On the immobilization of a monophosphaferrocene on a silica support

Article abstract of DOI:10.1002/zaac.201200113

Reactions of 2-formyl-3, 4-dimethyl-1-phosphaferrocene with aminopropyl (AP) functionalized hexagonal mesoporous silica (HMS) were studied in the absence and presence of an excess of the mild reducing agent Na[BH ₃CN]. Starting material and reaction products were characterized by analytical and multinuclear (¹³C, ²⁹Si, ³¹P) MAS NMR spectroscopic data. Both reactions proceeded by partial derivatization of surface-bound NH₂ groups to yield products with tethered phosphaferrocene units. Evaluation of spectroscopic data and the results of homogeneous model reactions confirmed that condensation of both components gives imines, which are obviously kinetically labile and allow leakage of the tethered phosphaferrocene into solution. Reactions carried out in the presence of Na[BH₃CN] proceed as reductive amination to yield secondary amines with enhanced kinetic stability. The reductive amination is accompanied by formation of minor amounts of phosphorus-containing by-products as well as the hydrolysis of surface bound ethoxy groups present in the starting material.

Full text of DOI:10.1002/zaac.201200113

ARTICLE
DOI: 10.1002/zaac.201200113
On the Immobilization of a Monophosphaferrocene on a Silica Support
Samith Komath Mallissery[a] and Dietrich Gudat*[a]
Keywords: Phosphaferrocenes; Modified silica; Reductive amination; Surface functionalization; CP-MAS NMR spectroscopy
Abstract. Reactions of 2-formyl-3,4-dimethyl-1-phosphaferrocene
tion of both components gives imines, which are obviously kinetically
with aminopropyl (AP) functionalized hexagonal mesoporous silica labile and allow leakage of the tethered phosphaferrocene into solution.
(HMS) were studied in the absence and presence of an excess of the
mild reducing agent Na[BH3CN]. Starting material and reaction prod-
Reactions carried out in the presence of Na[BH3CN] proceed as re-
ductive amination to yield secondary amines with enhanced kinetic
ucts were characterized by analytical and multinuclear (13C, 29Si, 31P) stability. The reductive amination is accompanied by formation of
MAS NMR spectroscopic data. Both reactions proceeded by partial
derivatization of surface-bound NH2 groups to yield products with
minor amounts of phosphorus-containing by-products as well as the
hydrolysis of surface bound ethoxy groups present in the starting mate-
tethered phosphaferrocene units. Evaluation of spectroscopic data and rial.
the results of homogeneous model reactions confirmed that condensa-
receive currently some attention,[7,8] no such attempts with
Introduction
phosphaferrocenes or their complexes seem to have been re-
ported. Considering that phosphaferrocenes exhibit reasonable
stability and can tolerate a wide range of experimental condi-
tions, we set out to perform a model study in order to evaluate
the prospect of anchoring a functionalized monophosphaferro-
cene to the surface of a modified silica material.
Phosphaferrocenes like compounds 1 and 2 (Scheme 1)
were first prepared in the late 1970ies,[1,2] and the elaboration
of their chemistry has since then received continuing inter-
est.[3] An interesting aspect of these compounds is that the
phosphorus atom is on one hand sufficiently nucleophilic to
support the formation of metal complexes, whereas the ring
system allows on the other hand easy derivatization by electro-
philic substitution.[2,3] This reactivity permits introduction of
extra functionality into the phospholyl moiety, and the combi-
nation of both factors makes phosphaferrocenes an ideal plat-
form for the construction of planar chiral ligands.[4]
Results and Discussion
Even though pure phosphaferrocenes are air-stable species
[1,2]
they resemble other types of low-coordinate phosphorus
compounds since reactions must usually be performed under
exclusion of air and moisture in order to prevent degradation.
Having recently found that even chemically quite stable phos-
phinines (phospha-benzenes) may undergo extensive decom-
position during attempts to link them to the surfaces of func-
[9]
tionalized silica
or gold nanoparticles,[10] we reckoned that
the choice of a proper carrier material and coupling reaction
might also be a critical prerequisite for the successful surface
anchoring of phosphaferrocenes.
Scheme 1.
Mesoporous silicas seemed a suitable material as their sur-
face was proven to display a certain tolerance toward low-
coordinate phosphorus compounds,[9] and can easily be func-
tionalized by coupling with trialkoxysilanes bearing various
functional groups, including phosphines and their com-
plexes.[7] Hexagonal mesoporous silicas (HMS),[11] although
having a less ordered structure than MCM-41 or SBA-15 mate-
rials, are further very easily accessible, and their preparation
does not require strongly acidic conditions. Last, but not least,
aminopropyl (AP) functionalized HMS[12,13] were shown to re-
act under particularly mild conditions with aldehydes to give
materials with covalently tethered imine functions.[14] In view
of these aspects, we decided to explore the prospect of linking
Starting with the first discovery of the catalytic activity of a
phosphaferrocene complex,[5] these ligands have now found a
variety of applications in catalysis.[6] To the best of our knowl-
edge, all catalytic transformations studied so far are based on
homogeneous reactions, and although attempts to immobilize
catalysts in order to improve their separation and recycling
* Prof. Dr. D. Gudat
Fax: +49-711-685-64241
E-Mail: gudat@iac.uni-stuttgart.de
[a] Institut für Anorganische Chemie
Universität Stuttgart
Pfaffenwaldring 55
70550 Stuttgart, Germany
Z. Anorg. Allg. Chem. 2012, 638, (7-8), 1141–1145
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1141

S. Komath Mallissery, D. Gudat
ARTICLE
formyl phosphaferrocene (3)[2] to aminopropyl moieties Q2 groups that had still been observable in the 29Si NMR spec-
grafted to the surface of HMS.
trum of HMS disappeared. Two additional signals at –68.5 and
The preparation of hexagonal mesoporous silica[11] and its –61.5 ppm with intensity ratios of approx. 55:45 (Figure 3) are
functionalization with 3-aminopropyl-triethoxysilane[13] was attributed to T3 and T2 groups [Tm = RSi(OSi)m(OR)3–m, R =
carried out as described. The material was characterized by ele- Et, H] of grafted organosilane moieties.[16] The presence of two
mental analysis, nitrogen adsorption/desorption studies, and 13
C
distinguishable Tn species is in accord with the different rela-
and 29Si CP MAS NMR spectroscopic data. The loading of tive intensities of the 13C NMR signals of ethoxy and propyli-
amine functionalities (elemental analysis revealed the presence dene moieties and the atomic ratio of C:N = 4.2:1 calculated
of 1.7 mmol·g–1 nitrogen, corresponding to an anchoring effi- from the analytical data. All findings together suggest that the
ciency of 58%) and the reduction of the BET surface area sample contains comparable amounts of surface bound
(811 m2·g–1 before and 536 m2·g–1 after amine functionaliza- RSi(OSi)3 (R = 3-aminopropyl, Si = silicon atom in the silica
tion) are similar to previously reported data.[13] A 13C CP-MAS network; calculated C:N ratio = 3:1) and R(EtO)Si(OSi)2 units
NMR spectrum (Figure 1) shows three signals at 6.5, 22.8, and (calculated C:N ratio = 5:1).
40.7 ppm attributable to propyl carbon atoms and two further,
less intense lines at δ = 13.5 and 56.6 ppm that we assign to
residual ethoxy groups. The observed chemical shifts and the
asymmetry of the signal at δ = 22.8 ppm match previously re-
ported values for functionalized silica produced by gas-phase
deposition of aminopropyl-triethoxysilane.[15] The 29Si MAS
and 29Si CP-MAS NMR spectra (Figure 2) show signals arising
from Q4 and Q3 units [Qn = Si(OSi)n(OH)4–n] of the silica
framework[16] at –111.0 and –104.0 ppm, whereas the signal of
Figure 2. Solid state 79.49 MHz 29Si NMR spectra of AP-function-
alized HMS before (traces a, b) and after reaction with 3/Na[BH3CN]
(traces c, d). The labels indicate the assignment of the signals to Tn
and Qn groups (see text). Conditions: trace (a): CP-MAS spectrum, νrot
8 kHz, recycle delay 7.5 s, contact time 1.4 ms, 3072 transients; trace
(b): MAS spectrum, νrot 8 kHz, recycle delay 120 s, 1536 transients;
trace (c): CP-MAS spectrum, νrot 8 kHz, recycle delay 7.5 s, contact
time 1.8 ms, 2048 transients; trace (d): MAS spectrum, νrot 8 kHz,
recycle delay 240 s, 768 transients).
In order to graft phosphaferrocene 3 to the surface of AP-
functionalized HMS, both components were reacted for three
days in anhydrous methanol at 55 °C. Elemental analysis of
the solid residue obtained after decantation of the supernatant
liquid and repeated washing with methanol revealed a signifi-
cant change in the atomic composition (atomic ratio C:N =
7.0:1), which indicated that anchoring of the phosphaferrocene
to the surface had indeed taken place. This assumption was
confirmed by the observation of a signal at –76 ppm in a 31P
Figure 1. Solid state 13C CP-MAS NMR spectra of AP-functionalized
MAS NMR spectrum, which compares well with the signals
of 3 (δ31P –62[2]), and of the single product observed upon
HMS before (top; νrot 8 kHz, recycle delay 5 s, contact time 3 ms,
4096 transients) and after (bottom; νrot 8 kHz, recycle delay 5 s, contact
reaction of 3 with an equimolar quantity of n-propylamine in
time 1.8 ms, 12288 transients) reaction with 3/Na[BH3CN]. The
methanol (δ31P –75[17]). However, when 31P NMR spectra of
sketches illustrate the assignment of the signals to the different carbon
sites.
washing liquids continued to show – even after repeated wash-
1142
www.zaac.wiley-vch.de
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Z. Anorg. Allg. Chem. 2012, 1141–1145

Immobilization of a Monophosphaferrocene on a Silica Support
lytical data reveal a ratio C:N of 5.8:1) and the still lower BET
surface area (397 m2·g–1) as compared to AP-modified HMS
give strong evidence that anchoring of phosphaferrocene has
taken place, and the failure to observe spectroscopically detect-
able amounts of 3 in washing liquids (after any unreacted start-
ing material was removed) suggest that the immobilized mate-
rial is indeed connected to the surface by a stable covalent
tether. However, the observation of a smaller C:N ratio than in
the previously described imine coupling product also indicates
that the presence of the reducing agent has an adverse effect
on the phosphaferrocene loading.
The presence of grafted phosphaferrocene moieties on the
surface is confirmed by the MAS NMR spectroscopic data. A
31P NMR spectrum recorded with direct excitation of 31P and
1
high power H decoupling displays a signal around –75 ppm
with a shoulder on the left side and indicates the presence of
two phosphaferrocene species with slightly different chemical
environment (spectral deconvolution gave chemical shifts of
–76.1 and –70.3 with relative intensities of 77:13). A further
weak resonance at δ = +42 ppm (10% of total spectral inten-
sity) reveals a minor amount of decomposition products (aris-
ing presumably from nucleophilic attack at the phosphorus
atom and subsequent decomplexation of the phosphole ring[2]).
Figure 3. Expansion of the solid state 79.49 MHz 29Si CP-MAS NMR
spectra of AP-functionalized HMS before (top trace) and after (bottom
trace) reaction with 3/Na[BH3CN]. The grey lines show the results of
fits of the experimental lineshapes to a sum of two mixed Gaussian/
Lorentzian lines representing distinguishable T2/T3 environments.
Computation of the integrated relative intensities yielded in both cases
a ratio of n(T3):n(T2) = 55:45.
A
31P CP-MAS NMR spectrum showed the same signals but
with much lower intensity. This effect reflects presumably the
known reduction of CP-efficiency at high spinning speeds,
which is due to the splitting of the Hartmann-Hahn matching
profile into separated sidebands and has previously been ob-
served in CP-MAS NMR spectra of other surface bound phos-
phorus compounds;[20] since the spectrum recorded without CP
ings – a signal attributable to 3, we concluded that the imine
formation is reversible, and the chemical tether is too labile to
prevent leaking of immobilized phosphaferrocene into solu-
tion.
Since it is well known that cleavage of labile aldimines may
be suppressed by their reduction to kinetically inert amines,
we sought a way to improve our immobilization protocol by
employing this strategy. After some testing we found that re-
ductive amination of 3 is feasible if the coupling with a pri-
mary amine is carried out in the presence of a large excess of
the mild reducing agent Na[BH3CN].[18] Thus, phosphaferro-
cenylamine (4) (Scheme 2) was obtained in a single step from
3 and benzylamine as red oil that was characterized by spectro-
scopic data, and formation of an analogous coupling product
was confirmed by 31P NMR spectroscopic monitoring of the
reaction of 3 and 3-aminopropyl-triethoxysilane.[19]
Figure 4. Solid state 161.9 MHz 31P MAS NMR spectrum of AP-func-
tionalized HMS after reaction with 3/Na[BH3CN] (νrot = 14 kHz, 1024
transients, relaxation delay 5 s) The top trace shows the experimental
spectrum, the middle trace the result of a simulation of the experimen-
tal lineshape as the sum of three spinning sideband manifolds, and the
bottom trace the difference between experiment and fit. The labels
indicate the assignment of the isotropic lines to an unknown decompo-
sition product (D) and two distinguishable phosphaferrocene sites (P1,
P2). The integrated intensities indicate a molar ratio of D:P1:P2 =
Scheme 2.
Coupling of 3 to the surface of AP-modified HMS was car-
ried out using the same protocol as in the synthesis of 4, and
the resulting brownish material was characterized by elemental
analysis, nitrogen adsorption/desorption, and CP MAS NMR
spectroscopic data. The different atomic composition (the ana- 10:13:77.
Z. Anorg. Allg. Chem. 2012, 1141–1145
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.zaac.wiley-vch.de
1143

S. Komath Mallissery, D. Gudat
ARTICLE
was of sufficient quality, we did not optimize the Hartmann– be much lower than in attempts to immobilize other types of
Hahn match (Figure 4Ͻyigr4 pos="x22"Ͼ).
low-coordinate phosphorus compounds,[9] and the phosphafer-
The tethering of intact phosphaferrocene is also evident rocene core remained largely intact. These findings led us to
from the 13C CP-MAS NMR spectrum (Figure 1). Comparison conclude that (i) the phosphaferrocene moiety is obviously less
with the spectrum of 4 allows assigning all carbon sites in the susceptible towards side-reactions during functionalization and
phosphaferrocene moiety. Four additional signals are attribut- seems therefore more suitable for immobilization than other
able to the carbon atoms in the propylidene spacer and the types of low-coordinate phosphorus compounds, (ii) the appli-
methylene group linking the amine and phosphaferrocene cation of as mild reaction conditions as possible is apparently
functionalities. Resonances of ethoxy units, which had been a crucial prerequisite for the immobilization of any type of
present in the AP-modified silica, are no longer visible. Similar low-coordinate phosphorus compounds, and (iii) still further
findings had previously been noted during the post-functionali- efforts are required to prepare well defined materials that con-
zation of similar modified silicas[22] and indicate that these tain no decomposition products and allow to exploit the full
groups hydrolyze during the grafting process. Comparison of potential of phosphaferrocenes in heterogeneous catalysis.
relative signal intensities reveals that the resonances of the car-
bon atoms in the propylidene spacer are much stronger than
Experimental Section
those in the phosphaferrocene moiety. Even though no quanti-
tative evaluation was feasible, this finding supports the earlier
mentioned hypothesis that only incomplete functionalization of
the anchoring groups had been accomplished.
General Remarks: All manipulations were carried out in an argon
atmosphere using Schlenk techniques. Solvents were dried prior to use
by common procedures. Preparation of phosphaferrocene aldehyde
3,[4] HMS,[11] and functionalization of the latter with 3-aminopropyl-
triethoxysilane [13] were carried out as described. Solution NMR spec-
tra were recorded with a Bruker Avance 250 spectrometer (1H:
250.1 MHz, 13C: 62.8 MHz, 31P: 101.2 MHz) at 303 K and solid-state
The 29Si MAS and CP-MAS NMR spectra resemble closely
those of the AP-functionalized silica before treatment with
phosphaferrocene and indicate that the silica framework has
not undergone any structural changes during the grafting pro-
cess. In particular, the ratio of T2 and T3 sites is essentially
unaffected (Figure 3), which suggests that the observed loss of
ethoxy groups does not turn the former RSi(OEt)(OSi)2 into
RSi(OSi)3 groups as had previously been proposed,[22] but may
possibly produce RSi(OH)(OSi)2 units with a further free sil-
anol function. The coexistence of such T2 and T3 sites on the
silica surface implies that tethered phosphaferrocene moieties
may exhibit different chemical environments, and suffices to
explain the shoulder on the 31P NMR signal of surface bound
phosphaferrocenes.
NMR spectra with
a
Bruker Avance 400 spectrometer (13C:
100.5 MHz, 29Si: 79.49 MHz, 31P: 161.9 MHz) equipped with a 4 mm
MAS probe. MAS experiments were performed using standard ZrO2
rotors and spinning speeds between 8 and 9 kHz for 13C and 29Si NMR
spectra and 14 kHz for 31P NMR spectra. Cross polarization was ap-
plied using a ramp-shaped contact pulse and mixing times between 0.8
and 5 ms unless noted. Chemical shifts were referenced to ext. TMS
(1H,13C) or 85% H3PO4 (Ξ = 40.480747, 31P). EI-MS: Varian MAT
711, 70eV. Elemental analysis: Perkin–Elmer 24000 CHN/O Analyser.
Nitrogen adsorption-desorption experiments were carried out with a
Micromeritics ASAP 2000 particle size analyzer.
Additional experiments revealed that although the amount
of side products (as indicated by the observation of a minor
31P NMR signal around 40 ppm) varied somewhat between
different reactions, their formation could not be completely
avoided. Pre-treatment of the functionalized silica with
HN(SiMe3)2 (in order to deactivate silanol functions on the
surface) did not lead to a significant improvement, thus sug-
gesting that the observed side products might arise from side-
reactions during the reductive amination step.
2-(Benzylamino)methyl-3,4-dimethyl-phosphaferrocene (4): Ben-
zylamine (295 mg, 2.76 mmol) was added to 2-formyl-3,4-dimethyl-
1-phosphaferrocene (720 mg, 2.76 mmol) in MeOH (20 mL). Solid
NaBH3CN (1.70 g, 27.6 mmol) was added, and the mixture was heated
at 55 °C for 5 h. Deoxygenated water (20 mL) was added. The organic
phase was separated and filtered through a short plug of neutral alu-
mina. The solvent was removed under vacuum to obtain an orange oil.
Yield 720 mg, 74%. 31P{1H} NMR (CD3OD): δ = –76.5 ppm. 1H
NMR (CD3OD): δ = 2.10 (s, 3 H, CH3), 2.17 (s, 3 H, CH3), 3.16 (dd,
2
3JPH = 13.5 Hz, 2 H, CH2N), 3.32 (s, 2 H, CH2N), 3.63 (d, JPH
=
37.0 Hz, 1 H, PCH), 4.06 (s, 5 H, Cp), 7.21–7.35 (m, 5 H, C6H5) ppm.
13C{1H} NMR (CD3OD): δ = 13.8 (s, CH3), 17.3 (s, CH3), 53.73 (d,
Conclusions
1
2JPC = 1.7 Hz, CH2N), 54.76 (s, CH2N), 73.01 (s, Cp), 76.70 (d, JPC
2
1
It was shown that persistent grafting of phosphaferrocene
aldehyde 3 to the surface of AP-functionalized silica may be
accomplished by coupling both components in the presence of
the mild reducing agent Na[BH3CN]. Solid-state NMR studies
and model reactions carried out in solution suggest assigning
the formed product the structure of an aminomethylated phos-
phaferrocene. The NMR studies indicate further that only par-
tial derivatization of the surface functionalities was achieved,
and that the grafting process is accompanied by hydrolytic
cleavage of surface bound ethoxy functions. Although forma-
tion of phosphorus-containing by-products could not com-
= 58.3 Hz, α-CH), 94.64 (d, JPC = 4.9 Hz, β-C), 96.26 (d, JPC
=
2
58.9 Hz, α-C), 97.43 (d, JPC = 6.8 Hz, β-C), 123.75, 129.47, 130.72,
140.5 (C6H5) ppm. MS (EI = 70 eV): m/e (%) = 351.1 (100) [M]+,
244.0 (44) [M – NHCH2C6H5 + H]+.
Immobilization of 3 on AP-functionalized HMS: (a) With Imine
Coupling: Compound 3 (400 mg, 1.52 mmol) was stirred with AP-
functionalized HMS (2.00 g) in anhydrous MeOH (15 mL) at 55 °C
for 3 d. The material was filtered, washed with methanol (3ϫ50 mL),
and dried in vacuo for 8 h. Analytical data: found C 13.49 H 2.67 N
2.22 %. MAS 31P{1H} NMR: δ = –76 ppm. Repeated washings with
MeOH produced washing liquids, whose 31P NMR spectra continu-
pletely be avoided, the amount of these species was found to ously displayed a signal attributable to 3.
1144 www.zaac.wiley-vch.de
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Z. Anorg. Allg. Chem. 2012, 1141–1145

Immobilization of a Monophosphaferrocene on a Silica Support
P. C. J. Kamer, P. W. N. M. van Leeuwen, J. Am. Chem. Soc.
2001, 123, 8468–8476; f) Z. Lu, E. Lindner, H. Mayer, Chem.
Rev. 2002, 79,102, 3543; g) P. McMorn, G. Hutchings, J. Chem.
Soc. Rev. 2004, 33, 108; h) J. M. Thomas, R. Raja, J. Organomet.
Chem. 2004, 689, 4110–4124; i) N. C. Mehendale, J. R. A. Si-
etsma, K. P. de Jong, C. A. van Walree, R. J. M. Klein Gebbink,
G. van Koten, Adv. Synth. Catal. 2007, 349, 2619–2630; j) Y.
Yang, B. Beele, J. Blümel, J. Am. Chem. Soc. 2008, 130, 3771–
3773.
(b) With Reductive Amination: Compound 3 (400 mg, 1.52 mmol)
was added to AP-functionalized HMS (2.00 g) in anhydrous MeOH
(15 mL). Solid Na[BH3CN] (470 mg, 7.6 mmol) was added and the
reaction mixture was heated to 50 °C for 3 d. The suspension was
filtered, the residue was washed with MeOH (3ϫ50 mL), and dried
in vacuo for 8 h. Analytical data: found C 11.69 H 2.94 N 2.34 %.
MAS NMR spectra are displayed in Figures 1–4. 13C CP-MAS NMR
(assignments refer to the labels shown in Figure 1): δ = 95 (h, i, j),
72.9 (g), 70.1 (k), 47.6 (c), 41.6 (d), 23.8 (b), 19.6 (e,f), 7.1 (a) ppm.
[8] F. R. Hartley, in Supported Metal Complexes (Eds.: R. Ugo, B. R.
James, D. Reidel), Dordrecht, Holland, 1985 and literature cited
therein. An overview on state-of-the-art techniques is given in
Chem. Rev. 2002, 79,102.
[9] S. Komath Mallissery, M. Nieger, D. Gudat, Z. Anorg. Allg.
Chem. 2010, 636, 1356–1360.
[10] S. Komath Mallissery, D. Gudat, Dalton Trans. 2010, 39, 4280–
4284.
[11] P. T. Panev, T. J. Pinnavaia, Science 1995, 267, 865.
[12] C. Yang, Chin. Chem. Lett. 2003, 14, 96–99.
[13] A. R. Silva, K. Wilson, A. C. Whitwood, J. H. Clark, C. Freire,
Eur. J. Inorg. Chem. 2006, 1275.
[14] a) K. Sarkar, K. Dhara, M. Nandi, P. Roy, A. Bhaumik, P.
Banerjee, Adv. Funct. Mater. 2009, 19, 223–234; b) M. Nandi, P.
Roy, H. Uyama, A. Bhaumik, Dalton Trans. 2011, 40, 12510–
12518.
[15] S. Ek, E. I. Iiskola, L. Niinisto, J. Vaittinen, T. T. Pakkanen, A.
Root, J. Phys. Chem. B 2004, 108, 11454–11463.
Acknowledgement
Financial support by the Deutsche Forschungsgemeinschaft (Graduate
College 448 “Modern NMR Methods in Material Science”; scholarship
for S.K.M.) is gratefully acknowledged. We thank Mrs. Tahira Yasmin
(Institute of Physical Chemistry) for performing the nitrogen adsorp-
tion-desorption measurements, Mr. J. Trinkner (Institute of Organic
Chemistry) for measurement of mass spectra, and Mrs. B. Förtsch (In-
stitute of Inorganic Chemistry) for elemental analyses.
References
[1] F. Mathey, A. Mitschler, R. Weisss, J. Am. Chem. Soc. 1977, 99,
3537.
[2] a) G. deLauzon, B. Deschamps, J. Fischer, F. Mathey, A.
Mitschler, J. Am. Chem. Soc. 1980, 102, 994; b) G. deLauzon, B.
Deschamps, F. Mathey, Nouv. J. Chim. 1980, 4, 683.
[16] G. Engelhardt, D. Michel, High Resolution Solid-State NMR of
Silicates and Zeolites, Wiley, New York, 1987.
[17] Although the occurrence of hydrolytic cleavage of the imine moi-
ety during work-up prevented isolation of the product, the ex-
pected constitution was further confirmed by 1H NMR spectro-
scopic data obtained directly from the reaction mixture: δ1H =
[3] For comprehensive reviews on the chemistry of phosphaferroc-
enes, see: a) F. Mathey, Coord. Chem. Rev. 1994, 137, 1; b) F.
Mathey, J. Organomet. Chem. 1990, 400, 149; c) F. Mathey, J.
Fischer, J. H. Nelson, Struct. Bonding (Berlin) 1983, 55, 154.
[4] a) C. Ganter, L. Brassat, C. Glinsböckel, B. Ganter, Organometal-
lics 1997, 16, 2862–2867; b) C. Ganter, L. Brassat, B. Ganter,
Chem. Ber./Recueil 1997, 130, 1771–1776; c) S. Qiao, D. A.
Hoic, G. C. Fu, Organometallics 1998, 17, 773–774.
[5] C. E. Garrett, G. C. Fu, J. Org. Chem. 1997, 62, 4534.
[6] a) S. Qiao, G. C. Fu, J. Org. Chem. 1998, 63, 4168–4169; b) C.
Ganter, C. Kaulen, U. Englert, Organometallics 1999, 18, 5444–
5446; c) C. Ganter, C. Glinsböckel, B. Ganter, Eur. J. Inorg.
Chem. 1998, 1163–1168; d) K. Tanaka, S. Qiao, M. Tobisu,
M. M.-C. Lo, G. C. Fu, J. Am. Chem. Soc. 2000, 122, 9870–9871;
e) K. Tanaka, G. C. Fu, J. Org. Chem. 2001, 66, 8177–8186; f)
M. Ogasawara, K. Yoshida, T. Hayashi, Organometallics 2001,
20, 3913–3917; g) R. Shintani, M. M.-C. Lo, G. C. Fu, Org. Lett.
2000, 2, 3695–3697; h) R. Shintani, G. C. Fu, Org. Lett. 2002, 4,
3699–3702; i) R. Shintani, G. C. Fu, Angew. Chem. Int. Ed. 2003,
42, 4082–4085.
3
2
8.17 (d, JHP = 8.8 Hz, CH = N), 3.95 (d, JHP = 37.4 Hz, PCH),
4.20 (s, Cp), 3.12 (m, NCH2) 2.31 (s, CH3), 2.21 (s, CH3), 1.57
(m, CH2), 0.89 (t, CH3).
[18] For a review of the use of cyano-borohdride in reductive amina-
tions see: R. O. Hutchins, M. K. Hutchins, in Comprehensive Or-
ganic Synthesis (Eds.: B. N. Trost, I. Fleming), Pergamon Press,
New York 1991, vol. 8, pp 25–78.
[19] The 31P NMR spectrum showed the signals of two products at
–75.4 ppm (major product) and 76.1 ppm (minor product). As the
specific reactivity of the Si(OEt)3 moiety made chromatographic
separation unfeasible, the nature of the by-product remains in the
dark.
[20] S. Reinhard, J. Blümel, Magn. Reson. Chem. 2003, 41, 406–416.
[21] It is further worthwhile to mention that the intensity of the signal
of the isolated NCH2 group is substantially higher than those of
the carbon atoms in the phosphaferrocene unit and matches the
intensities of the signals attributable to the carbon atoms in the
propylidene spacer. A concise interpretation of this effect is cur-
rently unfeasible.
[7] For some representative examples see: a) W. E. Rudzinski, T. L.
Montgomery, J. S. Frye, B. L. Hawkins, G. E. Maciel, J. Catal.
1986, 98, 444–456; b) H. Gao, R. J. Angelici, J. Am. Chem. Soc.
1997, 119, 6937; c) O. Kröcher, R. A. Köppel, M. Fröba, A.
Baiker, J. Catal. 1998, 178, 284–298; d) H. Gao, R. J. Angelici,
Organometallics 1999, 18, 989; e) A. J. Sandee, J. N. H. Reek,
[22] J. Li, L. Wang, T. Qi, Y. Zhou, C. Liu, J. Chu, Y. Zhang, Micro-
porous Mesoporous Mater. 2008, 110, 442–450.
Received: March 14, 2012
Published Online: June 04, 2012
Z. Anorg. Allg. Chem. 2012, 1141–1145
© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.zaac.wiley-vch.de
1145