independently constructed from IR absorbencies of the characteristic
band vs concentration. For the qualitative visual experiment, a large
excess (about 10 molar excess with respect to the amount of acceptor)
of manganese carbonyl compounds was added into each zeolite slurry
consisting of approximately 0.1 g in 5 mL of solvent.
-
and PPN+ salts of Mn(CO)4P(OPh)3 were similarly purified by
recrystallizing the corresponding salt in a mixed solvent of THF and
n-hexane at -35 °C. TMA+Cl-, TEA+Cl-, and TBA+Cl- were dried
-
at 100 °C for 15 h in vaccuo. Na+ BPh4 from Aldrich was used as
received. All the purified solvents were stored in Schlenk flasks under
argon and kept in the glovebox. Tetrahydrofuran (THF) and diethyl
ether were purified by distillation over sodiobenzophenone. Acetonitrile
was stirred over potassium premanganate for 15 h at room temperature
and then separated by filtration. The permanganate treated acetonitrile
was distilled, and the distillate was then refluxed and redistilled over
phosphorus pentoxide under an argon atmosphere. n-Hexane was
stirred repeatedly with sulfuric acid until the acid layer remained
colorless. The separated solvent was then washed with water and
subsequently with dilute aqueous sodium bicarbonate solution. The
washed n-hexane was dried with anhydrous calcium chloride and
distilled from sodium under an argon atmosphere.
Preparation of MV2+-Doped Zeolites-Y. The MV2+-doped zeo-
lites-Y were prepared by the aqueous ion exchange of sodium zeolite-Y
with MV2+ ions. The amounts of MV2+ ions intercalated in zeolite-Y
were controlled to be 0.1, 0.2, 0.5, 1.0, and 2.0 per supercage.
Typically, MV(1.0)Y (zeolite-Y exchanged with 1 MV2+ per supercage
on an average basis) was obtained by adding 1.13 g (4.65 mmol) of
MVC12 into a slurry containing 10 g of NaY (4.64 mmol of supercage)
and 0.9 L of distilled deionized water in a 1-L flask. The slurry was
Instrumentation. FT-IR spectra were obtained with a NaCl liquid
cell (0.1 mm) on a Nicolet 10 DX FT spectrometer with 2 cm-1
resolution. 1H NMR spectra were recorded on a JEOL FX 90Q FT
spectrometer operating at 89.55 MHz. The diffuse reflectance spectra
of the colored zeolites were measured on a Shimadzu UV-3101PC
spectrometer equipped with an integrating sphere. The scanning
electron micrograph of zeolite-Y was obtained on a JEOL JSM-35C
microscope.
then shaken for 15 h at ambient temperature. The resultant MV2+
-
doped zeolite-Y was filtered through a sintered glass filter and washed
with copious amounts of water until the chloride test with AgNO3 was
negative. The amounts of unexchanged MV2+ ion in the combined
wash were quantified by UV-vis spectrophotometry monitored at λmax
) 257 nm (ꢀ ) 20 417).13 Quantitative analysis revealed that essentially
all the added MV2+ ion was exchanged into the zeolite. MV(2.0)Y
was prepared by a two-step exchange. The first exchange was carried
out by shaking 5 g of NaY (2.32 mmol of supercage) and 0.76 g of
Acknowledgment. K.B.Y. thanks the R. A. Welch Founda-
tion for financial support during the sabbatical leave in Houston.
This research was also carried out under the auspices of the
NSF-KOSEF sponsored United States/Korea Cooperative Re-
search Program and the Aimed Basic Research Program of
KOSEF. We thank T. M. Bockman for many helpful discussions.
(43) Osborne, A. G.; Stiddard, M. H. B. J. Chem. Soc. 1964, 643.
(44) (a) Hieber, W.; Freyer, W. Chem. Ber. 1959, 92, 1765. (b) Lewis,
J.; Manning, A. R.; Miller, J. R. J. Chem. Soc. A 1966, 845. (c) Drew, D.;
Darensbourg, D. J.; Darensbourg, M. Y. Inorg. Chem., 1975, 14, 1579.
JA962645Y