One-pot carbon monoxide-free hydroformylation of internal olefins to terminal aldehydes

Article abstract of DOI:10.1002/adsc.200404228

A one-step hydroformylation of mixtures of internal and terminal olefins yielding terminal aldehydes without the need for carbon monoxide has been developed. Treatment of the olefin mixtures with Rh catalysts and pinacolborane leads to isomerization and hydroboration in one step. Homologation and subsequent oxidation regiospecifically afford the terminal aldehyde. Good overall yields are obtained for all substrates examined.

Full text of DOI:10.1002/adsc.200404228

COMMUNICATIONS
One-Pot Carbon Monoxide-Free Hydroformylation of Internal
Olefins to Terminal Aldehydes
David R. Edwards, Cathleen M. Crudden,* Katherine Yam
Department of Chemistry, Queenꢀs University, 90 Bader Lane, Kingston, ON K7L 3N6, Canada
Fax: (þ1)-613-533-6669, e-mail: cruddenc@chem.queensu.ca
Received: July 14, 2004; Accepted: November 5, 2004
Abstract: A one-step hydroformylation of mixtures
of internal and terminal olefins yielding terminal al-
dehydes without the need for carbon monoxide has
been developed. Treatment of the olefin mixtures
with Rh catalysts and pinacolborane leads to isomer-
ization and hydroboration in one step. Homologa-
tion and subsequent oxidation regiospecifically af-
ford the terminal aldehyde. Good overall yields are
obtained for all substrates examined.
internal olefins are converted into terminal aldehydes.
In fact, after some optimization of the individual steps,
we have been able to accomplish this transformation
in high yield, as illustrated in Scheme 1.
Keywords: aldehydes; boronates; homologation; hy-
droboration; hydroformylation
Scheme 1.
Since the hydroboration was reported to proceed
smoothly with dichloromethane as solvent, we set out
to develop a one-pot procedure with the solvent of the
The hydroformylation of olefins is the oldest and largest first transformation functioning as the reagent in the ho-
chemical process currently in use employing homoge- mologation step. After homologation, oxidation would
[
1]
nous transition metal catalysts. A major limitation of give the product of hydroformylation without the need
this process is the requirement for terminal olefin feed- for carbon monoxide. The homologation has precedent
[
8]
stocks in order to access the terminal aldehyde, which is in the work of Matteson on boronic esters in general
[
2]
desired for most applications. Innovations geared to- and by us specifically with the homologation of substi-
wards the use of economically more feasible olefin mix- tuted phen-1-ethylpinacol boronate during the asym-
[6]
tures, such as raffinate II containing a mixture of butene metric synthesis of Ibuprofen and Naproxen. Thus,
isomers, are highly desirable. Cobalt-based catalytic sys- we report herein a one-pot carbon monoxide-free hy-
tems have the dual problem of high temperature and droformylation procedure for mixtures of olefins, which
pressure requirements as well as low regioselectivity to- produces linear aldehydes in high yields.
[
2]
wards terminal aldehydes. Greater advances have
Our initial efforts focused on hydroborating 1-octene
been made with rhodium-based catalysts. For example, with Wilkinsonꢀs catalyst according to the reported pro-
[
5]
van Leeuwen et al. have described a series of diphos- cedure of Srebnik et al. [Eq. (1)]. We found the reac-
phine ligands that catalyze the isomerization/hydrofor- tion to be quite sluggish. Even after 120 minutes, conver-
mylation of internal octenes to the linear aldehyde in ex- sion to 1 was only 6%. Freshly prepared and recrystal-
[
3]
[9]
cellent yields. Similarly, Beller et al. have reported an lized catalyst gave identical results. Catalyst that had
Rh/NAPHOS system that affects the same transforma- been stored outside the glove box and showed signs of
3
1
tion as well as the related hydroaminomethylation reac- phosphine oxide in its P NMR gave significantly better
tion. Both these synthetic processes share a common results. In fact, GC yields of up to 92% could be obtained
limitation of decreased activity and selectivity when in- when a stream of air was passed through a solution of
[4]
ternal olefins, such as 4-octene, are employed.
RhCl(PPh ) in dichloromethane. Since this method
3 3
For this reason, we were intrigued by a report of Sreb- gave variable yields, a more predictable procedure was
nik et al. in which 1-octene and trans-4-octene were both sought.
hydroborated with complete regioselectivity to the line-
[
5]
ar boronate 1 using pinacolborane (HBPin). Com-
bined with our recently reported homologation chemis-
[6]
try, we felt this would provide an interesting possibility
of performing a CO-free hydroformylation in which
[
7]
ð1Þ
50
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/adsc.200404228
Adv. Synth. Catal. 2005, 347, 50–54

Hydroformylation of Internal Olefins to Terminal Aldehydes
COMMUNICATIONS
[a]
Table 1. Catalyst screening for hydroboration/isomerization of 1-octene.
[
b]
[c]
Entry
Catalyst
Additive (Phosphine:Rh)
Yield [%]
Time [h]
1
2
3
4
5
6
7
8
9
RhCl(PPh3)3
RhCl(PPh3)3
[RhCl(PPh ) ]
2 2
–
–
–
6
2
2
3
0.5
0.5
0.5
0.5
0.5
0.5
0.5
[d]
31–92
[e]
88
71
98
83
79
28
83
77
3
[Rh(m-Cl)(C H ) ]
PPh (1: 1)
PPh (1.25: 1)
PPh (1.5: 1)
PPh (2: 1)
DPPB (1.25: 1)
P(OPh) (1.25: 1)
PPh (1.25: 1)
2
4
2
2
2
2
2
2
2
3
[Rh(m-Cl)(C H ) ]
2
4
2
3
[Rh(m-Cl)(C H ) ]
2
4
2
3
[Rh(m-Cl)(C H ) ]
2
4
2
3
[
f]
[Rh(m-Cl)(C H ) ]
2
4 2
[Rh(m-Cl)(C H ) ]
2
4
2
3
1
0
[Rh(COD) ] BF
2
2
4
3
[
a]
Reactions were run on a 1 mmol scale, 1.0 M in dichloromethane under nitrogen with 1% catalyst in entries 1–3 and 10, and
2% catalyst for entries 4–9.
b]
[
[
[
[
[
Molar ratio of phosphine to Rh (in the monomeric form for entries 4–9).
GC yield using decane as internal standard.
c]
d]
e]
f]
Air was bubbled through the reaction mixture until the color changed from pale yellow to black.
Isolated yield.
DPPB¼1,4-bis(diphenylphosphino)butane.
Table 2. Optimization of homologation reaction conditions.
[a]
Entry
Base (equivs.)
Temperature [ 8C]
Additive
Conversion [%]
1
2
3
4
5
n-BuLi (2)
n-BuLi (3)
n-BuLi (3)
n-BuLi (3)
ꢀ100
ꢀ100
ꢀ100
ꢀ78
ZnCl2
ZnCl2
–
–
–
82
89
89
88
100
[
b]
LDA (2)
ꢀ40
[
[
a]
b]
1
Conversion was determined by H NMR assay of the ratio of starting material to a-chloropinacol boronate.
LDA¼lithium diisopropylamide.
[
RhCl(PPh ) ] was examined as a catalyst precursor be- action flask was then cooled to the desired temperature
3 2 2
cause of its latent free coordination site and decreased and base was added slowly according to the Matteson
[
10]
phosphine to Rh ratio compared to Wilkinsonꢀs catalyst. protocol.
The resulting homologated product was
Gratifyingly, hydroboration of 1-octene proceeded with then worked up and examined for conversion of 1 to 2
1
a 90% conversion to 1 after 3 hours reaction time. In an ef- by H NMR, see Table 2.
fort to further optimize the Rh:phosphine ratio, [RhCl-
(
C H ) ] /PPh combinations were examined (Table 1). It
2 2 2 2 3
was determined that a ratio of 1.25 phosphine to 1 Rh
was optimal. Use of 1–2 mole percent of catalyst precursor
resulted in near quantitative isolated and GC yields of lin-
ear boronate 1. Attempts to decrease catalyst loading be-
low this resulted in observation of the isomerization prod-
uct 2-octene. Triphenyl phosphite and DPPB were also ex-
ð2Þ
amined in the reaction. Neither of these phosphines nor Matteson has demonstrated that the addition of ZnCl2
even the use of a cationic Rh species gave improved results aids in the breakdown of the borate complex and per-
vs. the Rh dimer and PPh . Results are summarized in Ta- mits the reaction to take place with high diastereoselec-
3
ble 1.
tivity when the boronate ester is derived from a chiral
[11]
Having established a suitable protocol for the regiose- diol. Since our reaction employs an achiral boronate
lective and nearly quantitative conversion of 1-octene to ester and is not diastereoselective, ZnCl was unnecessa-
2
the corresponding linear pinacol boronates, our atten- ry. Entries two and three reveal that omission of the zinc
tion turned to the task of converting 1 to the a-chloropi- additive made no difference to the reaction. The slightly
nacol boronate 2 [Eq. (2)]. Invariably, the crude reac- milder temperature of ꢀ788C could also be employed
tion mixture resulting from the first transformation using this in situ homologation procedure, entry 4.
was diluted to a 0.1 M concentration in dry THF. The re- LDA (lithium diisopropylamide) was able to affect
Adv. Synth. Catal. 2005, 347, 50–54
asc.wiley-vch.de
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
51

COMMUNICATIONS
David R. Edwards et al.
Table 3. Optimization of oxidation reaction conditions.
[a]
Entry
Oxidant (equivs.)
Conditions
Overall Yield [%]
[
b]
1
2
3
4
5
6
7
TMANO (3.0)
PhH, 608C, 9 h
PhH, 608C, 24 h
PhH, 608C, 9 h
74–76
0
[c]
NMO (3.0)
Me N(O)C H (3.0)
NaBO (1.2)
H O /NaOH (1.2)
Na CO /1.5 H O (1.2)
Na CO /1.5 H O (2.0)
0
42
<20
65
80
2
11 24
[d]
H O , 258C, 2 h
3
2
[
d]
H O , 258C, 2 h
2
2
2
[d]
H O , 258C, 2 h
2
[d]
2
3
2
2
H O , 258C, 2 h
2
3
2
2
2
[
[
[
[
a]
b]
c]
d]
GC yield over all three steps using decane as internal standard.
Trimethylamine N-oxide.
N-Methylmorpholine N-oxide.
Crude reaction mixture was further diluted to a 0.2 M concentration in water.
this transformation with quantitative conversion of 1 to Having identified a suitable protocol for the conversion
2
at the more desirable temperature of ꢀ408C, entry 5. of 1-octene, a linear olefin, to nonyl aldehyde, it re-
Nonetheless, attempts to oxidize 2 in the presence of the mained to be shown that the same procedure could be
resulting two equivalents of diisopropylamine failed. applied to mixtures of octene isomers. This requires
Carrying out a simple extractive work-up procedure to that the olefin isomerization be faster than hydrobora-
[
16]
remove the amine allowed the oxidation to proceed, tion.
A stock solution consisting of equimolar
however, this did not fit with our initial goal of develop- amounts of 1-, cis-2-, trans-2-, and trans-4-octene was
ing a one-pot procedure so the procedure described in subjected the one-pot hydroboration/homologation/ox-
entry 4 was adopted.
idation reaction [Eq. (4)]. In order to further optimize
The crude reaction mixture containing 2 was treated the procedure, we carried out the hydroboration with
withvariousoxidizingreagents tocompletethesequence. a slight increase in dichloromethane (0.5 M instead of
Reactions were monitored by GC and the overall yield of 1.0 M in dichloromethane). This simple switch allowed
aldehyde from 1-octene was calculated. Amine oxides us to decrease the amount of basein the following homo-
have previously been employed for oxidation in organo- logation step, and did not adversely affect the hydrobo-
borane chemistry and we tested three such reagents, tri- ration reaction itself. Presumably the addition of extra
methylamine N-oxide (TMANO), N-methylmorpholine dichloromethane was sufficient to drive the homologa-
N-oxide (NMO), N,N-dimethylundecylamine N-oxide, tion reaction to high conversion without the require-
[
12]
entries 1, 2, 3 of Table 3. Unfortunately none of these ment for three equivalents of base. Employing the new
reagents was soluble in THF. TMANO could only be em- protocol, the overall yield of the one-pot reaction was
ployed following a solvent switch, in which THF was par- 85–86% [Eq. (4)].
tially removed and benzene subsequently added to the
crude reaction flask. Reflux for 9 h furnished nonyl alde-
hyde in an overall yield of 74–76% over the three steps.
Other amine oxides tested showed no conversion to the
aldehyde. Since the solvent switch is cumbersome, we ex-
amined other oxidants that would be soluble in THF/
[
13]
CH Cl . Neither Kabalkaꢀs perborate nor basic hydro-
2
2
[14]
ð4Þ
peroxide led to sufficient amounts of aldehyde, entries
and 5. However, we were delighted to see that sodium
percarbonate, a very inexpensive oxidant, cleanly fur- To further probe the scope of the reaction and demon-
4
[
15]
nished nonyl aldehyde in good overall yield, entry 6.
strate its potential as a viable option for the production
The oxidation could be carried out without switching sol- of aldehydes from olefin mixtures, a series of internal
vents, requiring only the addition of water as a co-solvent. olefins of varying carbon length were subjected to the
The reaction proceeded faster and with better conversion optimized protocol, Table 4. In all substrates examined,
using two equivalents of the reagent, entry 7.
good overall yields of thelinear aldehyde were obtained.
Most gratifying were the results obtained for 2-hexene,
in which an excellent yield of 90–95% was observed
over the three steps.
In conclusion, a one-pot protocol for the conversion of
olefins to linear aldehydes without the need for carbon
monoxide has been developed. A highlight of the de-
ð3Þ
52
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 50–54

Hydroformylation of Internal Olefins to Terminal Aldehydes
COMMUNICATIONS
Table 4. Hydroformylation of various internal olefins.
then 1-octene (1.02 mmol, 160 mL) was added by syringe fol-
lowed immediately by pinacolborane (1.2 mmol, 175 mL).
The clear reddish-yellow solution went dark within five mi-
nutes and reactions were left in the glove box for a further 30
minutes. The reaction flask was then removed from the glove
box, the mixture was diluted with 10 mL dry THF and cooled
to ꢀ788C. Freshly titrated n-BuLi (2 equivs.) was then added
slowly down the side of the flask. Reactions were then left to
slowly warm to room temperature overnight. The following
morning the reaction mixture was further diluted with 5 mL
distilled water and 2 equivalents of sodium percarbonate
(2.04 mmol, 320 mg) were added at 08C and the reaction was
left to warm to room temperature over 2 hours. Decane was
added as internal standard and a small aliquot removed for
yield determination by GC.
[a]
Entry
Olefin
Yield of Aldehyde [%]
1
2
3
4
2-hexene
2-heptene
octenes
3-nonene
90–95
79–83
85–86
60–64
[
b]
[
a]
GC yield using decane as internal standard except for en-
try 2, where octane was used.
Equimolar mixture of 1-octene, 2-octene and trans-4-oc-
tene.
[
b]
vised methodology is the ability to convert mixtures of
olefins to the more desirable linear aldehyde in up to
9
5% yield. This represents a significant step towards
Acknowledgements
the development of hydroformylation procedures capa-
ble of functioning on cheaper feedstocks comprised of
internal as well as terminal carbon-carbon double
bonds. Although our method does require cryogenic
temperatures for the homologation step and one full
equivalent of hydroborating reagent, it provides a signif-
icantly safer and easier method to carry out hydroformy-
lations. Thus regular synthetic laboratories not equip-
ped to handle CO can use our method to affect hydrofor-
mylations without the need for high pressure vessels and
special safety equipment. The reaction also works best
with simple, easily handled catalysts such as [Rh(m-
Cl)(C H ) ] and triphenylphosphine.
We acknowledge the Natural Sciences and Engineering Re-
search Council of Canada (NSERC) for funding of this research
in terms of operating grants to CMC and an undergraduate re-
search fellowship to KY. We thank the Walter Sumner fund for
a fellowship to DRE.
References and Notes
[
1] G. O. Spessard, G. L. Miessler, Organometallic Chemis-
try, John Prentice Hall, New Jersey, 1996, p. 255.
2
4 2 2
[
2] S. Bhaduri, D Mukesh, Homogeneous Catalysis Mecha-
nisms and Industrial Applications, Wiley-Interscience,
New York, 2000, p. 1; A. M. Trzeciak, J. J. Ziolkowski,
Coord. Chem. Rev. 1999, 190–192, 883.
Experimental Section
[
3] L. A. van der Veen, P. C. J. Kramer, P. W. N. M. van
Leeuwen, Angew. Chem. Int. Ed. 1999, 38, 336.
General Remarks
[
4] M. Beller, B. Zimmermann, H. Geissler, Chem. Eur. J.
1
999, 5, 1301; R. Jackstell, H. Klein, M. Beller, K. Wiese,
[
RhCl(C H ) ] was prepared from commercial rhodium chlor-
2 2 2 2
[17]
D. Rottger, Eur. J. Org. Chem. 2001, 3871; H. Klein, R.
Jackstell, K. Wiese, C. Borgmann, M. Beller, Angew.
Chem. Int. Ed. 2001, 40, 3408; A. Seayad, M. Ahmed,
H. Klein, R. Jackstell, T. Gross, M. Beller, Science
2002, 297, 1676.
ide trihydrate by the reported procedure. n-BuLi was freshly
titrated against N-benzylbenzamide to the deep blue endpoint.
All other reagents were commercially available and were puri-
[18]
fied according to Perrin and Perrin techniques prior to use.
Hydroborations took place in the oxygen-free environment
of a glove box. Yields of all aldehydes were obtained via an
Agilent 6850 GC referenced to a commercially available au-
thentic sample. Calibration curves were constructed by run-
ning a series of samples containing the aldehyde and internal
standard, decane, at four different concentrations. All attempts
to obtain isolated yields of 2 led to partial decomposition of the
alpha-chloropinacol boronates. Conversion of 1 to 2 was calcu-
[5] S. Pereira, M. Srebnik, J. Am. Chem. Soc. 1996, 118, 909.
[6] A. C. Chen, L. Ren, C. M. Crudden, J. Org. Chem. 1999,
6
4, 9704; C. M. Crudden, Y. B. Hleba, A. C. Chen, J. Am.
Chem. Soc. 2004, 126, 9200; A. C. Chen, L. Ren, C. M.
Crudden, Chem. Commun. 1999, 611.
7] For an excellent review of carbonylations that can be car-
ried out without carbon monoxide see: T. Morimoto, K.
Kakiuchi, Angew. Chem. Int. Ed. 2004, 43, 5580.
8] D. S. Matteson, D. J. Majumdar, Tetrahedron. 1998, 54,
[
1
lated based on crude H NMR integration ratios for proton sig-
nals alpha to boron.
[
[
1
0555.
9] D. A. Evans, G. C. Fu, B. A. Anderson, J. Am. Chem.
Soc. 1992, 114, 6679.
Nonyl Aldehyde
[
10] D. S. Matteson, D. J. Majumdar, Organometallics 1983, 2,
529.
[
RhCl(C H ) ] [0.01 mmoles, 3.9 mg, 2% (per Rh)] was weigh-
2 4 2 2
1
ed into a 50-mL two-necked round-bottomed flask inside a
glove box. To this PPh (0.025 mmoles, 6.6 mg, 2.5%) was added
followed by 2 mL of deoxygenated and dry dichloromethane.
The reddish-yellow solution was stirred for five minutes and
[11] D. S. Matteson, D. J. Majumdar, J. Am. Chem. Soc. 1980,
102, 7588.
[12] J. Soderquist, M. R. Najafi, J. Org. Chem. 1986, 51, 1330.
3
Adv. Synth. Catal. 2005, 347, 50–54
asc.wiley-vch.de
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
53

COMMUNICATIONS
David R. Edwards et al.
[
[
[
[
13] G. Kabalka, T. Shoup, N. Goudgaon, J. Org. Chem. 1989,
Also note that isomerization can take place thermally
in certain cases, for example, during the synthesis of 9-
BBN by hydroboration of 1,5-cyclooctadiene: E. F.
Knights, H. C. Brown, J. Am. Chem. Soc. 1968, 90, 5280.
[17] Inorganic Syntheses, Vol. 28, p. 86.
5
4, 5930.
14] G. W. Kabalka, H. C. Hedgecock, J. Org. Chem. 1975, 40,
776.
15] G. Kabalka, P. Wadgaonkar, T. Shoup, Organometallics
990, 9, 1316.
16] For isomerizations with BH and RhCl , see: T. C. Mor-
1
1
[18] W. L. F. Armarego, D. P. Perrin, Purification of Labora-
tory Chemicals, Butterworth-Heinemann, Oxford, 2000.
3
3
rill, C. A. DꢀSouza, Organometallics 2003, 22, 1626.
54
ꢁ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 50–54