Dinitrogen Reduction, Sulfur Reduction, and Isoprene Polymerization via Photochemical Activation of Trivalent Bis(cyclopentadienyl) Rare-Earth-Metal Allyl Complexes

Article abstract of DOI:10.1021/acs.organomet.5b00613

Dinitrogen can be reduced by photochemical activation of the trivalent rare-earth-metal bis(pentamethylcyclopentadienyl) allyl complexes (C₅Me₅)₂Ln(η³-C₃H₄R) (Ln = Y, Lu; R = H, Me) to form the (N=N)^2- complexes [(C₅Me₅)₂Ln]₂(μ-η²:η²-N₂). This demonstrates that productive organolanthanide photochemistry is not limited to complexes of the unusual (η³-C₅Me₄H)^- ligand in the heteroleptic complexes (C₅Me₅)₂(C₅Me₄H)Ln and (C₅Me₅)(C₅Me₄H)₂Ln. Photolytic activation of (C₅Me₅)₂Ln(η³-C₃H₅) (Ln = Y, Lu) in the presence of isoprene provides a rare photopolymerization route to polyisoprene. Sulfur can also be reduced by photolysis of (C₅Me₅)₂Ln(η³-C₃H₅) (Ln = Y, Lu) to generate the (S)^2- complexes, [(C₅Me₅)₂Ln]₂(μ-S), which have variable Ln-S-Ln angles depending on crystallization conditions.

Full text of DOI:10.1021/acs.organomet.5b00613

Article
pubs.acs.org/Organometallics
Dinitrogen Reduction, Sulfur Reduction, and Isoprene
Polymerization via Photochemical Activation of Trivalent
Bis(cyclopentadienyl) Rare-Earth-Metal Allyl Complexes
Megan E. Fieser, Casey W. Johnson, Jefferson E. Bates, Joseph W. Ziller, Filipp Furche,*
and William J. Evans*
Department of Chemistry, University of California, Irvine, California 92697-2025, United States
*
S Supporting Information
ABSTRACT: Dinitrogen can be reduced by photochemical
activation of the trivalent rare-earth-metal bis-
3
(
pentamethylcyclopentadienyl) allyl complexes (C Me ) Ln(η -
5
5 2
2−
C H R) (Ln = Y, Lu; R = H, Me) to form the (NN)
3
4
2
2
complexes [(C Me ) Ln] (μ-η :η -N ). This demonstrates that
5
5 2
2
2
productive organolanthanide photochemistry is not limited to
3
−
complexes of the unusual (η -C Me H) ligand in the
5
4
heteroleptic complexes (C Me ) (C Me H)Ln and (C Me )-
5
5 2
5
4
5
5
3
(
(
C Me H) Ln. Photolytic activation of (C Me ) Ln(η -C H )
5 4 2 5 5 2 3 5
Ln = Y, Lu) in the presence of isoprene provides a rare
3
photopolymerization route to polyisoprene. Sulfur can also be reduced by photolysis of (C Me ) Ln(η -C H ) (Ln = Y, Lu) to
generate the (S) complexes, [(C Me ) Ln] (μ-S), which have variable Ln−S−Ln angles depending on crystallization
conditions.
5
5 2
3
5
2−
5
5 2
2
INTRODUCTION
The fact that photochemical activation with lanthanides
■
could lead to the potent reduction of dinitrogen was surprising,
Recently, the traditional view that rare-earth-metal complexes
have minimal photochemistry was overturned by the
since Ln3+ ions are not effective at absorbing light.
2−6
This is
2
−
because the Laporte-forbidden nature of 4f → 4f transitions is
not relaxed by vibronic coupling, due to the contracted nature
of the 4f orbitals. Although lanthanide complexes can display
observation that dinitrogen can be reduced to (NN) by
3
+
photolysis of the Ln mixed-ligand tris(cyclopentadienyl) rare-
earth-metal complexes (C Me ) (C Me H)Ln (1-Ln; Ln = Y,
5
5 2
5
4
3
+
1
outstanding emission properties, particularly with Eu and
Lu, Dy) and (C Me )(C Me H) Lu. The products of the two-
5
5
5
4
2
3+
2
2
Tb in the red and green parts, respectively, of the visible
spectrum, this emission requires sensitizers to absorb the
energy. Laporte-allowed ligand to metal charge transfer
electron reductions of N , namely [(C Me ) Ln] (μ-η :η -N )
2
5
5 2
2
2
2
2
(
2-Ln) and [(C Me )(C Me H)Ln] (μ-η :η -N ), respectively,
5 5 5 4 2 2
−
are formed in reactions in which (C Me H) is oxidized to
C Me H) (Scheme 1).
5 4 2
5
4
1
(
LMCT) and ligand to ligand charge transfer (LLCT)
(
4
transitions are known for the lanthanides, but photoactivation
does not generally lead to productive lanthanide-based
Scheme 1. Reduction of Dinitrogen with
2
−9
transformations.
Although the 5f valence orbitals have a
(
(
C Me )(C Me H) Lu (Reaction a) and
5 5 5 4 2
better radial extension than the 4f orbitals, there are still
relatively few examples of photochemically activated reactions
C Me ) (C Me H)Ln (1-Ln; Ln = Y, Lu, Dy) (Reaction b)
5
5 2
5
4
under Photolytic Conditions and Formal Half-Reactions
1
0−13
3+
with actinide complexes.
The other rare-earth ions, Sc
3+
and Y , also have limited metal-based photochemistry since
they are d species. However, recently Chirik et al. have shown
0
the dinitrogen can be reduced by photolysis of (C Me H) Zr-
5
4
2
14
(
aryl) complexes.
2
Density functional theory (DFT) analysis of the unexpected
rare-earth-metal photochemistry in Scheme 1 indicated that
3
−
absorptions involving the unusual (η -C Me H) ligand were
5
4
1
responsible. Specifically, two absorptions at 412 and 437 nm
that are present in the heteroleptic (C Me ) (C Me H)Ln and
5
5 2
5
4
Received: July 16, 2015
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.organomet.5b00613
Organometallics XXXX, XXX, XXX−XXX

Organometallics
C Me )(C Me H) Ln complexes, but not in the homoleptic
Article
(
(C
5
Me
5
)
2
YCl
2
K(THF) (1.937 g, 3.577 mmol) in toluene (125 mL).
After the mixture was stirred overnight, the solvent was removed under
vacuum and a pale yellow tacky solid resulted. Dropwise addition of
5
5
5
4
2
analogues, (C Me ) Ln or (C Me H) Ln, appear to be
5
5
3
5
4
3
involved. Due to the rare nature of this photochemical
activation, it was desirable to test the generality of this
photoactivity beyond the two classes of mixed-ligand tris-
1
,4-dioxane (5 mL) to a stirred slurry of this solid in 125 mL of
hexanes gave an immediate color change to bright yellow. After the
mixture was stirred for 12 h, the yellow slurry was centrifuged and
filtered. The solvent was removed under reduced pressure to yield 4-Y
as a yellow microcrystalline solid (1.05 g, 70%). Yellow crystals
suitable for X-ray crystallography were grown from a saturated hexanes
(
cyclopentadienyl) complexes and three metals in Scheme 1.
Extending the reactions of Y, Dy, and Lu to the larger metals in
the lanthanide series is problematic, because the synthesis of
pure (C Me ) (C Me H)Ln and (C Me )(C Me H) Ln com-
1
5
5 2
5
4
5
5
5
4
2
solution at −35 °C. H NMR (C D ): δ 2.19 (d, 4H, CH C(CH )-
6
6
2
3
1
3
plexes for the larger lanthanides (La, Ce, Pr, Nd, Sm) is
CH ), 1.96 (s, 33H, (C Me ) , CH C(CH )CH ). C NMR (C D ):
2 5 5 2 2 3 2 6 6
15
complicated by ligand rearrangement. Hence, it was of
particular interest to determine if this photoactivity depended
δ 169.8, 117.1, 64.6, 29.4, 11.6. IR: 3054 m, 2966 s, 2908 s, 2859 s,
725 s, 2391 vw, 2287 vw, 1509 m, 1437 w, 1379 w, 1316 m, 1247 m,
1161 w, 1025 s, 951 w, 849 m, 767 s, 677 m cm . Anal. Calcd for
C H Y: C, 69.55; H, 9.00. Found: C, 69.49; H, 9.24.
2
−
1
3
−
on the unusual (η -C Me H) ligand found in each of these
5
4
24
37
complexes.
Given the η nature of the photoactive (η -C Me H) ligand,
Photolytic Reaction Conditions. All photolysis experiments were
3
3
−
5
4
conducted in a hood with aluminum foil covered windows with a
Hanovia medium-pressure 450 W mercury vapor lamp (PC451050/
610741). The 5.5 in. long lamp was clamped to hang inside a 13 in. ×
3
the structurally related allyl complexes (C Me ) Ln(η -
5
5 2
1
6−18
C H )
(3) were candidates for possible photoactivation.
3
5
Reactions analogous to Scheme 1 with allyl complexes would
provide a more facile synthetic route to the reduced dinitrogen
1
.5 in. diameter cavity of a double-walled quartz water cooling jacket.
Tap water at 24 °C flowed through the jacket at a rate of 6.4 L/min,
and the temperature in the hood was kept between 25 and 27 °C.
Samples were placed adjacent to the outer wall of the cooling jacket in
the middle of the hood.
Photolysis of (C Me ) Lu(η -C H ) (3-Lu) To Form
[(C Me ) Lu] (μ-η :η -N ) (2-Lu). In a nitrogen-filled glovebox, 3-
Lu (12 mg, 0.025 mmol) was dissolved in C D6 (1 mL) and
transferred to a J. Young NMR tube equipped with a greaseless
stopcock. After irradiation for 24 h, the yellow solution became deep
2
2
complexes [(C Me ) Ln] (μ-η :η -N ) (2), since the allyl
5
5 2
2
2
complexes are precursors to the heteroleptic
19,20
(
C Me ) (C Me H) Ln complexes
as well as the loosely
5
5 3−x
5
4
x
3
5
5
2
3
5
ligated tetraphenylborate cationic metallocene complexes
2
2
1
8,21−23
5 5 2 2 2
[
(C Me ) Ln][(μ-Ph) BPh ],
which are also used as
24
5
5 2
2
2
2
−
6
precursors to the (NN) complexes.
We report here that the allyl complexes are photoactive and
this reactivity can be used not only to reduce dinitrogen but
also to polymerize isoprene. Photopolymerization of isoprene is
1
24
red. H NMR analysis revealed full conversion to 2-Lu.
3
Photolysis of (C Me ) Y(η -C H ) (3-Y) To Form
(C Me ) Y] (μ-η :η -N ) (2-Y). As described for 3-Lu, a yellow
5
5
2
3
5
25−28
2
2
[
rare and generally involves radical initiators.
To show the
5 5 2 2 2
solution of 3-Y (10 mg, 0.02 mmol) became red-orange after
generality of the photoactivity of these common metallocene
2
4
irradiation for 24 h and generated a 1.5:1 mixture of 3-Y and 2-Y.
rare-earth-metal allyl complexes, we also report that S can be
8
3
reduced by Ln3+ allyl complexes. To our knowledge, reduction
Photolysis of (C
5
Me
5
)
2
Y(η -C
3 4
H R) (4-Y) To Form
(C Me ) Y] (μ-η :η -N ) (2-Y). As described above for 3-Lu,
2
2
[
5
5 2
2
2
of S by a trivalent rare-earth-metal complex has only been
8
photolysis of a yellow solution of 4-Y (10 mg, 0.02 mmol) for 14 h
2
2
29
1
24
reported before with {[(Me Si) N] (THF)Y} (μ-η :η -N )
and by alkyl thiolate disproportionation in cyclopentadienyl
and guanidinate rare-earth-metal alkyl systems. Density
functional theory is used to explain the photochemical activity
of this allyl metallocene system.
3
2
2
2
2
gave a 1:2 mixture of 4-Y and 2-Y by H NMR spectroscopy.
30
Large-Scale Yttrium Photolyses. Large-scale reactions were run
in 250 mL round-bottom vessels containing either 3-Y (500 mg, 1.25
mmol) or 4-Y (518 mg, 1.25 mmol) dissolved in toluene (50 mL).
After the stirred yellow solutions had been irradiated for 24 h, they
became red-orange. The solutions were centrifuged and filtered before
the solvent was removed under vacuum. The resulting red-orange
3
1
EXPERIMENTAL DETAILS
1
■
powders were shown by H NMR spectroscopy to contain the starting
material and the 2-Y product in ratios of 1.2:1 for 3-Y and 1:1.1 for 4-
The syntheses and manipulations described below were conducted
under argon or nitrogen with rigorous exclusion of air and water using
glovebox, vacuum line, and Schlenk techniques. Solvents were dried
over columns containing Q-5 and molecular sieves. NMR solvents
Y.
Gas Chromatography/Mass Spectrometry (GC/MS) Studies.
Photolyses were run in methylcyclohexane to analyze the gas present
in the headspace above the solution. GC/MS analysis of the gas
formed after irradiation of 3-Y (50 mg, 0.1 mmol) and 4-Y (61 mg,
(
Cambridge Isotope Laboratories) were dried over sodium potassium
32
alloy, degassed, and vacuum-transferred prior to use. KC Me ,
5
5
3
17
17
18
(
(
(
C Me ) Ln(η -C H ) (2-Ln; Ln = Y,
Lu,
Dy ), and
0.12 mmol) in methylcyclohexane (7 mL) under N revealed
2
5
5
2
3
5
1
7
C Me ) YCl K(THF) were prepared according to the literature.
significant molecular ion peaks at m/z 41.16 and 56.03, respectively.
These matched the data on pure gas samples of propene and
isobutylene. Analogous reactions run with ∼50 mg of 3-Y or 4-Y in 4
5
5
2
2
2-Methylallyl)magnesium chloride (0.5 M, in THF), 1,4-dioxane,
propene, and isobutylene were purchased from Sigma-Aldrich and
were used as received. Sulfur (Fisher) was purified by sublimation
before use. Isoprene (Acros) was dried over molecular sieves and
mL of toluene-d revealed peaks at m/z 42.14 and 57.18.
8
Isoprene Reactivity. In an argon-filled glovebox, ∼50 mg of 3-Ln
(Ln = Y, Lu) and a magnetic stir bar were added to a sealable 50 mL
side arm Schlenk flask equipped with a greaseless stopcock. The flask
was added to the high-vacuum line, and argon was removed under
vacuum. Isoprene (approximately 10 mL) was vacuum-transferred into
the flask. After irradiation for 12 h, gel had formed on the side of the
flask near the light source. The amount of gel continued to increase
over the course of 8 days of irradiation. Excess isoprene was removed
under vacuum to leave a thick gel that was analyzed by gel permeation
chromatography at the Goodyear Tire Co. to have low molecular
1
degassed by three freeze−pump−thaw cycles. H NMR spectra were
obtained on a Bruker DRX500 MHz spectrometer with a BBO probe
13
at 25 °C unless otherwise specified. C NMR spectra were obtained
on a Bruker DRX500 MHz spectrometer operating at 126 MHz with a
TCI cryoprobe at 25 °C. IR samples were prepared as KBr pellets and
the spectra were obtained on a Varian 1000 FT-IR system. Absorption
spectra were collected using a Cary 50 Scan UV−vis spectrometer at
2
5 °C in toluene or hexane. Mass spectra were recorded on a Thermo
Scientific Trace GC Ultra. Elemental analyses were performed on a
1
13
PerkinElmer 2400 Series II CHNS analyzer.
weight with M and M <6000. H NMR and C NMR spectra in
CDCl of the polymer that had been washed with MeOH and treated
n w
3
(
C Me ) Y(η -CH CMeCH ) (4-Y). In a nitrogen-filled glovebox, a
5
5 2
2
2
3
solution of (2-methylallyl)magnesium chloride 0.5 M in THF (7.20
mL, 3.60 mmol) was added dropwise to a stirred suspension of
with HCl contained resonances for cis-1,4-, trans-1,4-, and 3,4-
polyisoprene with the largest peaks for 3,4-polyisoprene.
B
DOI: 10.1021/acs.organomet.5b00613
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
3
Table 1. X-ray Data Collection Parameters for (C Me ) Y(η -CH CMeCH ) (4-Y), [(C Me ) Y] (μ-S) (5-Y(tol)), and Two
5
5 2
2
2
5
5 2
2
Forms of [(C Me ) Lu] (μ-S) (5-Lu(tol) and 5-Lu(hex))
5
5 2
2
4
-Y
5-Y(tol)
5-Lu(tol)
5-Lu(hex)
formula
fw
C H Y
C H SY2
C H Lu S
C H Lu S
2
4
37
40 60
40 60
2
40 60
2
414.44
88(2)
750.76
922.88
922.88
temp (K)
143(2)
88(2)
88(2)
wavelength (Å)
cryst syst
space group
a (Å)
0.71073
triclinic
0.71073
tetragonal
0.71073
0.71073
triclinic
tetragonal
P42 c
P1
̅
P4
̅
̅
P1
̅
1
1
11.0137(17)
14.194(2)
15.681(2)
2195.6(6)
67.4577(18)
75.8753(19)
84.5088(19)
4
14.7891(7)
14.7891(7)
18.9371(9)
4141.9(4)
90
14.6701(7)
14.6701(7)
18.9233(9)
4072.5(4)
90
10.5471(5)
11.3849(6)
16.6188(9)
1894.36(17)
77.2811(7)
76.8625(7)
85.1302(7)
2
b (Å)
c (Å)
3
V (Å )
α (deg)
β (deg)
90
90
γ (deg)
90
90
Z
4
4
ρcalcd (Mg/m3)
μ (mm−1)
R1 (I > 2σ(I))
wR2 (all data)
1.254
0.204
1.505
1.618
2.660
2.861
4.894
5.261
0.0312
0.0286
0.0633
0.0141
0.0334
0.0269
0.0808
0.0650
[
(C Me ) Y] (μ-S) (5-Y). In an argon-filled glovebox, 3-Y (204 mg,
5 5 2 2
RESULTS
■
0
.51 mmol) was dissolved in toluene (10 mL). S (8 mg, 0.03 mmol)
8
3
To determine if complexes beyond (C Me ) (η -C Me H)Ln
1-Ln) and (C Me )(C Me H) Ln could be photoactivated to
was added to the yellow solution, and the mixture was transferred to a
sealable 50 mL side arm Schlenk flask equipped with a greaseless
stopcock. During irradiation for 4 h, the stirred bright yellow solution
became pale yellow. In an argon-filled glovebox, the solvent was
removed under vacuum and the solids were washed with pentane to
leave [(C Me ) Y] (μ-S) (5-Y) as a pale yellow solid (84 mg, 52%).
Colorless single crystals of 5-Y suitable for X-ray diffraction were
grown from toluene at −35 °C. H NMR (C D ): δ 2.07 (s, 60H,
C Me ). C NMR (C D ): δ 119.16 (s, C Me ), 11.68 (s, C Me ). IR:
961 s, 2906 s, 2858 s, 2725w, 1638 w, 1490 w, 1437 m, 1378 m, 1259
w, 1186 w, 1162 w, 1086 m, 801 w cm . Anal. Calcd for C H SY :
C, 63.99; H, 8.06. Found: C, 64.51; H, 8.50. Eight attempts at
elemental analysis on this compound, including use of crystals
examined by X-ray diffraction, did not give better analytical data than
this (see the Supporting Information).
5
5 2
5
4
(
5
5
5
4
2
2
2
reduce N
reduction of (C
to [(C
Me
Me
Ln(η -C
)
Ln]
(μ-η :η -N
) (2-Ln), photo-
) (3-Ln; Ln = Y, Lu) was
2
5
5
2
2
2
3
)
H
3
5
5
2
5
explored. To aid in identifying organic byproducts of the allyl
5
5
2
2
photolysis, the methylallyl analogue of 3-Ln, namely
3
(
C Me ) Y(η -CH CMeCH ) (4-Y), was synthesized from
5
5 2
2
2
1
17
6
6
(methylallyl)magnesium chloride and (C Me ) YCl K(THF)
5 5 2 2
13
5
5
6
6
5
5
5
5
and was fully characterized. The X-ray crystal structure (Figure
) is similar to that of 3-Y, as detailed in the Supporting
Information.
2
1
−1
40
60
2
[
(C Me ) Lu] (μ-S) (5-Lu). As described for 5-Y above, 3-Lu (124
5 5 2 2
mg, 0.255 mmol) was combined with S in toluene (10 mL). During
8
irradiation for 5 h, the stirred bright yellow solution became pale
yellow. In an argon-filled glovebox, the solvent was removed under
vacuum to leave a pale yellow solid, [(C Me ) Lu] (μ-S) (5-Lu; 65
5
5
2
2
mg, 55%). Colorless single crystals of 5-Lu suitable for X-ray
diffraction were grown from toluene (5-Lu(tol)) and from hexane
1
(
5-Lu(hex)) at −35 °C. H NMR (C D ): δ 2.22 (s, 60H, C Me ).
6
6
5
5
1
3
C NMR (C D ): δ 120.96 (s, C Me ), 11.35 (C Me ). IR: 2953 s,
6
6
5
5
5
5
2
1
910 s, 2854 s, 2726 w, 1648 w, 1439 m, 1376 m, 1259 w, 1113 w,
061 m, 1020 w, 992 w. As found for 5-Y, obtaining satisfactory
elemental analyses was difficult. The best data of five attempts
(
Supporting Information) are given here. Anal. Calcd for C H SLu :
40 60 2
C, 52.06; H, 6.55. Found: C, 53.75; H, 6.32.
5
3
Figure 1. Thermal ellipsoid plot of (η -C Me ) Y[(η -CH CMeCH )
(
5
5
2
2
2
X-ray Crystallographic Data. Crystallographic information for
complexes 4-Y, 5-Y(tol), 5-Lu(tol), and 5-Lu(hex) is given in Table 1,
and details are given in the Supporting Information.
4-Y) drawn at the 50% probability level. The second independent
molecule of 4-Y in the unit cell and hydrogen atoms are omitted for
clarity.
Computational Details. DFT calculations were carried out on
3
3
(
(
C Me ) Y(η -C H ) (3-Lu) (C Me ) Lu(η -C H ) (3-Y), and
5 5 2 3 5 5 5 2 3 5
3
C Me ) Y(η -CH CMeCH ) (4-Y) using the hybrid meta-general-
5
5
2
2
2
3
3
34
UV−visible absorption spectra were taken of 3-Y, 3-Lu, and
4-Y (Figure 2) to determine how they compared with that of
photoactive 1-Ln. The spectra of the allyl complexes are similar
ized gradient approximation functional TPSSh. Small-core ECPs
35
and def-TZVP basis sets were used for Y and Lu. All calculations
3
6
were performed using the Turbomole quantum chemistry software.
Time-dependent DFT (TDDFT) calculations
1
3
7,38
to those of 1-Y and 1-Lu in that they contain two broad
were also per-
absorptions, with the lowest energy absorption being near the
405 nm emission of the medium-pressure mercury vapor lamp
used for irradiation.
formed to simulate the UV−vis spectrum for 3-Lu, 3-Y, and 4-Y. A full
description of the computational methods is reported in the
Supporting Information.
C
DOI: 10.1021/acs.organomet.5b00613
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
of 3-Y with neat isoprene without irradation showed no
polymerization.
Photoreduction of Sulfur. Irradiation of 3-Y and 3-Lu
under argon in toluene in the presence of elemental sulfur, S ,
8
led to the formation of the bridging (S)2 complexes
−
[
(C Me ) Ln] (μ-S) (5) (eq 2) (Figure 3). X-ray crystallog-
5 5 2 2
3
Figure 2. UV−vis spectra of (C Me ) Y(η -C H ) (3-Lu; dotted blue
5
5
2
3
5
Figure 3. Thermal ellipsoid plots of [(C Me ) Lu] (μ-S) (5-Lu)
3
3
5
5
2
2
line), (C Me ) Y(η -C H ) (3-Y; dashed red line), and (C Me ) Y(η -
5
5
2
3
5
5
5 2
crystallized from (a) hexane (5-Lu(hex)) and (b) toluene (5-Lu(tol))
drawn at the 50% probability level. Hydrogen atoms are omitted for
clarity.
CH C(Me)CH ) (4-Y; solid green line).
2
2
Photoactivated Reduction of Dinitrogen. Photolysis of
solutions of 3-Y, 3-Lu, and 4-Y in benzene under dinitrogen
2−
2
2
produced the (NN) complexes [(C Me ) Ln] (μ-η :η -
5
5 2
2
raphy revealed an unusual feature of the lutetium product: 5-Lu
crystallized in two different forms depending on the
crystallization solvent, hexane vs toluene, even though neither
solvent cocrystallized in either structure. As shown in Table 1,
when 5-Lu was crystallized from toluene, crystals of 5-Lu(tol)
1
N ) (Ln = Y (2-Y), Lu (2-Lu)), which were identified by H
2
2
4
NMR spectroscopy (eq 1). Only in the case of 2-Lu was the
̅
1
formed in the P42 c space group and were isomorphous with 5-
Y(tol), also crystallized from toluene. When 5-Lu was
crystallized from hexane, crystals of 5-Lu(hex) were isolated
̅
in space group P1.
reaction quantitative in 24 h. For 3-Y and 4-Y, conversions of
only 40−45% were observed in 24 h with the N reduction
2
1
product 2-Y being the only metallocene product by H NMR
spectroscopy. Using cyclooctane as an internal standard, it
appeared that 4-Y generated approximately 50% more 2-Y than
All three crystalline materials, 5-Y(tol), 5-Lu(tol), and 5-
Lu(hex), have bimetallic structures with similar bond lengths
after accounting for the difference in the eight-coordinate ionic
radii of Y (1.019 Å) and Lu (0.977 Å), as shown in Table 2.
Bond angles are also similar in the three complexes except for
the Ln−S−Ln angles. Complexes 5-Y(tol) and 5-Lu(tol) have
Ln−S−Ln angles of 171.91(6) and 173.27(5)°, respectively,
but the Ln−S−Ln angle in 5-Lu(hex) is 166.68(5)°. Since the
linearity of such angles is sometimes used to argue for multiple
3
-Y in the same time period.
Gas chromatography/mass spectrometry (GC/MS) analysis
42
of the gas present in the headspace above a large-scale
irradiation of 3-Y under N in methylcyclohexane gave a
2
molecular ion peak that matched the m/z value and retention
time of a standard sample of propene. The analogous
experiment in toluene-d gave a peak with an m/z value that
8
was 1 mass unit higher, as is appropriate for propene-d.
Analogous photolysis of 4-Y in methylcyclohexane gave a peak
at the m/z value of isobutene, and the corresponding reaction
Table 2. Selected Bond Lengths (Å) and Bond Angles (deg)
of 5-Y and 5-Lu (Crystallized from both Toluene and
Hexane)
in toluene-d gave a peak 1 mass unit higher. These results are
8
consistent with formation of propene and isobutene by
hydrogen abstraction from solvent by photolytically generated
39,40
5-Y(tol)
5-Lu(tol)
5-Lu(hex)
allyl and methylallyl radicals.
No evidence for dimers of
Ln(1)−Cnt1
Ln(1)−Cnt2
2.341
2.351
2.288
2.297
2.300
allyl and methylallyl radicals was observed. Hydrogen
abstraction from solvent would be expected to be prevalent,
since there is a much higher concentration of solvent in
comparison to the radical concentration.
2.300
Ln(2)−Cnt3
2.288
Ln(2)−Cnt4
2.297
Ln(1)−S(1)
Ln(2)−S(1)
Ln−S−Ln
Cnt1−Ln(1)−Cnt2
Cnt3−Ln(2)−Cnt4
S(1)−Ln(1)−Cnt1
S(1)−Ln(1)−Cnt2
S(1)−Ln(2)−Cnt3
S(1)−Ln(2)−Cnt4
2.5433(3)
2.5026(2)
2.4868(10)
2.4974(10)
166.68(5)
140.4
Photopolymerization of Isoprene. Photolysis of 3-Y and
3
-Lu in neat isoprene that had been vacuum-transferred into
1
171.91(6)
138.3
173.27(5)
138.2
the reaction vessel generated a viscous white product. H NMR
and 1 C NMR analysis in CDCl indicated the presence of
3
41
3
139.7
polyisoprene in the form of cis-1,4-, trans-1,4-, and 3,4-isomers,
with the 3,4-isomer being predominant. Gel permeation
chromatography (GPC) analysis of the molecular weight of
the material indicated that the polymer had a low molecular
weight with M and M values under 6000. The blank reaction
111.0
110.7
111.0
110.8
111.2
108.4
109.5
110.9
n
w
D
DOI: 10.1021/acs.organomet.5b00613
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
bonding and covalency, it is significant that for 5-Lu the value
of this angle can vary significantly depending on the solvent of
crystallization.
the HOMO of the allyl complexes, which is localized on the
2
allyl ligand, and the LUMO, which is mainly d in character.
z
The calculations suggest that this would generate an allyl radical
Density Functional Theory (DFT) Calculations. DFT
and the excited-state Y2+ species “(C Me ) Y”, which should be
5
5 2
3
calculations on (C Me ) Ln(η -C H ) (3-Ln; Ln = Lu, Y)
capable of reducing dinitrogen on the basis of data on the
5
5 2
3
5
recently discovered Y2+ complexes.
43,44
Identification of
indicate that the HOMO has electron density located on the
2
allyl ligand while the LUMO resembles a d orbital of the metal
propene and isobutene by GC/MS in the headspace over the
photolysis solution is consistent with formation of allyl radicals
z
center (Figure 4). Time-dependent DFT (TDDFT) calcu-
39,40
followed by hydrogen abstraction.
Reactions done in
toluene-d8 that generate propene-d and isobutene-d are
consistent with radical abstraction from solvent. This result is
parallel to the postulated pathway of photolysis of the
(
C Me ) (C Me H) Ln complexes, which have (C Me H)
5 5 3−x 5 4 x 5 4 2
1
byproducts from formation of C Me H radicals.
5
4
Irradiation of 3-Y and 3-Lu in neat isoprene led to the
formation of low-molecular-weight polyisoprene. NMR evi-
dence showed that a mixture of cis-1,4-, trans-1,4-, and 3,4-
polyisoprene had formed, with the 3,4-isomer being predom-
inant. This is consistent with radical polymerization of isoprene,
as opposed to rare-earth-metal-based coordination polymer-
Figure 4. Molecular orbital plots of (a) the HOMO and (b) the
LUMO for 3-Lu, using a contour value of 0.06.
45−51
ization, which can give specific isomers.
Although selective
polymerization was not observed in this case, the photo-
25−28
3
7,38
lations
were performed to compare with the UV−vis
polymerization of isoprene is rare.
Complexes 3-Y and 3-Lu also can be photoactivated to
reduce S to form the bridging sulfide complexes 5-Y and 5-Lu
spectra of 3-Ln. Excitations near 360 and 390 nm were
predicted for 3-Lu and excitations near 370 and 390 nm were
predicted for 3-Y. These excitation energies match the
experimental spectra, but the oscillator strength for the low-
energy excitation (389 nm) for 3-Lu is not as large as would be
expected in comparison to the experimental spectrum. Table S8
in the Supporting Information summarizes the excitation
energies and oscillator strengths. The lowest energy absorption
8
(eq 2). While the reduction is not as challenging as that of
dinitrogen (S to S2 is only −0.476 V vs SCE ), this reduction
suggests that this photochemical rare-earth-metal process is not
restricted to just the reduction of dinitrogen. To our
−
52
knowledge, this is just the third example of S being reduced
8
to (S)2 or (S
−
)
2
2−
by a trivalent rare-earth-metal complex.
29−31
for both 3-Y and 3-Lu is the HOMO−LUMO ligand to metal
Structural analysis of the sulfide-bridged complexes provided
another example of the variability of organometallic rare-earth-
metal structures on the basis of crystallization conditions.
3
2
charge transfer (LMCT) from the η -allyl ligand to a d orbital.
z
This is analogous to the first two excitations of 1-Y and 1-Lu3,
which were also predicted to be LMCT transitions from the η
[(C Me ) Lu] (μ-S) (5-Lu) was found to crystallize from
5 5 2 2
2
z
hexane with a Lu−S−Lu angle more linear, 173.27(5)°, than
the 166.68(5)° angle found in crystals obtained from toluene.
ligand to a d orbital that led to the observed photoreduction
1
chemistry. The next excitations for 3-Y and 3-Lu under 300
5
−
nm are all predicted to be LMCT from the (η -C Me ) ligands
Similar situations have recently been reported for (C
5
Me
5 2
) Y-
5
5
53
54
2
(NC Me ) and [(C Me ) DyH] , where different crystal-
to the d orbital, which is also analogous to the calculations of
-Ln.
Further DFT studies of the methylallyl complex 4-Y reveal a
4
4
5
5 2
2
z
1
lization methods led to different structures of the same
complex. Evidently, in some combinations of ligands and
metals, the energy surface is very shallow and several structures
are possible. This emphasizes the dangers of drawing
conclusions from a single structural feature in a single structure
of a crystal. For example, it could be argued that the more linear
1
HOMO and LUMO similar to those seen for both 3-Y and 3-
Lu (Figure S6 in the Supporting Information). TDDFT
predicts excitations near 360 and 390 nm similar to those
found for 3-Y. The oscillator strength of the excitation for 4-Y
near 390 nm is slightly larger than that for 3-Y, while the
oscillator strength of the excitation for 4-Y near 360 nm is
almost half that for 3-Y. This trend matches the observed
differences in the experimental UV−vis spectra.
1
71.91(6)° angle in 5-Y in comparison to the 166.68(5)° angle
in 5-Lu(hex) was due to enhanced covalency with yttrium,
since it is a transition metal. However, 5-Lu(tol) has a
173.27(5)° angle that is even closer to linear than that in the
yttrium structure.
DISCUSSION
■
3
CONCLUSION
The allyl complexes (C Me ) Y(η -C H ) (3-Y), (C Me ) Lu-
■
5
5 2
3
5
5
5 2
3
3
(
η -C H ) (3-Lu), and (C Me ) Y(η -CH C(Me)CH ) (4-Y)
These studies have shown that productive rare-earth-metal-
based photochemistry is not limited to complexes containing
3
5
5
5 2
2
2
are similar to the photochemically active compounds
3
3
−
(
C Me ) Ln(η -C Me H) (1-Ln) and (C Me )(C Me H) Ln
the unusual (η -C Me H) ligand or to heteroleptic tris-
5
5 2
5
4
5
5
5
4
2
5 4
in Scheme 1 in terms of their yellow color, their UV−vis
(cyclopentadienyl) species. In fact, photochemical reduction of
dinitrogen is possible with a common class of rare-earth-metal
metallocenes, the (C Me ) Ln(C H ) allyl complexes, which
5
spectra, and their overall structures, which include two (η -
−
3
C Me ) rings and one η ligand. As such, they were good
5
5
5
5 2
3
5
candidates to undergo the photochemistry observed in Scheme
. As shown in eq 1, each of the allyl complexes can be
are frequently used as precursors to a variety of bis-
(cyclopentadienyl) rare-earth-metal compounds. The results
1
2
−
photoactivated to reduce dinitrogen to (NN) .
not only show that powerful reductions such as N to (N
2
2−
DFT calculations suggest that the source of the photoactivity
is a LMCT band near 390 nm arising from a transition between
N) can be accomplished with this synthetically accessible
class of complexes, but also suggest that a wider range of rare-
E
DOI: 10.1021/acs.organomet.5b00613
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
earth-metal compounds may be photochemically active.
Photolytic methods should be more generally considered with
the rare-earth metals not only for reduction, but also for
photopolymerization, as demonstrated here with isoprene.
(12) Thomson, R. K.; Cantat, T.; Scott, B. L.; Morris, D. E.; Batista,
E. R.; Kiplinger, J. L. Nat. Chem. 2010, 2, 723−729.
(13) Gardner, B. M.; Patel, D.; Lewis, W.; Blake, A. J.; Liddle, S. T.
Angew. Chem., Int. Ed. 2011, 50, 10440−10443.
(14) Margulieux, G. W.; Semproni, S. P.; Chirik, P. J. Angew. Chem.,
Int. Ed. 2014, 53, 9189−9192.
ASSOCIATED CONTENT
Supporting Information
■
(15) Fieser, M. E. Ph.D. Thesis, University of California, Irvine, CA,
2015.
*
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.organo-
met.5b00613. These data can also be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif (CCDC file nos. 1411848−
(16) Evans, W. J.; Ulibarri, T. A.; Ziller, J. W. J. Am. Chem. Soc. 1990,
1
(
12, 2314−2324.
17) Evans, W. J.; Kozimor, S. A.; Brady, J. C.; Davis, B. L.; Nyce, G.
W.; Seibel, C. A.; Ziller, J. W.; Doedens, R. J. Organometallics 2005, 24,
269−2278.
18) Demir, S.; Zadrozny, J. M.; Nippe, M.; Long, J. R. J. Am. Chem.
2
(
1
411851, respectively).
Soc. 2012, 134, 18546−18549.
Computational details, converged calculated structures
for 3-Y, 3-Lu, and 4-Y, and selected bond distances for 4-
Y, 5-Y(tol), 5-Lu(tol), and 5-Lu(hex) (PDF)
Cartesian coordinates for calculated structures (XYZ)
Crystallographic data (CIF) for 4-Y, 5-Y(tol), 5-Lu(tol),
and 5-Lu(hex) (CIF)
(19) Demir, S.; Mueller, T. J.; Ziller, J. W.; Evans, W. J. Angew. Chem.,
Int. Ed. 2011, 50, 515−518.
(20) Mueller, T. J.; Fieser, M. E.; Ziller, J. W.; Evans, W. J. Chem. Sci.
2
011, 2, 1992−1996.
21) Evans, W. J.; Seibel, C. A.; Ziller, J. W. J. Am. Chem. Soc. 1998,
20, 6745−6752.
22) Evans, W. J.; Davis, B. L.; Champagne, T. M.; Ziller, J. W. Proc.
Natl. Acad. Sci. U. S. A. 2006, 103, 12678−12683.
23) Hamaed, H.; Lo, A. Y. H.; Lee, D. S.; Evans, W. J.; Schurko, R.
W. J. Am. Chem. Soc. 2006, 128, 12638−12639.
24) Schmiege, B. M.; Ziller, J. W.; Evans, W. J. Inorg. Chem. 2010,
9, 10506−10511.
25) Tang, D. K.; Ho, S. Y. J. Polym. Sci., Polym. Chem. Ed. 1984, 22,
1357−1363.
26) Salman, S. R.; Al-Jarrah, M. M. F. Polym.-Plast. Technol. Eng.
(
1
(
(
AUTHOR INFORMATION
Corresponding Authors
E-mail for F.F.: filipp.furche@uci.edu.
E-mail for W.J.E.: wevans@uci.edu.
■
*
*
(
4
(
Notes
(
The authors declare no competing financial interest.
1
(
(
989, 28, 1009−1014.
27) Derouet, D.; Thuc, C. N. H. J. Rubber Research 2008, 11, 78−96.
ACKNOWLEDGMENTS
■
28) Gobran, R. H.; Berenbaum, M. B.; Tobolsky, A. V. J. Polym. Sci.
For support of this research, we thank the U.S. National
Science Foundation for the experimental studies under grant
CHE-1265396 (W.J.E.) and the Department of Energy for the
theoretical studies under grant DE-SC0008694 (F.F.). We also
thank Dr. John Greaves and Dr. Beniam Berhane at the
University of California, Irvine Mass Spectrometry Facility, for
assistance with mass spectrometry, Professor A. S. Borovik for
assistance with UV−vis spectroscopy, Jordan F. Corbey for help
with crystallography, Dr. Margaret Flook for polymer analysis,
and Ryan R. Langeslay for experimental assistance.
1960, 46, 431−440.
(29) Corbey, J. F.; Fang, M.; Ziller, J. W.; Evans, W. J. Inorg. Chem.
2015, 54, 801−807.
(30) Li, Y.; Pi, C.; Zhang, J.; Zhou, X.; Chen, Z.; Weng, L.
Organometallics 2005, 24, 1982−1988.
31) Zhang, Z.; Zhang, L.; Li, Y.; Hong, L.; Chen, Z.; Zhou, X. Inorg.
Chem. 2010, 49, 5715−5722.
32) Evans, W. J.; Kozimor, S. A.; Ziller, J. W.; Kaltsoyannis, N. J. Am.
Chem. Soc. 2004, 126, 14533−14547.
33) Staroverov, V. N.; Scuseria, G. E.; Tao, J.; Perdew, J. P. J. Chem.
Phys. 2003, 119, 12129−12137.
34) Andrae, D.; Hauessermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
Theor. Chim. Acta 1990, 77, 123−141.
35) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7,
297−3305.
36) Furche, F.; Ahlrichs, R.; Hat
Weigend, F. WIREs Comput. Mol. Sci. 2014, 4, 91−100.
37) Bauernschmitt, R.; Ahlrichs, R. Chem. Phys. Lett. 1996, 256,
(
(
(
(
REFERENCES
■
(
3
(
(
1) Fieser, M. E.; Bates, J. E.; Ziller, J. W.; Furche, F.; Evans, W. J. J.
Am. Chem. Soc. 2013, 135, 3804−3807.
̈
tig, C.; Klopper, W.; Sierka, M.;
(
2) Carnall, W. T. In Handbook on the Physics and Chemistry of Rare
Earths; Bunzli, J.-C. G., Pecharsky, V. K., Eds.; Elsevier: Amsterdam,
979; Vol. 43, p 171.
3) Bu
̈
(
1
(
(
(
̈
nzli, J.-C. G. Acc. Chem. Res. 2006, 39, 53−61.
454−464.
4) Vogler, A.; Kunkely, H. Inorg. Chim. Acta 2006, 359, 4130−4138.
(38) Bates, J. E.; Furche, F. J. Chem. Phys. 2012, 137, 164105.
(39) Gordon, A. S.; Smith, S. R.; McNesby, J. R. J. Am. Chem. Soc.
1959, 81, 5059−5061.
5) Leonard, J. P.; Nolan, C. B.; Stomeo, F.; Gunnlaugsson, T. In
Photochemistry and Photophysics of Coordination Compounds II; Balzani,
V., Campagna, S., Eds.; Springer: Berlin, 2007; Vol. 281, p 1.
(40) Chen, Y.; Tschuikow-Roux, E. J. Phys. Chem. 1993, 97, 3742−
3749.
(41) Tanaka, Y.; Takeuchi, Y.; Kobayashi, M.; Tadokoro, H. J. Polym.
(
6) Lanthanide Luminescencence; Photophysical, Analytical and Bio-
logical Aspects, 1st ed.; Hanninen, P., Harma, H., Eds.; Springer: New
York, 2011; Vol. 1.
7) Zinnen, H. A.; Pluth, J. J.; Evans, W. J. J. Chem. Soc., Chem.
Commun. 1980, 810−812.
8) Ogawa, A.; Ohya, S.; Doi, M.; Sumino, Y.; Sonoda, N.; Hirao, T.
Tetrahedron Lett. 1998, 39, 6341−6342.
9) Sumino, Y.; Harato, N.; Tomisaka, Y.; Ogawa, A. Tetrahedron
003, 59, 10499−10508.
10) Kalina, D. G.; Marks, T. J.; Wachter, W. A. J. Am. Chem. Soc.
977, 99, 3877−3879.
11) Bruno, J. W.; Kalina, D. G.; Mintz, E. A.; Marks, T. J. J. Am.
Chem. Soc. 1982, 104, 1860−1869.
̈
̈
̈
Sci. Part A-2 1971, 9, 43−57.
(
(42) Shannon, R. D. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr.,
Theor. Gen. Crystallogr. 1976, 32, 751−767.
(43) Fang, M.; Lee, D. S.; Ziller, J. W.; Doedens, R. J.; Bates, J. E.;
Furche, F.; Evans, W. J. J. Am. Chem. Soc. 2011, 133, 3784−3787.
(44) MacDonald, M. R.; Ziller, J. W.; Evans, W. J. J. Am. Chem. Soc.
2011, 133, 15914−15917.
(
(
2
(
1
(45) Kaita, S.; Doi, Y.; Kaneko, K.; Horiuchi, A. C.; Wakatsuki, Y.
Macromolecules 2004, 37, 5860−5862.
(46) Zhang, L.; Luo, Y.; Hou, Z. J. Am. Chem. Soc. 2005, 127, 14562−
14563.
(
F
DOI: 10.1021/acs.organomet.5b00613
Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(
47) Rodrigues, I.; Mihalcea, I.; Volkringer, C.; Loiseau, T.; Visseaux,
M. Inorg. Chem. 2012, 51, 483−490.
(
(
48) Jian, Z.; Cui, D.; Hou, Z. Chem. - Eur. J. 2012, 18, 2674−2684.
49) Li, L.; Wu, C.; Liu, D.; Li, S.; Cui, D. Organometallics 2013, 32,
3
203−3209.
50) Jende, L. N.; Hollfelder, C. O.; Maichle-Mo
Anwander, R. Organometallics 2015, 34, 32−41 and references therein.
51) Zhang, G.; Wei, Y.; Guo, L.; Zhu, X.; Wang, S.; Zhou, S.; Mu, X.
Chem. - Eur. J. 2015, 21, 2519−2526.
52) CRC Handbook of Chemistry and Physics, 64th ed.; Weast, R. C.,
Ed.; CRC Press: Boca Raton, FL, 1983−1984; p D-159.
53) Webster, C. L.; Bates, J. E.; Fang, M.; Ziller, J. W.; Furche, F.;
Evans, W. J. Inorg. Chem. 2013, 52, 3565−3572.
54) Liu, S.; Gao, S.; Ziller, J. W.; Evans, W. J. Dalton Trans. 2014,
3, 15526−15531.
(
̈
ssmer, C.;
(
(
(
(
4
G
DOI: 10.1021/acs.organomet.5b00613
Organometallics XXXX, XXX, XXX−XXX