Cyclic denitrogenation of N-heterocycles applying a homogeneous titanium reagent

Article abstract of DOI:10.1021/ja075326n

A transient titanium alkylidyne (PNP)Ti≡C^tBu (PNP = N[2-P(CHMe₂)₂-4-methylphenyl]₂^-) offers an excellent opportunity to study an industrially important reaction such as the hydrodenitrogenation (HDN) of N-heterocycles present in crude oil. This type of molecule has allowed us to quantitatively remove the nitrogen, in a cyclic manner and under mild conditions, from heterocycles such as pyridine and picolines, resulting in formation of the corresponding arene and a titanium nitrogen-based product. Applying this complex for N-removal of pyridines avoids the use of drastic reaction conditions, over saturation of substrate (hydrogenation), and the use of hydrogen, all of which are common limitations in the HDN industrial setting. Isotopic labeling and mechanistic insights into both the C-N bond rupture and denitrogenation of N-heterocycles are also discussed in this report. Copyright

Full text of DOI:10.1021/ja075326n

Published on Web 09/29/2007
Cyclic Denitrogenation of N-Heterocycles Applying a Homogeneous Titanium
Reagent
Alison R. Fout, Brad C. Bailey, John Tomaszewski, and Daniel J. Mindiola*
Department of Chemistry, Indiana UniVersity, Bloomington, Indiana 47405
Received July 17, 2007; E-mail: mindiola@indiana.edu
Scheme 1. Ring-opening of N-Heterocycles
Catalytic conversion of N-heterocycles present in petroleum or
coal-based liquids to ammonia and nitrogen-free carbon based
products, a process referred to as hydrodenitrogenation (HDN), is
an important petrochemical transformation since it reduces the
emissions of NOx upon combustion of these fuels.1 Unfortunately,
the mechanism by which the HDN heterogeneous catalysts, typically
a sulfided CoMo/Al2O3 or NiMo/Al2O3 surface, break the strong
C-N bonds in N-heterocycles is far from being understood.2
Although HDN, hydrodesulfurization (HDS), and hydrodeoxygen-
ation (HDO) are procedures performed simultaneously during the
hydrotreating process, current technology applies conditions opti-
mized only for HDS.1c,2a,3 As a result, sulfur removal is more
efficient since HDN would only occur for a fraction of substrates
such as aliphatic amines, given that these are easier to hydrodenitro-
genate.2b,3,4 Hence, the current practice of HDN needs to be
understood and possibly optimized, since robust N-heterocycles
often carry on untouched under the current standard conditions
(2000 psi H2, 300-450 °C).1d Understanding the intrinsic details
behind some of the key steps in HDN, especially C-N bond
cleavage, would benefit from prototypical homogeneous models
which could ultimately perform the analogous transformation under
much milder conditions.1c,4c,5
Scheme 2. Denitrogenation of N-Heterocycles
Scheme 3. Proposed Mechanism for the Denitrogenation of
N-Heterocycles
While studying intermolecular C-H activation reactions of
hydrocarbons, we recently discovered that compound (PNP)Tid
t
CHtBu(CH2 Bu) (1)6 (PNP ) N[2-P(CHMe2)2-4-methylphenyl]2 can
ring-open both pyridine or picolines at room temperature over 12
h to afford azametallabicyclic systems of the type (PNP)Ti(C-
(tBu)CC4H3RNH) (R ) H, 2a; R ) Me, 2b and 2c; Scheme 1).7 In
the case of 3-methylpicoline only one regioisomer, 2c, was
generated. The likely intermediates leading to 2a have already been
probed by independent experiments as well as high-level DFT
analysis.7 We report here that the former pyridine nitrogen of
complexes such as 2a-c can be denitrogenated, under mild condi-
tions and in a cyclic manner, when treated with an electrophile
such as (CH3)3SiCl. Synthetic, isotopic labeling and preliminary
mechanistic studies addressing the denitrogenation process are
presented.
{Si(CH3)3}(Cl) (3)-15N is quantitatively formed (15N NMR: 553.9
ppm, 2JN-P ) 2.3 Hz)8a thus unarguably supporting the notion that
the trimethylsilylimide nitrogen in 3 is derived from pyridine
denitrogenation.
Formation of 3 from 2a-c and (CH3)3SiCl is proposed to occur
first by silylation of the R-nitrogen composing the azametallabi-
cyclic ring (intermediate A), followed by a ring expansion applying
a 1,3-hydrogen shift to afford the eight-membered ring intermediate
B. By means of an electrocyclic rearrangement, we then propose
that B would ring-contract to generate intermediate C, which then
would [2+2] retrocyclo-extrude the substituted arene product to
finally engender formation of the imide, 3 (Scheme 3). Electrocyclic
rearrangements involving cyclooctatriene to the bicyclic ring-fused
system are well-known, and similar ring-contraction reactions in
eight-azamembered metal rings have been documented previously
by Wigley and co-workers.10 We have found that the rate of the
reaction for the 2a f 3 conversion does not appear to be dependent
on the concentration of the (CH3)3SiCl at 80 °C (0.018 M, k )
10.2(37) × 10-5 s-1; 0.18 M, k ) 9.18(23) × 10-5 s-1; 79.9 M, k
) 11.7(12) × 10-5 s-1)9 thus implying that the proposed conversion,
2 f A, is likely not the slowest step in the denitrogenation sequence.
We cannot however, rule out the possibility of a pre-equilibrium
phenomena along the steps 2 f C.
Despite compound 2 being remarkably stable to heat (120 °C
for 4 days under N2 or Ar), treatment of such with (CH3)3SiCl
(excess or 1 equiv) over 72 h at 65 °C results in formation of (PNP)-
TidN{Si(CH3)3}(Cl) (3), concurrent with extrusion of the substi-
t
tuted arene BuAr (Ar ) C6H5, 4-MeC6H4, 3-MeC6H4) (Scheme
2). While the former product has been independently synthesized
and characterized,8 the latter organic side-material was confirmed
1
by a combination of H, and 13C NMR spectral data. When using
1 equiv of electrophile, vacuum transfer of the volatiles reveals
exclusive formation of the arene by multinuclear NMR spectroscopy
with no other hydrocarbon-based products being detected by GC-
MS, and 1H NMR spectroscopy.9 If the same reaction is monitored
with the isotopomer (PNP)Ti(C(tBu)CC4H415NH) (2a)-15N, derived
1
from 15N enriched pyridine9 and 1 (15N NMR: 131.8 ppm, JN-H
) 58.4 Hz, 3JN-H ) 6.0 Hz), the 15N enriched imide (PNP)Ti)15N-
9
12640
J. AM. CHEM. SOC. 2007, 129, 12640-12641
10.1021/ja075326n CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
Scheme 4. Protonation of 2 and Subsequent Tautomerization
5), transmetallation transpires to afford the known Ti(III) bisalkyl
t
precursor (PNP)Ti(CH2 Bu)2 (7)11 (Scheme 5). One-electron oxida-
tion of 6 with AgOTf furnishes (PNP)TidCHtBu(OTf) (8),11 and
t
subsequent treatment of the latter with 1 equiv with LiCH2 Bu
produces 16 thereby closing the homogeneous cycle for the
denitrogenation of pyridine or picolines (Scheme 5). In fact, for
the denitrogenation of pyridine, the overall yield for the recycling
of titanium reagent was 18%.
Scheme 5. Denitrogenation of N-Heterocycles Using a Recyclable
The overall conversion of 1 to 3 in the presence of (CH3)3SiCl
advocates that a complete “CtBu” for “N” metathetical reaction with
pyridines and picolines has taken place, whereby a transient titanium
alkylidyne, (PNP)TitCtBu, denitrogenates the N-heterocycles
concurrent with transfer of the alkylidyne moiety. In the process,
three strong C-N bonds composing the N-heterocycle ring have
been cleaved under mild conditions and without the need of a
reducing agent such as H2. As a result, the hydrocarbon based-
product generated in our set of denitrogenation reactions is not
saturated. We speculate that the use of an electrophile such as
(CH3)3SiCl was needed to make formation of the titanium imide
much more thermodynamically within reach. Our work therefore
demonstrates that “N” removal of pyridine can be achieved under
mild conditions and in a cyclic manner, and that such a process is
promoted by an azophilic Ti center as well as highly polarized Ti-C
multiple bond. We are currently performing more detailed kinetic
studies on the 2 f 3 conversion.
Titanium Reagent
Acknowledgment. We thank the Sloan Foundation and the NSF
(Grant CHE-0348941) for financial support of this research.
Supporting Information Available: Experimental preparation,
reactivity (all compounds), and additional discussion. This material is
available free of charge via the Internet at http://pubs.acs.org.
To ascertain if the electrophile adds to the R-N in compounds
such as 2a we explored a close analogue. Accordingly, we found
that when complex 2a is treated with an H+ source such as [HNMe2-
Ph][B(C6F 5)4], denitrogenation does not occur. Instead, protonation
and tautomerization ensues to produce [(PNP)Ti(NH2(CtBu)C5H4)]-
References
(1) (a) Fish, R. H.; Thormodsen, A. D.; Moore, R. S.; Perry, D. L.; Heinemann,
H. J. Catal. 1986, 102, 270. (b) Fish, R. H.; Michaels, J. N.; Moore, R.
S.; Heinemann, H. J. Catal. 1990, 123, 74. (c) Ho, T. C. Catal. ReV. -
Sci. Eng. 1988, 30, 117. (d) Satterfield, C. N. In Heterogeneous Catalysis
in Industrial Practice, 2nd ed.; McGraw-Hill: New York, 1991. (e) Katzer,
J. R.; Sivasubramanian, R. Cat. ReV. - Sci. Eng. 1979, 20, 155.
(2) (a) Laine, R. M.; Thomas, D. W.; Cary, L. W. J. Org. Chem. 1979, 44,
4964. (b) Laine, R. M. Catal. ReV. - Sci. Eng. 1983, 25, 459.
(3) Sivasubramanian, R.; Crynes, B. L. Ind. Eng. Chem. Process Des. DeV.
1979, 18, 175.
(4) (a) Laine, R. M. Ann. N.Y. Acad. Sci. 1983, 415, 271. (b) Laine, R. M. J.
Mol. Catal. 1983, 21, 119. (c) Laine, R. M. New J. Chem. 1987, 11, 543.
(d) Laine, R. M.; Thomas, D. W.; Cary, L. W. J. Am. Chem. Soc. 1982,
104, 1763. (e) Kabir, S. E.; Day, M.; Irving, M.; McPhillips, T.; Minassian,
H.; Rosenberg, E.; Hardcastle, K. I. Organometallics 1991, 10, 3997. (f)
Shvo, Y.; Laine, R. M. J. Chem. Soc., Chem. Commun. 1980, 753. (g)
Adams, R. D.; Chen, L.; Wu, W. Organometallics 1993, 12, 4962. (h)
Adams, R. D.; Chen, G. Organometallics 1993, 12, 2070. (i) Adams, R.
D.; Chen, G. Organometallics 1992, 11, 3510.
(5) (a) Sanchez-Delgado, R. A. In Organometallic Modelling of the Hydrode
Sulfurization and Hydrodenitrogenation Reactions; Kluwer Academic
Publishers: Boston, MA, 2002. (b) Weller, K. J.; Fox, P. A.; Gray, S. D.;
Wigley, D. E. Polyhedron 1997, 16, 3139. (e) Kleckley, T. S.; Bennett,
J. L.; Wolczanski, P. T.; Lobkovsky, E. B. J. Am. Chem. Soc. 1997, 119,
247.
(6) (a) Bailey, B. C.; Fan, H.; Baum, E. W.; Huffman, J. C.; Baik, M.-H.;
Mindiola, D. J. J. Am. Chem. Soc. 2005, 127, 16016. (b) Bailey, B. C.;
Fan, H.; Huffman, J. C.; Baik, M.-H.; Mindiola, D. J. J. Am. Chem. Soc.
2007, 129, 8781-8793.
(7) Bailey, B. C.; Fan, H.; Huffman, J. C.; Baik, M.-H.; Mindiola, D. J. J.
Am. Chem. Soc. 2006, 128, 6798.
1
[B(C6F5)4] (4), a complex characterized by a combination of H,
13C, 31P, 11B, and 19F NMR spectra (Scheme 4). Formation of 4
and its tautomer were spectroscopically evident when (PNP)Ti-
(C(tBu)CC4D4ND) (2a)-d5 (prepared from NC5D5) was treated with
[HNMe2Ph][B(C6F5)4] to yield the mixture of tautomers (4)-d5
(Scheme 4). This result implies that an electrophile is likely to
coordinate first to the former pyridine or picoline nitrogen in 2.
Changing the anion from Cl to B(C6F5)4 does not appear to impede
both the ring-expansion, the electrocyclic rearrangement, and ring-
extrusion steps for the A f B f C f 3 sequence, since addition
of [Et3Si][B(C6F5)4] to 2a-c also effects arene extrusion (Scheme
3).9 However, we have been unable to characterize the metal-based
byproduct for the latter reaction.9
We have also determined that product 3 can be deiminated with
electrophiles such as TiCl4(THF)2, NbCl5, and PCl5 to afford the
known complex (PNP)TiCl3 (5).11 In all cases however, separation
of the imide oligomer from 5 was hampered by their similar
solubilities. We speculated whether an insoluble product would
facilitate purification of 3 for further reaction chemistries since
substitution of an imide for two chlorides would allow us to possibly
recycle 5 back to 1. To achieve this goal, we turned our attention
to MoCl5, since the reported nitride Mo2Cl7(µ2-N)12 is a highly
insoluble compound. Gratifyingly, treatment of 3 with ∼2 equiv
of MoCl5 resulted in clean formation of 5 in 75% yield concurrent
with precipitation of the black insoluble product that we propose
to be Mo2Cl7(µ2-N)12 (ClSi(CH3)3 was observed by 1H NMR
(8) (a) Bailey, B. C.; Fout, A. R.; Fan, H.; Tomaszewski, J.; Huffman, J. C.;
Gary, J. B.; Johnson, M. J. A.; Mindiola, D. J. J. Am. Chem. Soc. 2007,
129, 2234-2235. (b) Bailey, B. C.; Basuli, F.; Huffman, J. C.; Mindiola,
D. J. Organometallics 2006, 25, 2725.
(9) See Supporting Information for complete details.
(10) Gray, S. D.; Weller, K. J.; Bruck, M. A.; Briggs, P. M.; Wigley, D. E. J.
Am. Chem. Soc. 1995, 117, 10678.
(11) Bailey, B. C.; Huffman, J. C.; Mindiola, D. J.; Weng, W.; Ozerov, O. V.
Organometallics 2005, 24, 1390.
(12) Godemeyer, T.; Dehnicke, K. Z. Anorg. Allg Chem. 1988, 558, 114.
t
spectroscopy). Upon treatment of 5 with 3 equiv of LiCH2 Bu, a
reduction (presumably via formation of (PNP)TiCl2 (6), Scheme
JA075326N
9
J. AM. CHEM. SOC. VOL. 129, NO. 42, 2007 12641