An improved bidentate complex of iridium as a catalyst for hydrogen isotope exchange

Article abstract of DOI:10.1002/jlcr.921

Iridium(I) complexes 1, containing bidentate phosphines, and 3, with arsine ligands, are generated in situ. These species mediate hydrogen isotope exchange in a variety of aromatic substrates including benzyl ketones. Although the catalytic activities of complexes 1 and 3 are generally unexceptional, a logical step leads to the use of [ethylene-1,2-bis(diphenylarsine)](cyclooctadiene) iridium(I) tetrafluoroborate (5), which is an efficient catalyst for both aryl and benzyl ketones, and mediates exchange to a substantial extent in other substrates also. Copyright

Full text of DOI:10.1002/jlcr.921

JOURNAL OF LABELLED COMPOUNDS AND RADIOPHARMACEUTICALS
J Label Compd Radiopharm 2005; 48: 285–294.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jlcr.921
Research Article
An improved bidentate complex of iridium as a catalyst
for hydrogen isotope exchange
John M. Herbert*, Andrew D. Kohler and Alan H. McNeill
Isotope Chemistry and Metabolite Synthesis Department, Sanofi-aventis, Willowburn
Ave., Alnwick, Northumberland NE66 2JH, UK
Summary
Iridium(I) complexes 1, containing bidentate phosphines, and 3, with arsine ligands,
are generated in situ. These species mediate hydrogen isotope exchange in a variety
of aromatic substrates including benzyl ketones. Although the catalytic activities
of complexes 1 and 3 are generally unexceptional, a logical step leads to the use
of [ethylene-1,2-bis(diphenylarsine)](cyclooctadiene)iridium(I) tetrafluoroborate (5),
which is an efficient catalyst for both aryl and benzyl ketones, and mediates exchange
to a substantial extent in other substrates also. Copyright # 2005 John Wiley & Sons,
Ltd.
Key Words: iridium complexes; deuterium exchange; isotope exchange
Introduction
The majority of the iridium complexes developed as catalysts for ortho-
hydrogen isotope exchange are effective only where the directing heteroatom
in the substrate is separated from the arene by a single carbon. Thus, aryl
ketones, benzoate esters, benzamides, and a number of arylheterocycles can be
exchanged with reasonable efficiency.1–4 In principle, exchange in homologous
species such as phenylacetone or derivatives of phenylacetic acid should be
possible but, in practice, complexes of simple phosphines do not mediate such
processes readily, although exchange with acetanilide and its derivatives does
occur in some cases.3,4
Substrates such as ethyl phenylacetate are, in fact, exchanged by bidentate
phosphine complexes such as 1a.5 This observation can be rationalised by
considering the mechanism of exchange (Scheme 1). The adduct A is believed
*Correspondence to: J. M. Herbert, Isotope Chemistry and Metabolite Synthesis Department, Sanofi-
aventis, Willowburn Ave., Alnwick, Northumberland NE66 2JH, UK. E-mail: john.herbert@sanofi-
synthelabo.com
Copyright # 2005 John Wiley & Sons, Ltd.
Received 29 October 2004
Revised 26 November 2004
Accepted 2 December 2004

286
J. M. HERBERT ET AL.
L-L
L-L
+
a. Ph2PCH2CH2PPh2
b. Ph2PCH2CH2CH2PPh2
d. o-C6H4(PPh2)2
L
-
BF4
Ir
c. Ph2PCH2PPh2
L
e. 2,2′-bis(diphenylphosphinyl)-
1,1′-binaphthyl) [(+)-BINAP]
g. 9,9-dimethyl-4,5-bis(diphenyl-
phosphino)xanthene [Xantphos]
f. 1,1′-bis(diphenylphosphino)-
ferrocene (dppf)
1
h. Cy2PCH2CH2PCy2
+
+
L′
L′
L-L
L = PR3,
= Py
′= chelate or
H
D
CO(R)CH2Ph
CO(R)CH2Ph
A
D
L
Ir(cod)LL +
′
Ir
Ir
L
D2
O
D
L′
D
D2
R
+
PhCH2COR
L′
B
D
Ir
L
D
L,L
′
= PR3
+
+
+
L′
L′
Ir
D2
L′
Ir
HD
L
DD
L
D
D
Ir
L
O
D
O
D
R
R
Scheme 1. Partial Mechanism for Deuterium Exchange in Phenylacetone or
Ethyl Phenylacetate
to undergo cyclometalation via the agostic complex B: there is evidence for the
formation of such agostic species and of the corresponding metallacycles from
earlier work by Crabtree and co-workers.6 However, the same work also
describes deuterium exchange in both benzo[h]quinoline, which forms a
metallacycle with iridium species, and with 8-methylquinoline where the
iridium complex clearly contained an agostic bond from a methyl hydrogen to
the metal, but cyclometalation did not occur. It is therefore unclear whether
the cyclometalation step is essential to the exchange process in all cases, or
whether exchange may simply follow weakening of a carbon–hydrogen bond
by agostic interaction. It must also be uncertain whether cyclometalation
absolutely has to be preceded by an agostic M–H interaction since, in
principle, the proximity of ortho hydrogens to the metal centre in a species
such as A must facilitate oxidative addition of the C–H bond to some extent.
Shu and Heys5 suggested that because the coordinating centres in 1a were
constrained to lie cis, the corresponding active intermediates were less
congested and better able to accommodate a six-membered metallacycle than
were species where two simple phosphine ligands are trans-disposed. An
improved explanation is that agostic interaction effectively forms a ring in
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2005; 48: 285–294

AN IMPROVED BIDENTATE
287
intermediate B, and the larger seven-membered ring can only be accommo-
dated if the phosphine ligands are constrained by chelate formation. In either
case, this rationale also explains the activity of Crabtree’s catalyst,
Ir(cod)(Py)(PCy3)+.PF6ꢀ (2), as an exchange catalyst for phenylacetone and
acetanilide,3,4 since hydrogenation of 2 gives species in which the tricyclohex-
ylphosphine and pyridine ligands have also been found to lie cis.7
In the course of an initial survey of the activity of simple phosphine
complexes, we found that the monophosphine complex Ir(cod)(MePPh2)+,
where it is assumed that the vacant fourth coordination is occupied either by
adventitious water or the substrate, mediated complete ortho exchange in
acetanilide.2 However, analogues of this complex have much poorer activity,
apparently degrading to form iridium metal within 24 h of exposure to
deuterium. The most electron-rich complex of this series, Ir(cod)(P-t-Bu3)+, is
more active than most toward phenylacetone, but even in this case exchange
proceeds only to 20% of the theoretical maximum. In view of the reaction
kinetics discussed later in this publication, it seems likely that it is the solution
stability of these systems that limits their activity. We also found that the
triphenylarsine complex 3a had an interesting spectrum of activity,2 mediating
some incorporation into phenylacetone whereas the corresponding phosphine
complex 4 did not. Since the bond radius of arsenic is greater than that of
phosphorus,8 this was also expected to result in less congestion at the metal
centre. Our efforts to improve incorporation into phenylacetone, in particular,
have therefore been focused in two areas: complexes with arsine ligands and
with bidentate ligands.
3. (a) L,L′ = AsPh3; (b) L,L′ = As(p-C6H4F)3; (c) L,L′ = As(p-C6H4Cl)3;
+
(d) L,L′ = MeAsPh2 (e) L,L′ = As(p-C6H4OMe)3; (f) L,L′ = AsEt3;
(g) L = AsPh3, L′ = PPh3.
L
-
BF4
Ir
L'
4. L,L′ = PPh3.
Results and discussion
Our initial results using 1a, formed in situ, to mediate deuterium exchange into
phenylacetone were unexceptional. It had been assumed by ourselves and by
previous workers,5 that the kinetics of exchange with complexes 1 would be
comparable to those of complexes such as 2 and 4, where exchange is largely
complete within 24 h.2,4 However, this assumption proves to have been
incorrect. Not only is exchange using 1a comparatively slow (Figure 1), but
exchange into phenylacetone is very much slower than into acetophenone.
This observation prompted us to re-examine the kinetics of exchange with the
dppp complex (1b; Figure 2) and of the bis(triphenylarsine) complex (3a;
Figure 3). In these cases also, exchange is slower than with 4. The differences
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2005; 48: 285–294

288
J. M. HERBERT ET AL.
between exchange processes with these catalysts are equally interesting.
Exchange of acetophenone with 3a follows the pattern observed previously
with 4, in that after an initial induction period, which will correspond to
hydrogenolysis of the cyclooctadiene ligand, exchange is rapid for a short
period. Nevertheless, the process effectively stops within 12 h, possibly as a
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
Acetophenone
Phenylacetone
0.2
0
60
0
80
100
120
20
40
Time, h
Figure 1. Variation of incorporation over time (2:1 substrate/catalyst) for ortho-
deuteration of acetophenone and phenylacetone with complex 1a
2
1.8
1.6
1.4
1.2
1
0.8
Acetophenone
0.6
0.4
0.2
0
Phenylacetone
-8
2
72
12
22
32
52
42
62
Time, h
Figure 2. Variation of incorporation over time (2:1 substrate/catalyst) for ortho-
deuteration of acetophenone and phenylacetone with complex 1b
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2005; 48: 285–294

AN IMPROVED BIDENTATE
289
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Acetophenone
Phenylacetone
20
80
0
100
120
40
60
Time, h
Figure 3. Variation of incorporation over time (2:1 substrate/catalyst) for ortho-
deuteration of acetophenone and phenylacetone with complex 3a
result of catalyst inactivation. With phenylacetone, as already observed, the
exchange is slower and reaches equilibrium after approximately 96 h. On the
other hand, with both 1a and 1b, exchange into acetophenone does not plateau
after the initial period of rapid exchange, but continues at a slower rate until
equilibrium is eventually reached. Unlike acetophenone, initial exchange into
phenylacetone with complexes 1 is very slow, significant deuterium incorpora-
tion being observed only after around 24 h. Even after this period, the catalysts
display different kinetics, with the process mediated by 1a continuing until it
reaches equilibrium after approximately 96 h, while that in the presence of 1b is
faster, but plateaus after only a brief additional time. This latter behaviour
cannot be due to loss of activity on the part of the catalyst, since exchange into
acetophenone continues for some time thereafter. It is unclear why there is
such a long induction period before exchange into phenylacetone is observed,
but the difference between systems where exchange continues for a longer
period, and those where exchange levels out within 24 h, may be one of
complex stability; even where the complex involved is the same, it may be that
different substrates are able to stabilise the active species to different extents.
Nevertheless, based on these observations, it was apparent that an extended
period of exchange is desirable when using complexes 1 or 3, and the results
that follow were all obtained after exposure times of around 120 h.
Exchange using complexes 1
Deuteration of simple substrates using complexes 1 was examined in the
manner described previously,2 but comparing homologous pairs of substrates,
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2005; 48: 285–294

290
J. M. HERBERT ET AL.
so that acetophenone was compared with phenylacetone and N,N-dimethyl-
benzamide with acetanilide and N,N-dimethylphenylacetamide. To ensure
comparability, all experiments were carried out at the same catalyst loading
(50 mole %) and at the same concentration (catalyst concentration 10 mM).
The results are summarised in Table 1.
Assuming that reduced steric demand in complexes 1 is indeed critical to
their ability to mediate isotope exchange via six-membered metallacycles, it is
reasonable to expect that both the size of the ligand and the distance between
the phosphorus centres (reflected in the ligand bite angle9) would be important
factors influencing the activity of a complex. Electronic effects will also be
different to those observed in complexes of simple phosphines; the distortion
caused by chelate formation increases the basicity of the metal centre,10 this
effect increasing as the size of the chelate ring decreases. Since iridium
complexes of the most basic phosphines are of limited value,11 it was expected
that the best results would be obtained using complexes with bis(phosphines)
of moderate basicity. Indeed, the results presented in Table 1 are broadly in
line with this expectation. Taking 1a as a benchmark, the smaller chelate ring
in 1c results in a generally poorer profile (although this may not be an
electronic effect), as does the use of a more basic ligand in complex 1h.
Complexes where the phosphorus centres are less basic are less effective for
exchange in ketones, but the observed exchange in amides is not greatly
affected either by small changes in ligand basicity or in chelate ring size. The
best activity profiles overall are obtained with 1a, and with the BINAP
complex, 1e, which is related to the series of complexes already examined by
Salter and co-workers.12 Increasing the chelate size a little further gives a
complex (1f) of generally modest activity, while the extreme case (1g) is
completely inactive. Nevertheless, the incorporation into both phenylacetone
and N,N-dimethylphenylacetamide is consistently less than 50% of the
theoretical maximum, and the prospect of finding a significantly improved
catalyst within this series does not appear bright.
Table 1. Deuteration Mediated by Complexes 1 (50% loading)a
Complex 1
1a
1b
1c
1d
1e
1f
1g
1h
- ligand bite angle
85
91
72
83
92
96
111
Substrate
Acetophenone
Phenylacetone
N,N-Dimethylbenzamide
N,N-Dimethylphenylacetamide
Acetanilide
1.7
0.6
1.8
0.7
2.0
1.5
0.3
1.6
0.6
1.6
1.1
0.0
0.7
0.6
0.2
1.1
0.3
1.5
1.8
0.6
1.6
0.6
1.7
1.3
0.2
0.9
0.5
1.0
0.0
0.0
0.0
1.2
1.0
1.2
0.7
1.3
1.3
0.0
a Figures are the average number of deuterium atoms incorporated per molecule of substrate, as
determined by mass spectrometry. Exchange was carried out over 120 h in all cases.
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2005; 48: 285–294

AN IMPROVED BIDENTATE
291
Exchange using complexes 3
Exchange reactions with complexes 3 were carried out exactly as described
above for complexes 1, the results being summarised in Table 2. We have
already demonstrated that the effect of the phosphine cone angle in 2 and its
analogues is small so that, in this series, the principal effect of changing the
nature of the arsine ligand should be electronic.11 However, exchange into
phenylacetone in the presence of complexes 3 was not improved by modifying
the ligands, systems other than 3a mediating no more than 25% of the
theoretical maximum level of exchange.
By comparison, deuterium incorporation into dimethylbenzamide and
acetanilide was generally more efficient than into the ketones, and the extent
of exchange into the two amides was comparable using any given catalyst.
Complexes 3 therefore appear better suited for use with amides rather than
with ketones, but still fail to reach the desired level of activity into either class
of substrate. In addition, it appears that longer iridium–arsenic bond is not, in
itself, sufficient to reduce steric demand to a point where formation of the
required six-membered metallacycle is an efficient process. An interesting
comparison can also be made with complex 3g, where replacement of one of
the arsine ligands of 3a by triphenylphosphine results in a complex whose
activity toward acetophenone and N,N-dimethylbenzamide is improved at the
expense of activity toward phenylacetone in particular. In this respect, 3g
behaves more like the phosphine complex 4 than like 3a.
Exchange using complex 5
Given the limitations observed with 1 and 3, a logical next step was the use of a
bidentate arsine complex such as 5. In this case, the enforced cis-relationship
of the coordinating centres, the longer Ir–As bond length, and the increased
effect on complex basicity resulting from chelate formation combine to favour
exchange into the ketone substrates. Consequently, high levels of exchange are
observed into both acetophenone and phenylacetone. At the same time,
exchange into other substrates using 5 remains respectable (Table 2). In
Table 2. Deuteration mediated by Complexes 3 and 5
Complex 1
3a
3b
3c
3d
3e
3f
3gb
5
Substrate
Acetophenone
Phenylacetone
N,N-Dimethylbenzamide
Acetanilide
1.5
1.0
1.8
1.6
1.7
0.5
1.5
1.4
1.0
0.2
1.4
1.7
1.4
0.3
0.6
0.8
1.6
0.3
0.3
0.6
1.0
0.0
1.4
1.4
1.8
0.5
1.6
1.4
1.9
1.7
1.4
1.5
a Figures are the average number of deuterium atoms incorporated per molecule of substrate, as determined
by mass spectrometry; b Using isolated complex.
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2005; 48: 285–294

292
J. M. HERBERT ET AL.
particular, 5 mediates 50% ortho-deuteration (1.0 D) into N,N-dimethylphe-
nylacetamide, this being the highest incorporation we have observed with this
substrate. An additional interesting feature of this system is that exchange into
phenylacetone occurs without the long induction period observed with
complexes 1 (Figure 4), although the process still requires in excess of 96 h
to reach equilibrium. Complex 5 therefore represents an important new lead
for the development of improved promoters of isotope exchange in benzyl
ketones and related substrates.
+
Ph
Ph
As
-
BF4
Ir
As
Ph
Ph
5
In summary, although bidentate complexes 1 are capable of mediating
limited exchange into homologous substrates such as phenylacetone, the
usefulness of complexes 3, with simple arsine ligands, for this purpose, is very
limited. However, complex 5, which combines enforced cis-disposition of the
coordinating centres with a comparatively long metal-arsenic bond, is a
promising new lead for isotopic exchange into a reasonable range of
substrates. The wider applicability of this complex, and the preparation of
analogues, is under investigation.
2
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
Acetophenone
Phenylacetone
0
120
140
160
180
200
60
40
80
100
20
Time, h
Figure 4. Variation of incorporation over time (2:1 substrate/catalyst) for ortho-
deuteration of acetophenone and phenylacetone with complex 5
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2005; 48: 285–294

AN IMPROVED BIDENTATE
293
Experimental
NMR spectra were recorded using a Bruker DRX-500 instrument. GC-MS
conditions and the general exchange procedure were as described previously,2
except that exchange was allowed to proceed for a minimum of 100 h in all
cases, to ensure that equilibrium had been reached. Bis(1,5-cyclooctadiene)-
diiridium(I) dichloride, bis(1,5-cyclooctadiene)iridium tetrafluoroborate (Strem),
phosphines, and ethylene-1,2-bis(diphenylarsine) (Aldrich) were obtained from
commercial suppliers. Phenylacetone4 and (1,5-cyclooctadiene)(triphenylarsi-
ne)(triphenylphosphine)iridium(I) hexafluorophosphate (3g)8 were prepared as
described previously.
Triarylarsines were prepared by the dropwise addition of arsenic(III)
chloride to 10 molar equivalents of the appropriate Grignard reagent under
nitrogen at ꢀ108C.13 Thus obtained were tris(4-fluorophenyl)arsine (90%),
m.p. 73–73.68C (lit.14 m.p. 748C), dH(CD3SOCD3) 7.23 (6H, dd, J ¼ 11:1,
10.2 Hz), 7.29 (6H, J ¼ 10:2, 7.2Hz), m/z 360 (M+ꢁ ), 170 (100%, 4-FC6H4As+ꢁ );
tris(4-chlorophenyl)arsine (100%), m.p. 103–1068C (lit.15 m.p. 104–1078C),
dH(CDCl3) 7.19 (6H, d, J ¼ 8:2 Hz), 7.30 (6H, d, J ¼ 8:2 Hz), m/z 408, 410,
412 (M+ ꢁ ), 186, 188 (100%, 2:1, 4-ClC6H4As+ ꢁ ); tris(4-methoxyphenyl)arsine
(100%), m.p. 155–1588C (lit.16 m.p. 156–1588C), dH(CD3SOCD3) 3.73 (9H, s),
6.92 (6H, d, J ¼ 8:1 Hz), 7.15 (6H, d, J ¼ 8:1 Hz), m/z 396 (M+ ꢁ ), 182 (100%,
4-MeOC6H4As+ ꢁ ). Diphenylmethylarsine was prepared as described by Doak
and Freedman.17 dH(CD2Cl2) 1.48 (3H, s), 7.29–7.34 (6H, m), 7.42 (4H, dd,
J ¼ 7:5, 2.0 Hz), m/z 244 (M+ ꢁ ), 154 (100%, PhAsH2+ ꢁ ).
Representative exchange procedure
A solution of bis(1,5-cyclooctadiene)iridium tetrafluoroborate and ligand
(1 mol. equiv.) in DCM was stirred for 5 min under nitrogen, then diluted with
DCM, to give a 5 mmol/ml solution of 1 or 5. This solution (2 ml) was added to
the substrate (20 mmol) in a 10 ml flask; the system was degassed and flushed
with deuterium (corresponding to at least 20 molar equivalents, or 10 molar
equivalents per exchange site), then sealed and stirred for 120 h. After this time
most of the volatiles were evaporated and the crude product was extracted
with ethyl acetate. Exchange runs using complexes 3 or 4 were carried out
using the procedure described previously for generation of the catalyst from
bis(1,5-cyclooctadiene)diiridium(I) dichloride.2
References
1. Heys JR, Shu AYL, Senderoff SG, Phillips NM. J Label Compd Radiopharm
1993; 33: 431; Chen W, Garnes KT, Levinson SH, Saunders D, Senderoff SG, Shu
AYL, Villani AJ, Heys JR. J Label Compd Radiopharm 1997; 39: 291.
2. Ellames GJ, Gibson JS, Herbert JM, McNeill AH, Cross PWC, Kerr WJ,
Mathers TW, Trewartha G. Tetrahedron 2003; 59: 3349.
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2005; 48: 285–294

294
J. M. HERBERT ET AL.
3. Hesk D, Das PR, Evans B. J Label Compd Radiopharm 1995; 36: 497.
4. Ellames GJ, Gibson JS, Herbert JM, McNeill AH. Tetrahedron 2001; 57: 9487.
5. Shu AYL, Heys JR. J Organomet Chem 1996; 524: 87.
6. Crabtree RH, Holt EM, Lavin M, Morehouse SM. Inorg Chem 1985; 24: 1986;
Crabtree RH, Lavin M, Bonneviot L. J Am Chem Soc 1986; 108: 4032; Lavin M,
Holt EM, Crabtree RH. Organometallics 1989; 8: 99.
7. Crabtree RH, Felkin H, Fillebeen-Khan T, Morris GE. J Organomet Chem 1979;
168: 183.
8. Bedford RB, Chaloner PA, Dewa SZ, Lopez G, Hitchcock PB, Momblona F,
Serrano JL. J Organomet Chem 1997; 527: 75.
9. Van Leeuwen PWNM, Kamer PCJ, Reek JNH, Dierkes P. Chem Rev 2000; 100:
2741.
10. Angelici RJ. Acc Chem Res 1995; 28: 51.
11. Ellames GJ, Gibson JS, Herbert JM, Kerr WJ, McNeill AH. J Label Compd
Radiopharm 2004; 47: 1.
12. Salter R, Chappelle M, Morgan A, Ackermann P, Moenius T, Studer M, Spindler
F, Togni A. In Synthesis and Applications of Isotopically Labelled Compounds, vol.
7, Pleiss U, Voges R (eds). Wiley: New York, 2001; 63; Salter R, Moenius T,
Ackermann P, Studer M, Morgan A, Chappelle M. J Label Compd Radiopharm
2001; 44: 956.
13. McCartney BA, Jacobsen BM, Vreeke M, Lewis ES. J Am Chem Soc 1990; 112:
3554.
14. De Ketelaere RF, Delbeke FT, Van Der Kelen GP. J Organomet Chem 1971; 28:
217.
15. Hellwinkel D, Kilthau G. Chem Ber 1968; 101: 121.
16. McEwen WE, Fountaine JE, Schulz DN, Shiau WI. J Org Chem 1976; 41: 1684.
17. Doak GO, Freedman LD. Synthesis 1974; 328.
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2005; 48: 285–294