Thermolyses of O-Phenyl Oxime Ethers. A New Source of Iminyl Radicals and a New Source of Aryloxyl Radicals

Article abstract of DOI:10.1021/jo049927y

Six O-phenyl ketoxime ethers, RR′C=NOPh 1-6, with RR′= diaryl, dialkyl, and arylalkyl, together with N-phenoxybenzimidic acid phenyl ether, PhO(Ph)C=NOPh, 7, have been shown to thermolyze at moderate temperatures with "clean" N-O bond homolyses to yield iminyl and phenoxyl radicals, RR′C=N^. and PhO^.. Since 1-6 can be synthesized at room temperature, these compounds provide a new and potentially useful source of iminyls for syntheses. The iminyl from 7 undergoes a competition between β-scission, to PhCN and PhO^., and cyclization to an oxazole. Rate constants, 10⁶ k/s^-1, at 90 °C for 1-6 range from 4.2 (RR′ = 9-fluorenyl) to 180 (RR′ = 9-bicyclononanyl), and that for 7 is 0.61. The estimated activation enthalpies for N-O bond scission are in satisfactory agreement with the results of DFT calculations of N-O bond dissociation enthalpies, BDEs, and represent the first thermochemical data for any reaction yielding iminyl radicals. The small range in k (N-O homolyses) is consistent with the known σ structure of these radicals, and the variations in k and N-O BDEs with changes in RR′ are rationalized in terms of iminyl radical stabilization by hyperconjugation: RR′C=N ^. ? R^.R′C≡N. Calculated N-H BDEs in the corresponding RR′C= NH are also presented.

Full text of DOI:10.1021/jo049927y

Th er m olyses of O-P h en yl Oxim e Eth er s. A New Sou r ce of Im in yl
Ra d ica ls a n d a New Sou r ce of Ar yloxyl Ra d ica ls
J essie A. Blake,† Derek A. Pratt,‡ Shuqiong Lin,† J ohn C. Walton,§ Peter Mulder,*,| and
K. U. Ingold†
National Research Council, Ottawa, Ontario K1A 0R6, Canada, University of Illinois at
Urbana-Champaign, Urbana, Illinois 61801, University of St. Andrews, St. Andrews KY16 9ST, U.K.,
and Leiden Institute of Chemistry, Leiden University, 2300 RA Leiden, The Netherlands
p.mulder@chem.leidenuniv.nl
Received J anuary 12, 2004
Six O-phenyl ketoxime ethers, RR′CdNOPh 1-6, with RR′ ) diaryl, dialkyl, and arylalkyl, together
with N-phenoxybenzimidic acid phenyl ether, PhO(Ph)CdNOPh, 7, have been shown to thermolyze
at moderate temperatures with “clean” N-O bond homolyses to yield iminyl and phenoxyl radicals,
RR′CdN• and PhO•. Since 1-6 can be synthesized at room temperature, these compounds provide
a new and potentially useful source of iminyls for syntheses. The iminyl from 7 undergoes a
competition between â-scission, to PhCN and PhO•, and cyclization to an oxazole. Rate constants,
106 k/s-1, at 90 °C for 1-6 range from 4.2 (RR′ ) 9-fluorenyl) to 180 (RR′ ) 9-bicyclononanyl), and
that for 7 is 0.61. The estimated activation enthalpies for N-O bond scission are in satisfactory
agreement with the results of DFT calculations of N-O bond dissociation enthalpies, BDEs, and
represent the first thermochemical data for any reaction yielding iminyl radicals. The small range
in k (N-O homolyses) is consistent with the known σ structure of these radicals, and the variations
in k and N-O BDEs with changes in RR′ are rationalized in terms of iminyl radical stabilization
by hyperconjugation: RR′CdN• T R•R′CtN. Calculated N-H BDEs in the corresponding RR′Cd
NH are also presented.
Kinetic studies on the thermal decomposition of a
number of O-benzyl oxime ethers in hydrogen donor
solvents, XH (such as tetralin and 9,10-dihydroan-
thracene), at 150 °C have shown that the dominant
process is a reverse radical disproportionation (RRD)
followed by â-scission of the N-O bond, reaction 1.1
class of free-radical initiators as well as a source of iminyl
radicals which have become of considerable synthetic
interest in recent years.3 Herein, we report on the
kinetics and products of the thermolyses of seven O-
phenyl oxime ethers in tert-butylbenzene mainly at 90
°C but with some work done at temperatures ranging
from 60 to 150 °C. RRD is demonstrated to be unimpor-
tant, and the activation enthalpies, Ea, for decomposition
are compared with DFT-calculated N-O bond dissocia-
tion enthalpies, BDEs.
XH + RR′CdNOCH2Ph f X• +
RR′C•NHOCH2Ph f RR′CdNH + •OCH2Ph (1)
In non-hydrogen-donor solvents, e.g., tert-butylbenzene,
and at higher temperatures (170 °C), these compounds
undergo a relatively “clean” homolysis of their O-C bond,
reaction 2.2
Resu lts
The compounds synthesized and studied in the present
work are shown in Chart 1. The synthesis of 1-6 involved
only a simple condensation of the corresponding ketone
with commercially available O-phenylhydroxylamine (hy-
drochloride) in anhydrous pyridine at room temperature,
reaction 4.
RR′CdNOCH2Ph f RR′CdNO• + •CH2Ph (2)
During these studies, it occurred to us that it might
be possible to achieve a clean homolysis of the N-O bond
at much lower temperatures by using O-phenyl oxime
ethers, reaction 3.
RR′CdO + H2NOPh - H O8RR′CdNOPh
(4)
2
Compound 7 was prepared from benzohydroxamic acid
in a four-step process based on its original synthesis by
RR′CdNOPh f RR′CdN• + •OPh
(3)
Reaction 3 would then provide a new, convenient, and
relatively low-temperature, nonazo and nonperoxide,
(3) See, e.g.: Boivin, J .; Schiano, A.-M.; Zard, S. Z. Tetrahedron Lett.
1994, 35, 249-252. Boivin, J .; Fouquet, E.; Zard, S. Z. Tetrahedron
1994, 50, 1745-1746. Zard, S. Z. Synlett 1996, 1148-1154. Uchiyama,
K.; Hayashi, Y.; Narasaka, K. Chem. Lett. 1998, 1261-1262. Calestani,
G.; Leardini, R.; McNab, H.; Nanni, D.; Zanardi, G. J . Chem. Soc.,
Perkin Trans. 1 1998, 1813-1824. Gagosz, F.; Zard, S. Z. Synlett 1999,
1978-1980. Guidon, Y.; Gue´rin, B.; Landry, S. R. Org. Lett. 2001, 3,
2293-2296. Mikami, T.; Narasaka, K. C. R. Acad. Sci. Paris, Chim./
Chem. 2001, 4, 477-486. Leardini, R.; McNab, H.; Minozzi, M.; Nanni,
D. J . Chem. Soc., Perkin Trans. 1 2001, 1072-1078. Gennet, D.; Zard,
S. Z.; Zhang, H. Chem. Commun. 2003, 1870-1871.
† National Research Council.
‡ University of Illinois at Urbana-Champaign.
§ University of St. Andrews.
| Leiden University.
(1) Blake, J . A.; Ingold, K. U.; Lin, S.; Mulder, P.; Pratt, D. A.;
Sheeler, B.; Walton, J . C. Org. Biomol. Chem. 2004, 2, 415-420.
(2) Pratt, D. A.; Blake, J . A.; Mulder, P.; Walton, J . C.; Korth, H.-
G.; Ingold, K. U. Manuscript in preparation.
10.1021/jo049927y CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/06/2004
3112
J . Org. Chem. 2004, 69, 3112-3120

Thermolyses of O-Phenyl Oxime Ethers
TABLE 1. Decom p osition of P h 2CdNOP h (1) a n d
F ldNOP h (2) in ter t-Bu tylben zen e a t 90 °C for Va r iou s
Tim e P er iod sa ,b
CHART 1
time (10-3 s)
1
Ph2CNH Ph2CO Ph2CNH + Ph2CO PhOH
54.0
86.4
97.2
122.0
172.8
244.8
50.7
33.8
31.4
21.4
11.1
4.6
66.3
68.4
72.4
63.0
65.0
66.1
5.1
5.0
6.0
9.3
7.7
71.4
73.4
78.4
72.3
72.7
77.6
28.5
31.1
34.8
31.2
32.4
35.7
11.5
time (10-3 s)
2
FlNH
FlO
FlNH + FlO
PhOH
68.4
118.8
176.4
230.4
248.4
72.5
57.9
47.1
39.1
35.1
83.0
76.2
80.0
81.9
78.3
5.6
3.1
3.7
4.6
4.2
88.6
79.3
83.7
86.5
82.5
20.7
17.1
18.1
19.8
18.3
Taylor and Kienzle4 but modified to avoid the use of a
(highly toxic) thallium salt; see the Experimental Section.
Th er m olyses of O-P h en yl Oxim e Eth er s. The oxime
ethers (8-13 mg) were dissolved in 0.5 mL of tert-
butylbenzene (tBB) in small Pyrex tubes. The solutions
were degassed via three or more freeze-pump-thaw
cycles on a vacuum line, and the tubes were flame-sealed
under vacuum. The tubes were then heated in a constant
temperature ((1 °C) oven at chosen temperatures for
various lengths of time. Following thermolyses, the tubes
were cooled to room temperature and opened, and the
contents were added to 3 mL of acetonitrile. The solutions
were subsequently filtered, when necessary, and ana-
lyzed by HPLC (see the Supporting Information).
Dia r yl Eth er s 1 a n d 2. The products of the thermoly-
ses of these two Ar2CdNOPh compounds in tBB were the
corresponding imines, Ar2CdNH, the corresponding ke-
tones, Ar2CdO (formed presumably by hydrolysis of the
imines by trace water in the acetonitrile), and phenol.
The identities of the imines and ketones were determined
by comparison of their HPLC retention times and UV
spectra with authentic materials, and their identities
were further confirmed by GC/MS on samples where all
the starting material had decomposed. Quantification of
the loss of the starting oxime ether and formation of
products was based on calibration of the HPLC’s detector
with the authentic compounds.
Product yields from the thermolyses of 1 and 2 in tBB
at 90 °C as a percentage of the quantity of oxime ether
decomposed after six and five, respectively, different time
intervals are given in Table 1 (for additional details, see
Table S1 in the Supporting Information). The major
product is always the diarylimine, and the yields of imine
plus ketone are fairly good ranging from 71% to 89%
(Table 1). The yields of phenol are always poor, ranging
from 17% to 35%.
Since tBB is not a hydrogen atom donor solvent, the
obvious question is: What is the H-atom donor that
converts iminyl and phenoxyl radicals to imine and
phenol? We suggest that this hydrogen donor is formed
in situ during the reaction via the dimerization and
subsequent oligomerization of the phenoxyl radicals. The
initial dimerization is known to involve both C-C and
C-O coupling at the ortho and para positions and to yield
much better H-atom donors than phenol itself.5 The
expected initial dimers arising from a C-C ortho and a
C-O ortho coupling are shown in Scheme 1. Initial
dimerization will be followed by oligomerization and the
a
Remaining Ar2CdNOPh is given as a percentage of the initial
percentage of decomposed
concentration and products as
a
b
Ar2CdNOPh. For additional data, see Table S1 (Supporting
Information).
SCHEME 1
consequent formation of compounds containing additional
labile hydrogens. Oligomerization is implied by the fact
that a precipitate was formed when these product solu-
tions in tBB were mixed with acetonitrile. This solid,
which had to be removed by filtration prior to analysis,
1
was dissolved in deuterated chloroform, and the H NMR
spectrum showed only the presence of aromatic residues.
No precipitate was formed when the thermolyses of 1 and
2 were run in tBB containing a deliberately added good
hydrogen donor, vide infra.
The induced decomposition of O-benzyl oxime ethers
in hydrogen donor solvents by the RRD process, reaction
1, was most pronounced for Ph2CdNOCH2Ph and Fld
NOCH2Ph.1 This is, presumably, a consequence of the
more favorable thermodynamics arising from the reso-
nance stabilization of the intermediate diarylalkoxyami-
nocarbinyl radical (see reaction 1). Although the solvent-
induced RRD reaction appears not to occur in tBB,2 there
was no assurance that the radicals formed by homolysis
of RR′CdNOPh (reaction 3) might not also be capable of
inducing decomposition. To asses this possibility with the
two oxime ethers where it was expected to be most
important, and to see if the product yields shown in Table
1 could be improved, we studied the decomposition of 1
and 2 in tBB to which was added either 2,2,5,7,8-
pentamethyl-6-hydroxychroman, PMHC (a vitamin E
analogue), or 1,2-diphenylhydrazine, DPHA, both of
which are known to be excellent H-atom donors to
(4) Taylor, E. C.; Kienzle, F. J . Org. Chem. 1971, 36, 233-235.
(5) See, e.g.: Ingold, K. U. In Essays on Free Radical Chemistry;
Special Publication 24, 1970; Chapter 11, Royal Chemistry Society,
London, U.K., ISBN 0-85186-009-5.
J . Org. Chem, Vol. 69, No. 9, 2004 3113

Blake et al.
TABLE 2. Decom p osition of P h 2CdNOP h (1) a n d
F ldNOP h (2) in ter t-Bu tylben zen e Con ta in in g a
Hyd r ogen Don or , XH, a t 90 °C for Va r iou s Tim e
P er iod sa ,b
TABLE 3. F ir st-Or d er Ra te Con sta n ts for Hom olyses of
th e N-O Bon d in O-P h en yl Oxim e Eth er s, RR′CdNOP h ,
1-7 fr om Loss of Eth er in ter t-Bu tylben zen e w ith ou t a n d
in P a r en th esis w ith 2.2 Mola r Equ iv of a Hyd r ogen
Don or , XH, a t Va r iou s Tem p er a tu r esa
time
Ph2NH +
Ph2CO
106k/s-1
XH
(10-3 s)
1
Ph2CNH Ph2CO
PhOH
R
R′
Ph
60 °C
90 °C
110 °C
PMHC
57.6
108.0
172.2
248.4
61.2
108.0
172.8
249.0
40.9
22.0
6.3
91.1
91.7
86.7
86.9
c
c
c
c
3.8
4.3
6.2
7.1
3.2
3.8
4.6
5.9
94.9
96.0
92.9
94.0
c
c
c
c
78.5
78.1
74.8
74.3
75.7
69.7
71.5
71.0
1
2
3
4
5
6-E
7
Ph
0.25 (0.31, -) 12 (15, 15)
110 (160, -)
9-fluorenyl
Me Me
cyclohexyl
bicyclononanyl 4.9 (6.1, -)
Ph
0.12 (-, 0.14)
1.4 (1.7, -)
4.2 (4.8, 4.4) 46 (-, 42)
12 (15, 16)
2.7
DPHA
42.4
21.1
8.1
56 (64, -)
180 (200, -)
Me
Ph
7.7 (7.9, -)
21b (25b,c
)
2.5
PhO
0.61
a
In parentheses, k values as determined in the presence of a
time
FlNH +
XH
(10-3 s)
2
FlNH
FlO
FlO
PhOH
hydrogen donor, XH ) PMHC (first entry), or XH ) DPHA. Rate
constants calculated from product formation are generally in
excellent agreement with those given in this table. They are
PMHC
72.0
100.8
172.8
244.8
64.8
108.0
172.8
249.0
63.6
57.7
44.4
32.5
71.7
63.8
48.5
32.5
45.9
51.0
56.0
53.7
85.4
92.9
90.6
86.1
7.0
8.8
10.4
11.3
8.7
10.4
11.2
11.4
52.9d
59.8d
66.4d
65.0d
94.1
103.3
101.8
97.5
56.4
68.5
76.9
78.0
82.9
97.4
97.7
106.5
b
presented in Table S6 (Supporting Information). At 120 °C, see
Table 4. c No PMHC but with XH ) tetralin as the solvent.
DPHA
SCHEME 2
a
Remaining Ar2CdNOPh is given as a percentage of the initial
percentage of decomposed
concentration and products as
a
b
Ar2CdNOPh
For additional data, see Table S2 (Supporting
Information). c Could not be determined because this imine elutes
d
at the same time as DPHA. Three unidentified products are
formed as well, which are attributed to trapping of the FldN•
radical by the relatively persistent phenoxyl radical from PMHC.
Assuming the same UV response factors as the imine and starting
FldNOPh, the yields of fluorenyl-containing products rise to 59.2,
68.4, 79.6, and 79.7% with increasing time.
those in the presence of a H-donor. This proves that there
is little or no RRD-induced decomposition of either 1 or
2 even with these strong H-donors.6 Rate constants for
the thermolyses of 1 and 2 determined from the rates of
loss of these compounds at 60, 90, and 110 °C are
presented in Table 3. These rate constants at 60 and 110
°C are, like those at 90 °C, also in good agreement with
those determined from the formation of imine plus ketone
or phenol (see also Table S6, Supporting Information).
radicals.6 The initial concentrations of 1 and 2 were
generally 86 mM, and 2.2 molar equiv of these two
H-donors were used, i.e., 189 mM. The results of these
experiments at 90 °C are given in Table 2. Additional
data, including data at 60 °C and 110 °C, are available
as Supporting Information (Table S2).
In tBB, even in the absence of an added H-donor, the
combined yields of imine plus ketone are fairly good (71-
89% of the O-phenyl ether decomposed) but there is a
large deficit in the yields of phenol (17-36%), see Table
1, due presumably to phenoxyl radical dimerization and
oligomerization, vide supra. In the presence of a hydrogen
donor (PMHC for 1 and DPHA for 2), the combined yields
of imine plus ketone become essentially quantitative and
phenol yields increase dramatically (to 70-100%); see
Table 2.
The same data that were used to formulate Tables 1
and 2 can be converted into rate constants for the
thermolyses of 1 and 2. These rate constants can be
derived both from the loss of the O-phenyl ether reactant
and from both the formation of imine plus ketone and
the formation of phenol. For each set of conditions,
reactant and product derived rate constants are in
excellent agreement (see Table S3, Supporting Informa-
tion). More significantly, the rate constants in the
absence of a hydrogen donor are essentially identical to
Since 1 and 2 have the iminyl moieties most suscep-
tible to RRD-induced decomposition of RR′CdNOCH2Ph
compounds1 it would appear to be highly unlikely that
the kinetics for the decomposition of 3-7 in tBB will be
contaminated by such a process. Nevertheless, we did
carry out experiments that confirm the absence of
significant RRD for all of these compounds, vide infra.
Dia lk yl Eth er s 3-5. The product imines from the
three O-phenyl dialkyl oxime ethers are not expected to
absorb strongly at 254 nm, the applied wavelength for
the HPLC analysis. Therefore, rate constants could only
be determined from the rate of loss of the reactant (see
Table 3) and from the rate of formation of phenol. Rate
constants derived by these two procedures were again
in excellent agreement (see Table S6, Supporting Infor-
mation). Compounds 4 and 5 were chosen to explore the
possibility that the observed N-O bond scission in
O-phenyl dialkyl (and alkyl aryl) oxime ethers was not
the straightforward reaction, 3, but instead involved an
initial tautomerism to the enamine followed by N-O
cleavage at an sp3 nitrogen atom, e.g., Scheme 2.
This potential alternative decomposition pathway is
not possible for 5 because the enamine intermediate
would have a double bond at its bridgehead, reaction 5,
which is prohibited by Bredt’s rule. Since 5 decomposes
more rapidly than 4 we can discount decomposition via
(6) For PMHC, see, e.g.: Burton, G. W.; Doba, T.; Gabe, E. J .;
Hughes, L.; Lee, F. L.; Prasad, L.; Ingold, K. U. J . Am. Chem. Soc.
1985, 107, 7053-7065. For DPHA, see, e.g.; Mahoney, L. R.; Ferris,
F. C.; DaRooge, M. A. J . Am. Chem. Soc. 1969, 91, 3883-3889.
Mahoney, L. R.; Menderhall, G. D.; Ingold, K. U. J . Am. Chem. Soc.
1973, 95, 8610-8614.
3114 J . Org. Chem., Vol. 69, No. 9, 2004

Thermolyses of O-Phenyl Oxime Ethers
TABLE 4. Decom p osition of P h O(P h )CdNOP h (7) in ter t-Bu tylben zen e a n d Tetr a lin a t 120 °C for Va r iou s Tim e
P er iod sa ,b
tBB
time (10-3 s)
7
8 imine
9 PhCN
10 oxazole
Σ(8 + 9 + 10)c
PhOH
9/10d
7.2
16.2
57.6
73.8
95.4
83.3
67.2
35.5
20.2
12.1
8
14
4
22
15
0
59
58
55
71
8
8
11
9
86
81
73
86
94
78
74
79
76
83
8.8
7.4
5.3
6.1
8.9
8
106 k/s-1
21
0.982
e
e
21
0.947
21
0.984
R2
tetralin
time (10-3 s)
7
8 imine
9 PhCN
10 oxazole
Σ(8 + 9 + 10)c
PhOH
9/10d
3.6
8.1
18.0
64.8
79.2
84.9
70.8
62.2
21.6
12.3
14
15
29
f
26
25
36
f
5
5
8
f
45
45
73
62
72
118
f
5.2
5.0
4.5
24
49
9
82
137
5.4
106 k/s-1
25
0.984
e
e
21
0.886
e
R2
a
Remaining 7 is given as a percentage of the initial concentration, and products are given as a percentage of decomposed 7. Rate
constants (and R2) calculated from loss of 7 and from formation of some products are included in italics. For additional data, see Table
a
S7 (Supporting Information). Nitrogen balance for product formation. c Ratio of the rates of â-scission to cyclization, see Scheme 3.
b
Not calculated because the large changes in the ratio [product]t/([7]0 - [7]t) suggest that additional chemistry is occurring. e Not measured.
d
the enamine tantomers of 3-6 and other O-phenyl alkyl
oxime ethers.
oxidative stress in biological systems. The search for a
thermally more stable ARTS-2 family led to 7 with the
expectation finding reactions 7 and 8:
PhO(Ph)CdNOPh slow8 PhO(Ph)CdN• + PhO• (7)
7
PhO(Ph)CdN• â-scission8
(8)
PhCN + PhO•
Alk yl Ar yl Eth er 6. This compound is formed as a
mixture of 6-E (83%, Me and PhO cis to one another) and
6-Z (17%) isomers. After heating at 90 °C for 24 h, 94%
of the Z-isomer but only 46% of the E-isomer had
decayed. The rate constant for decay of the Z-isomer at
90 °C is 33.2 × 10-6 s-1, but this rate constant probably
contains contributions from N-O bond homolysis and
from isomerization of the less to the more stable isomer.7
For this reason, Table 3 contains only rate constants
measured for decay of the E-isomer (which can, of course,
be determined reliably after most of the Z-isomer has
been destroyed). Full kinetic details are given in Tables
S4-S6 (Supporting Information).
9
Reaction 8 was expected to be fairly rapid because the
â-scission of ketiminyl radicals, or rather of sterically
crowded ketiminyls, is known to be a fairly facile process,
e.g.,9 the â-scission: (Me3C)2CdN• f Me3CCN + Me3C•,
occurs at a rate comparable to that for the well-known
â-scission: Me3CO• f Me2CO + Me•.
In the event, the thermolyses of 7 in tBB at tempera-
tures from 90 to 150 °C showed that reactions 7 and 8
represent 7’s major decomposition pathways, and there-
fore, 7 represents the first member of the new ARTS-2
family. Next to benzonitrile, 9, and phenol, two additional
products emerged, which were identified (with EIMS or
GCMS) as phenoxy phenyl ketimine, 8, and 2-phenyl-
benzoxazole, 10. Quantitative HPLC analyses were based
on authentic 7, 9, 10, and phenol, but for 8 we had to
use the nearest carbonyl analogue, phenyl benzoate. The
product yields from a thermolysis at 120 °C in tBB are
given in Table 4 where it can be seen that Σ([8] + [9] +
[10]), i.e., the “nitrogen balance” for product formation,
is fairly good (73-94%), a result which justifies our use
of the phenyl benzoate response to quantify 8.
N-P h en oxyben zim id ic Acid P h en yl Eth er , 7. We
have previously invented a clean, novel, aryloxyl radical
thermal source, ARTS-1 (reaction 6), and synthesized two
examples
k6
9
(ArOCH2ONd)2
8 2 ArO• + 2 CH2O + N2 (6)
(ArO• ) PhO• and R-tocopheroxyl).8 These compounds
decayed rather too rapidly at 37 °C (k6 )10-3 s-1) to be
useful for long-term studies of aryloxyl radical-induced
As with 1 and 2, a precipitate formed when the tBB
solutions of thermolyzed 7 were diluted with acetonitrile,
(7) Isomerization routes potentially include (i) N-O bond homolysis
followed by an in-cage collapse of the geminate iminyl/phenoxyl radical
pair to the more stable oxime ether isomer, (ii) inversion at nitrogen,
and (iii) rotation about the CdN double bond. Theoretical calculations
on the isomerizations of aldoximes suggest that the barriers for (ii)
and (iii) will be very high in energy (they are both ca. 61 kcal/mol for
the isomerization of vinylaldoxime), favoring (i). See: Alberti, A.;
Barbaro, G.; Battaglia, A.; Guerra, M.; Bernardi, F.; Dondoni, A.;
Pedulli, G. F. J . Org. Chem. 1981, 46, 742-750.
1
and again the H NMR spectrum of this solid dissolved
in CDCl3 showed only a complex pattern in the aromatic
region. Since the nitrogen balance is good, this precipitate
must be phenoxyl radical dimers and oligomers, the
formation of which is consistent with yields of phenol in
(8) Paul, T.; Ingold, K. U. Angew. Chem., Int. Ed. 2003, 41, 804-
806.
(9) Griller, D.; Mendenhall, G. D.; Van Hoof, W.; Ingold, K. U. J .
Am. Chem. Soc. 1974, 96, 6068-6070.
J . Org. Chem, Vol. 69, No. 9, 2004 3115

Blake et al.
TABLE 5. Decom p osition of P h O(P h )CdNOP h (7) a t 90 °C in ter t-Bu tylben zen e in th e Absen ce or P r esen ce of XH a n d
P MHC for Va r iou s P er iod s of Tim ea ,b
XH
time (10-3 s)
7
8 imine
9 PhCN
10 oxazole
Σ(8 + 9 + 10)c
PhOH
9/10d
none
82.8
169.2
354.6
435.6
1040.4
93.9
88.3
79.1
76.8
53.1
11
15
20
23
25
64
53
45
46
42
9
9
8
9
9
84
77
73
78
76
69
65
62
66
70
7.1
5.9
5.6
5.1
4.7
106k/s-1
0.61
0.998
e
e
0.55f
0.934
0.56f
0.945
R2
PMHC
504.0
595.5
72.2
66.8
59.7
42.0
21
20
21
21
20
15
15
16
8
8
8
9
49
43
44
46
120
117
128
145
(2.5)g
(1.9)g
(1.9)g
(1.8)g
761.5
1277.0
106k/s-1
0.67
0.998
0.68
0.988
0.74
0.876
0.60
0.971
0.57
0.945
R2
a
Remaining 7 is given as a percentage of the initial concentration, and products are given as a percentage of decomposed 7. Rate
constants (and R2) calculated from loss of 7 and from formation of some products are included in italics. For additional data, see Tables
b
S7 and S8 (Supporting Information). c Nitrogen balance for product formation. Ratio of the rates of â-scission to cyclization, see Scheme
d
3. e Not calculated because the large monotonic change with time of the ratio [product]t/([7]0 - [7]t) indicates that this product suffers
from time-dependent chemistry as the reaction progresses from t ) 0 to t. f Since the ratio [product]t/([7]0 - [7]t) is reasonably constant,
g
this rate constant was calculated by plotting ln({[product]t/[7]0 - [7]t}average × [7]0 - [product]t) vs t. These ratios reflect the loss of a
substantial fraction of the PhCN presumably formed in the reaction, see text.
SCHEME 3. Th er m a l Decom p osition P a th w a ys for
benzonitrile in degassed tBB at 120 °C for 5 days.
Compound 10 was produced, but there was no trace of
2-(4-tert-butylphenyl)benzoxazole. Since the phenoxyl/
phenol f phenol/phenoxyl reaction is fast,10,11 the absence
of the tert-butyl-substituted 10 proves that the formation
of 10 itself does not involve the phenoxyl radical. The
mechanism of formation of compound 10 is, therefore,
most probably a 5-exo intramolecular cyclization of the
first formed phenoxy phenyl ketiminyl radical followed
by oxidative aromatization of the resulting cyclohexadi-
enyl radical (Scheme 3).
P h O(P h )CdNOP h a
The partitioning of the phenoxy phenyl ketiminyl’s
unimolecular decay routes between â-scission and cy-
clization is given by the product ratio [9]/[10]. The mean
value of this ratio in tBB at 120 °C is 7.3. That is,
â-scission, reaction 8, which yields phenoxyl radicals, is
favored over cyclization. The â-scission/cyclization ratio
appears to have only a slight temperature dependence
with a mean value of 5.7 at 90 °C (Table 5) which rises
to ca. 9-10 at 150 °C (see also Table S9, Supporting
Information).
The hydrogen donor solvent, tetralin, accelerated the
decomposition of O-benzyl oxime ethers by RRD, reaction
1.1 Tetralin was therefore employed in a preliminary
exploration of RRD for 7 since extensive RRD would
certainly limit the utility of this compound as a kineti-
cally “steady”, thermal source of phenoxyl radicals.
a
XH is a hydrogen atom donor which may be the solvent, a
deliberately added XH (e.g., PMHC), and the phenoxyl radical
dimers.
Comparison of the products of thermolysis of 7 at 120
°C in tetralin reveals that there is less benzonitrile and
more phenol in this solvent than in tBB, as might be
expected (Table 4). However, the rate constants calcu-
lated from the decay of 7 and from product formation
show no significant differences in these two solvents
(Table 4). That RRD-induced decomposition of 7 is
unimportant was further confirmed by the thermolysis
the 74-83% range (Table 4). That is, in the presence of
a good hydrogen donor the yield of phenol should be
(100% + 9 (%)), see Scheme 3, i.e., 155-171%, but the
actual yields are only half as large (Table 4).
There are three potential sources of 10. The first is a
direct reaction between two of the products, phenol and
benzonitrile, followed by oxidative aromatization. This
possibility was eliminated by heating 81 mM phenol plus
81 mM benzonitrile in degassed tBB at 150 °C for ca. 24
h. No 10 was produced. The second is a direct reaction
between a phenoxyl radical and benzonitrile, followed by
oxidation. This possibility was eliminated by heating 86
mM 7, plus 88 mM 4-tert-butylphenol, plus 100 mM
(10) Foti, M.; Ingold, K. U.; Lusztyk, J . J . Am. Chem. Soc. 1994,
116, 9440-9447.
(11) (a) Lucarini, M.; Pedulli, G. F.; Cipollone, M. J . Org. Chem.
1994, 59, 5063-5070. (b) Lucarini, M.; Pedrielli, P.; Pedulli, G. F.;
Ciabiddu, S.; Fattuoni, C. J . Org. Chem. 1996, 61, 9259-9263.
3116 J . Org. Chem., Vol. 69, No. 9, 2004

Thermolyses of O-Phenyl Oxime Ethers
of 7 at 90 °C in tBB and in tBB containing 2.2 molar equiv
of the hydrogen atom donor, PMHC. There was again a
decrease in the yield of benzonitrile, a definite increase
in the yield of phenol, but with no significant changes in
the yields of imine and oxazole (Table 5). However, the
rate constants derived from loss of 7 and from formation
of the products were the same, within our experimental
accuracy, in the absence or presence of PMHC (Table 5),
indicating that there is no significant RRD-induced
thermal decomposition of 7 even in the presence of a good
H-atom donor. There is obviously some subsequent
chemistry among the products which, although it does
not appreciably change the rate of decomposition of 7,
does cause a loss of benzonitrile and, hence, a large
change in the benzonitrile/oxazole (9/10) ratio. This
secondary chemistry was not explored.
also holds for O-aryl oxime ethers. That is, the calculated
N-O BDE for O-4-methoxyphenyl acetoxime ether is 29.9
kcal/mol, which is 5.9 kcal/mol lower than the one for
O-phenyl acetoxime ether (vide infra). For the O-phenyl
ketoxime ethers 1-6, the activation enthalpies for N-O
bond homolysis decrease from ca. 34.2 kcal/mol for 2 to
ca. 31.2 kcal/mol for 5, vide infra. The corresponding O-4-
methoxyphenyl ketoxime ethers would therefore be ex-
pected to have activation energies for decomposition
lower by ca. 6 kcal/mol, i.e., 28.2 and 25.2 kcal/mol,
respectively, from which the respective rate constants can
be estimated, vide infra, to be ca. 1.6 × 10-2 and 1 s-1 at
90 °C and ca. 3 × 10-6 and 5 × 10-4 s-1 at 25 °C. This
implies that at room temperature the O-4-methoxyphenyl
analogue of 2 could probably be synthesized but the
analogue of 5 would be too unstable. These calculated
rate constants are certainly not precise. They are in-
tended solely to indicate that the thermal stability of
O-aryl ketoxime ethers derived from a specific ketone can
be “tuned” to provide the corresponding ketiminyl radical
at a temperature desired for some planned synthesis or
other experiment.
Discu ssion
Th er m a l Sta bilities of RR′CdNOP h , 1-6. The
thermal decomposition of O-phenyl ketoxime ethers has
been shown to occur by a simple, unimolecular N-O bond
homolysis unperturbed by any parallel RRD-induced
decomposition even in the presence of very good H-atom
donors. Thermal decomposition occurs at reasonable rates
at relatively low temperatures. This opens the prospect
of using O-phenyl ketoxime ethers as thermal initiators
of radical chain reactions and as thermal sources of
ketiminyl radicals for organic syntheses.
The relatively similar thermal stabilities of 1-6 are
fully consistent with the known σ radical nature of
iminyls. In these radicals, the unpaired electron resides
mainly in the nitrogen’s 2py atomic orbital and therefore
lies in the plane of the local molecular framework 12:9,14
For many potential uses, these compounds have an-
other advantage that we have not attempted to explore
and exploit in the present, very preliminary survey of
their properties. This advantage arises from their facile
synthesis at room temperature (see the Experimental
Section). The structure of the ketone from which the
O-phenyl ketoxime ethers are synthesized has only a
rather small effect on the rate of N-O bond homolysis;
i.e., the maximum difference in decomposition rates at
90 °C is the 43-fold faster decomposition of the bicy-
clononan-9-one derivative, 5, than the fluorenone deriva-
tive, 2, see Table 3. However, it is well established, both
by experiment11,12 and theory,13 that the O-X bond
dissociation enthalpies (BDEs) in 4-YC6H4O-X (X ) H,
CH3, CH2Ph) can be increased and decreased dramati-
cally by using Y groups which are electron-withdrawing
(EW) or electron-donating (ED), respectively. For example,
the O-X BDEs in these three classes of compounds are
ca. 6 kcal/mol lower when Y ) MeO than when Y )
H.11-13 Our DFT calculations suggest that this difference
The unpaired electron does not delocalize to the sp2
carbon atom, and hence, iminyl radicals are not stabilized
when R and/or R′ are aromatic groups. However, there
is a powerful hyperconjugative interaction, the presence
of which can be deduced from the large hydrogen hyper-
fine splitting (hfs) in H2CdN• (cf. 12 T 13, R, R′ ) H, H)
of about 87 G.14 Since a “free” hydrogen atom (which
necessarily has 100% of the unpaired electron in its 1s
orbital) has a hfs ≈ 507 G,15 the unpaired spin density
on each hydrogen in H2CdN• is about 17%. Such hyper-
conjugation will obviously stabilize ketiminyl radicals
and will be favored by R groups that are inductively
electron donating (ED) and, hence, increase the electron
density in the R-C bond. Similarly, hyperconjugation
will be disfavored for R groups that are electron with-
drawing (EW) and decrease the electron density in the
R-C bond. Alkyl groups are ED and are therefore
expected to stabilize ketiminyl radicals and hence weaken
the N-O bond in RR′CdNOPh compounds, whereas aryl
groups are EW and are expected to destabilize ketiminyls
and strengthen the N-O bonds. These concepts are
congruent with our experimental observations that the
rate constants for decomposition of O-phenyl ketoxime
ethers at 90 °C are smallest for the compounds derived
from benzophenone (1), fluorenone (2), and acetophenone
(6-E) and have magnitudes that span only a factor of 3;
(12) (a) Mulder, P.; Saastad, O. W.; Griller, D. J . Am. Chem. Soc.
1988, 110, 4090-4092. (b) Lind, J .; Shen, X.; Eriksen, T. E.; Mere´nyi,
G. J . Am. Chem. Soc. 1990, 112, 479-482. (c) J onsson, M.; Lind, J .;
Eriksen, T. E.; Mere´nyi, G. J . Chem. Soc., Perkin Trans. 2 1993, 1567-
1568. (d) Bordwell, F. G.; Zhang, X.-M.; Satish, A. V.; Cheng, J .-P. J .
Am. Chem. Soc. 1994, 116, 6605-6610. (e) Wayner, D. D. M.; Lusztyk,
E.; Ingold, K. U.; Mulder, P. J . Org. Chem. 1996, 61, 6430-6433. (f)
Dorrestijn, E.; Laarhoven, L. J . J .; Arends, I. W. C. E.; Mulder, P. J .
Anal. Appl. Pyrol. 2000, 54, 153-192. (g) Pratt, D. A.; de Heer, M. I.;
Mulder, P.; Ingold, K. U. J . Am. Chem. Soc. 2001, 123, 5518-5526.
(13) (a) J onsson, M.; Lind, J .; Mere´nyi, G.; Eriksen, T. E. J . Chem.
Soc., Perkin Trans. 2 1994, 2149-2154. (b) Wu, Y.-D.; Lai, D. K. W. J .
Org. Chem. 1996, 61, 7904-7910. (c) Brinck, T.; Haeberlein, M.;
J onsson, M. J . Am. Chem. Soc. 1997, 119, 4239-4244. (d) DiLabio, G.
A.; Pratt, D. A.; LoFaro, A. D.; Wright, J . S. J . Phys. Chem. A 1999,
103, 1653-1661. (e) DiLabio, G. A. J . Phys. Chem. A 1999, 103, 11414-
11424. (f) Pratt, D. A.; DiLabio, G. A.; Valgimigli, L.; Pedulli, G. F.;
Ingold, K. U. J . Am. Chem. Soc. 2002, 124, 11085-11092.
(14) Cochran, E. L.; Adrian, F. J .; Bowers, V. A. J . Chem. Phys. 1962,
36, 1938-1942. Brivati, J . A.; Root, K. D. J .; Symons, M. C. R.;
Tingling, D. J . A. J . Chem. Soc. A. 1969, 1942-1945. Symons, M. C.
R. Tetrahedron 1973, 29, 615-619.
(15) Fischer, H. In Free Radicals; Kochi, J . K., Ed.; Wiley: New
York, 1973; Vol. 2, p 439.
J . Org. Chem, Vol. 69, No. 9, 2004 3117

Blake et al.
see Table 3. The rate constant for decomposition of the
O-phenyl ether derived from acetone (3) is comparable
to that for the ether derived from benzophenone (Table
3). The compound derived from cyclohexanone (4) decays
roughly five times more rapidly than that derived from
acetone (3), while the compound derived from bicy-
clononan-9-one (5) decays roughly 20 times faster than
3. This order is consistent with the stabilization of
ketiminyl radicals by hyperconjugation (13), which will
increase along the series involving a formal contribution
from a methyl radical (14), a primary alkyl radical (15),
and a secondary alkyl radical, (16):
weaker and longer than Ph-C bonds which should favor
hyperconjugation, 13 with R′ ) Ph, R• ) PhO• being more
favored than R• ) Ph•. In the event, the rate constant
for N-O bond homolysis at 90 °C is ca. 20 times smaller
for 7 than for 1 (Table 3), a direction and a difference
which we interpret as further evidence for the large role
of inductive effects on the thermochemistry of iminyl
radical reactions.
In one sense, the thermal stability of 7 (intuitively
surprising and not predicted by theory, vide infra) proved
a “blessing” because this compound was prepared from
precursors by heating on a boiling-water bath for 6 h.
The half-life for 7 at 100 °C is ca. 3.3 days. If 7 had had
the same thermal stability as 1, the half-life at 100 °C
would have been ca. 3.4 h and its synthesis could not have
been achieved in reasonable yield. Until a low-temper-
ature synthesis of 7 is reported, “tuning” the thermal
stability of PhO(Ph)CdNOAr by addition of substituents
to the 4-position of the OAr group can only be in the
direction of thermally more stable N-phenoxybenzimidic
acid aryl esters. This could be achieved using EW
substituents, e.g., with Ar ) 4-CF3C6H4 or 4-MeC(O)C6H4,
etc. However, even with the present synthetic route, the
efficiency of â-scission of Ar′O(Ph)CdN• radicals relative
to their cyclization to form an oxazole (e.g., 10) and
H-atom abstraction to form the corresponding imine,
could be greatly improved over that found for PhO(Ph)Cd
N• by using an ED 4-substituent on the Ar′O group which
would stabilize the Ar′O• and hence weaken the Ar′O-C
bond (e.g., Ar′O ) 4-MeOC6H4O, vide supra).
N-O Bon d Scission for 1-7, Activa tion En th a l-
p ies, a n d Ca lcu la ted Bon d Dissocia tion En th a lp ies,
BDEs. The temperature ranges used in our studies of
the kinetics of decomposition of RR′CdNOPh (Table 3)
are far too limited to yield reliable Arrhenius parameters
for these N-O bond-cleavage reactions. However, the
measured rate constants should be fairly reliable, and
since 90 °C was a common temperature in our experi-
ments we have used rate constants at 90 °C to estimate
the activation enthalpies, Ea, for the reactions. The
Arrhenius preexponential factor for N-O bond scission
of RR′CdNOPh to RR′C) N• + PhO• has been assumed
to be the same as that for C-O bond scission of CH3OPh
Two brief attempts to explore further the relation
between the thermal stabilities of O-phenyl ketoxime
ethers, RR′CdNOPh, and the inductive effects of R and
R′ were not successful:
(i) (4-MeOC6H4)2CdNOPh was expected to be less
stable than Ph2CdNOPh (1) because of the strong ED
effect of 4-MeO groups. The compound was successfully
synthesized although the reaction of (4-MeOC6H4)2CO
with PhONH2 was extremely sluggish at room temper-
ature (ca. 10% yield after 48 h). Heating to 100 °C caused
the desired product to decompose (which would not have
occurred so rapidly with 1), yet still a large amount of
the starting ketone was recovered. The crude (4-
MeOC6H4)2CdNOPh turned dark brown in air at room
temperature and could not be purified. (ii) Ph(C6F5)Cd
NOPh was expected to be very stable because of the
strong EW effect of the five fluorine atoms. Unfortu-
nately, the ketone did not react with PhONH2 at room
temperature, and after heating to 100 °C for 5 h, there
was none of the desired product (MS) yet still some of
the starting ketone. Further exploration of the relation
between thermal stabilities and inductive effects was
deemed an inappropriate use of resources at the present
stage in the development of RR′CdNOPh and related
compounds into reagents useful for organic synthesis.
Th er m a l Sta bility of P h O(P h )CdNOP h , 7. In prin-
ciple, 7 should provide an additional test of the relation
between inductive effects and N-O BDEs because PhO
will be inductively more strongly EW than Ph, on which
basis 7 would be predicted to be thermally more stable
than Ph2CdNOPh (1). However, it appeared possible that
7 would be less stable than 1 because PhO-C bonds are
to CH3 + PhO•, viz., log(A/s-1) ) 15.2.12g,16,17 Activation
•
enthalpies calculated with this assumption are given in
Table 6. DFT calculated N-O BDEs have been included
for comparison. For 1, the calculated N-O BDE is 1.5
kcal/mol larger than the estimated activation enthalpy
while for 2-6 the calculated N-O BDEs are 2.8 ( 0.4
kcal/mol larger than the estimated activation enthalpies.
The direction of these differences may indicate that the
assumed value of log(A/s-1) of 15.2 is too low. However,
for PhO(Ph)CdNOPh (7), the calculated N-O BDE is 2.4
kcal/mol lower than the estimated activation enthalpy.
The reason for this “anomalous” BDE remains unclear
even after extensive additional computations. In contrast,
the calculated N-H BDE in PhO(Ph)CdNH is stronger
than in Me2CdNH, just as the Ea for PhO(Ph)CdNOPh
is larger than for Me2CdNOPh (Table 6).
(16) Sato, Y.; Yamakawa, T. Ind. Eng. Chem. Fundam. 1985, 24,
12-15.
(17) See also: Ciriano, M. V.; Korth, H.-G.; van Scheppingen, W.
B.; Mulder, P. J . Am. Chem. Soc. 1999, 121, 6375-6381.
3118 J . Org. Chem., Vol. 69, No. 9, 2004

Thermolyses of O-Phenyl Oxime Ethers
TABLE 6. Th er m od yn a m ic P a r a m eter s for
but the only reported N-H BDE for an imine of which
we are aware is Bordwell and J i’s21 value of 117 kcal/
mol for Ph2CdNH. This value was based on the imine’s
equilibrium acidity in DMSO and the oxidation potential
of its conjugate base. However, it is obviously much too
large because 117 kcal/mol is not only 9 kcal/mol stronger
than the N-H BDE in NH3,18 but also 117 kcal/mol would
make reaction 11 endothermic by ca. 13 kcal/mol, in
which case this reaction could not have provided iminyl
radicals for spectroscopic studies. The σ structure of
iminyl radicals is reflected in the similar magnitudes of
the quantities listed in the various columns in Table 6,
viz., k, Ea, N-O BDE, and N-H BDE.
RR′CdNOP h , 1-7, a n d RR′CdNHa
b
c
d
R
R′
Ea
N-O BDE
N-H BDE
1
2
3
4
5
6-E
7
Ph
Ph
33.4
34.2
33.4
32.3
31.2
33.7
35.5
34.9
37.0
35.8
34.9
34.4
36.1
33.1
91.7
92.4
90.2
9-fluorenyl
Me Me
cyclohexyl
bicyclononanyl
Ph
PhO
e
89.7
89.0
91.4
97.7
Me
Ph
a
b
Units, kcal/mol. Rate constants used are at 90 °C and are
taken from Table 3 for experiments with no added XH; for
homolysis a log(A/s-1) ) 15.2 is taken; see text. c DFT-computed
N-O BDE using (RO)B3P86/6-311G(d,p)//(U)B3P86/6-311G(d,p)
d
as in ref 22. DFT-computed imine N-H BDE using (RO)B3LYP/
We are currently exploring the utility of O-phenyl
ketoxime ethers as low-temperature thermal sources of
iminyl radicals for organic synthesis.
e
6-311+G(2d,2p)//(U)B3LYP/6-31G(d) as in ref 13e. 90.0 kcal/mol
by CBS-QB3.
Exp er im en ta l Section
The thermolysis of all seven of the O-phenyl oxime
ethers, 1-7, involves a “clean” homolysis of the N-O
bond and release of the resonance-stabilized phenoxyl
radical, reaction 3. Interestingly, the N-O BDEs for 1-7
are all less than the O-O BDEs of dialkyl peroxides (38-
40 kcal/mol). Such weak N-O bonds are consistent with
the detection by EPR spectroscopy of iminyl radicals
during the thermolyses of certain thionocarbonates19
(reaction 9) and the strong 31P CIDNP
Gen er a l P r oced u r e for th e Syn th esis of O-P h en yl
Oxim e Eth er s 1-6, Exem plified with 9-Flu or en on e Oxim e
Eth er , 2. O-Phenylhydroxylamine hydrochloride (1.0 g, 6.87
mmol) was dissolved in anhydrous pyridine (20 mL) under N2
at room temperature, and 9-fluorenone (1.24 g, 6.87 mmol) was
added to the solution in one portion. The resulting solution
was stirred at room temperature overnight, and the progress
of the reaction was monitored by TLC (hexane/ethyl acetate
) 10:1). Upon completion, the reaction mixture was poured
into distilled water (40 mL) and extracted with EtOAc (3 ×
30 mL), and the combined organic phases were washed several
times with saturated, aqueous CuSO4 solution to remove any
traces of pyridine. The solution was then dried (MgSO4) and
concentrated on a rotavap, and the desired product, 2, was
purified by column chromatography (hexane/ethyl acetate )
10:1): yellow solid (1.60 g, 86%); mp 95 °C; 1H NMR δH 7.15-
8.53 (m); 13C NMR δC 115.3, 120.5, 122.7, 123.4, 128.5, 128.8,
129.9, 130.3, 130.8, 131.0, 132.1, 135.7, 141.1, 142.3, 155.0,
160.0; HRMS calcd for C19H14ON 272.1075, found 272.1088.
RR′CdNOC(S)NMe2 9
∆8RR′CdN•(+ Me2NC(S)O•) (9)
effects seen during the rearrangement of Ph2CdNOPPh2
at 0 °C20 (reaction 10). The N-O BDEs
Ph2CdNOPPh2
9
8 [Ph2CdN• •OPPh2]
Ph2C)NP(O)Ph2 (10)
98
1
O-P h en yl ben zop h en on e oxim e eth er , 1: mp 53 °C; H
NMR δH 7.04-7.08 (m, 1H), 7.27-7.52 (m, 12H), 7.63-7.65
(m, 2H); 13C NMR δC 115.3, 122.7, 128.5, 128.7, 128.9, 129.6,
for O-alkyl oxime ethers are expected to be stronger than
those of O-phenyl oxime ethers by about the difference
between the O-H BDE for alcohols, 104.2 ( 0.9 kcal/
mol,18 and the O-H BDE for phenol, 87.3 ( 1.5 kcal/
mol,18 i.e., by 16.9 ( 2.4 kcal/mol. Our calculations of
N-O BDEs for some RR′CdNOCH2Ph2 show that this
expectation is fulfilled, i.e., for the same RR′ groups the
N-O ∆BDE (RR′CdN-OCH2Ph - RR′CdN-OPh) var-
ies from 15.2 (5, RR′ ) bicyclononanyl) to 16.6 (2, RR′ )
9-fluorenyl) kcal/mol. There are no reliable experimental
N-H BDEs for imines. Our calculations (Table 6) indi-
cate that they are roughly 55 kcal/mol stronger than the
N-O BDEs of the corresponding RR′CdNOPh com-
pounds. Relatively weak imine N-H bonds are consistent
with the facile generation of iminyl radicals for EPR
spectroscopic study by reaction 11,9
129.8, 130.4, 133.3, 136.4, 160.0, 160.3; HRMS calcd for C19H16
ON 274.1154, found 274.1232.
-
O-P h en yl a ceton e oxim e eth er , 3: liquid; 1H NMR δH 2.05
(s, 3H), 2.09 (s, 3H), 6.99-7.34 (m, 5H); 13C NMR δC 16.6, 22.3,
115.0, 122.1, 129.6, 158.9, 159.9; HRMS calcd for C9H12ON
150.0919, found 150.0990.
O-P h en yl cycloh exa n on e oxim e eth er , 4: mp 42-44 °C;
1H NMR δH 1.64-1.81 (m, 6H), 2.36-2.39 (m, 2H), 2.67-2.71
(m, 2H), 6.98-7.33 (m, 5H); 13C NMR δC 26.2, 26.3, 26.4, 27.5,
32.6, 115.0, 122.0, 129.6, 160.0, 164.3; HRMS calcd for C12H16
ON 190.1232, found 190.1185.
-
O-P h en yl bicyclo[3.3.1]n on a n -9-on e oxim e eth er , 5: mp
63-64 °C; 1H NMR δH 1.58-1.64 (m, 2H), 1.91-2.02 (m, 8H),
2.03-2.15 (m, 2H), 2.70 (s, 1H), 3.68 (s, 1H), 6.97-7.34 (m,
5H); 13C NMR δC 21.6, 30.6, 32.6, 34.0, 34.8, 36.5, 114.9, 121.8,
129.6, 160.1, 171.4; HRMS calcd for C15H20ON 230.1545, found
230.1560.
O-P h en yl a cetop h en on e oxim e eth er , 6: 6-E:6-Z ) 5:1;
1H NMR δH for isomer (a) 2.49 (s, 3H), 7.30-7.83 (m, 10H);
for isomer (b) 2.38 (s, 3H), 7.05-7.79 (m, 10H); 13C NMR δC
for the mixture of isomers: 13.8, 22.3, 115.0, 115.17, 115.22,
122.2, 122.4, 122.6, 126.9, 128.4, 128.6, 128.9, 129.6, 129.7,
130.1, 136.4, 158.0, 158.2, 160.0; HRMS calcd for C14H14ON
212.1075, found 212.1081.
Me3CO• + RR′CdNH
9
8 Me3COH + RR′CdN• (11)
(18) Ingold, K. U.; Wright, J . S. J . Chem. Educ. 2000, 77, 1062-
1064.
(19) Hudson, R. F.; Lawson, A. J .; Lucken, E. A. C. J . Chem. Soc.,
Chem. Commun. 1971, 807-808. Hudson, R. F.; Lawson, A. J .; Lucken,
E. A. C. J . Chem. Soc., Chem. Commun. 1972, 721-722. See also:
Brown, C.; Hudson, R. F.; Lawson, A. J . J . Am. Chem. Soc. 1973, 95,
6500-6502.
(20) Brown, C.; Hudson, R. F.; Maron, A.; Record, K. A. F. J . Chem.
Soc., Chem. Commun. 1976, 663-664.
Note: All the O-phenyl oxime ethers appear to be light
sensitive, especially in solution. Presumably, it is light sensi-
(21) Bordwell, F. G.; J i, G.-Z. J . Am. Chem. Soc. 1991, 113, 8398-
8401.
(22) DiLabio, G. A.; Pratt, D. A. J . Phys. Chem. A 2000, 104, 1938-
1943.
J . Org. Chem, Vol. 69, No. 9, 2004 3119

Blake et al.
tivity that causes all these compounds to become more strongly
colored than the crude materials after two or three column
purifications!
N-P h en oxyben zim id ic Acid P h en yl Eth er , 7. N-Phe-
noxybenzimidic acid chloride (1.59 g, 6.89 mmol), sodium
phenoxide trihydrate (1.17 g, 6.89 mmol), and DMSO (4.5 mL)
were stirred and heated under nitrogen on a boiling-water bath
for 6 h. The mixture was cooled, taken up in diethyl ether (40
mL), and extracted with water (4 × 5 mL). The ether layer
was evaporated, and the residue was dissolved in hexane (4
mL) and separated on a silica column using hexane/ethyl
acetate (97:3) as eluent. The oily product (1.4 g), which
solidified in the fridge, was taken up in EtOH (20 mL). This
solution was warmed to 50 °C, water was added, and the
mixture was cooled. The solid was filtered, dried, and dissolved
at 50 °C in a mixture of EtOH (30 mL) and H2O (6 mL) and
recrystallized to afford white crystals of the title compound
(0.9 g, 45%, which nevertheless contained e1% of 2,4,6-
triphenyl-1,3,5-triazine, the benzonitrile trimer (see ref 4), as
an impurity which was not removed): mp 54-55 °C; ESI-MS
290; IR ν/cm-1 1588, 1490, 1324, 1301, 1199, 1162, 1069, 1024,
1000, 966, 924, 767, 752, 688; 1H NMR δH 7.01-7.12 (4H, m),
7.18-7.22 (2H, m), 7.27-7.35 (4H, m), 7.41-7.48 (3H, m),
7.89-7.93 (2H, m); 13C NMR, 114.8, 116.6, 122.7, 123.4, 127.5,
128.8, 129.4,129.7, 129.8, 130.0, 131.1, 153.6, 155.8, 159.2.
N-P h en oxyben zim id ic Acid P h en yl Eth er , 7. This syn-
thesis is based on that of Taylor and Kienzle.4
P ota ssiu m Ben zoh yd r oxa m a te. Benzohydroxamic acid
(10.00 g, 73 mmol) was dissolved in hot anhydrous EtOH (60
mL), and a hot solution of KOH (4.1 g, 73 mmol) in anhydrous
EtOH (50 mL) was added portionwise with mixing. A white
precipitate formed immediately. Once all the KOH was added,
the mixture was stirred for another 30-60 min, cooled, and
filtered. The filtrate was partially evaporated, and the newly
formed precipitate was filtered as well. The combined solids
were dried under vacuum (1 Torr) for 2 h to give the salt (7.35
g). Additional product was recovered from the remaining
mother liquor and recrystallized from EtOH.
N-P h en oxyben za m id e. Diphenyliodonium chloride (9.48
g, 31.2 mmol) and potassium benzohydroxamate (5.46 g, 31.2
mmol) were placed in a 250 mL round-bottom flask. The flask
was purged with nitrogen, and t-BuOH (125 mL) was added.
The mixture was stirred and refluxed under nitrogen for 4 h,
filtered, and evaporated. The residue was taken up in diethyl
ether (50 mL) and extracted with 1 M aq NaOH (4 × 15 mL).
The extracts were combined and acidified with dilute HCl. The
precipitate was filtered and dried to give (4.0 g, 60%) of crude
material. This product was sufficiently pure for the next step,
but nevertheless was recrystallized from a mixture of hexane
and ethyl acetate: ESI-MS 214.
Ack n ow led gm en t. J .C.W. thanks the Royal Society
of Chemistry for a travel grant, and we thank Malgosia
Daroszewska for technical support.
N-P h en oxyben zim id ic Acid Ch lor id e. N-Phenoxyben-
zamide (1.88 g, 8.82 mmol) and PCl5 (1.94 g, 9.30 mmol) were
dissolved in CCl4 (68 mL), cooled to 0 °C, and stirred under
nitrogen for 6 h. The solvent was evaporated, and the residue
was taken up in diethyl ether (27 mL) and washed twice with
water (7 and 3 mL). After evaporation, the syrupy residue
solidified in the freezer. The product was triturated under 5
mL of cold 50% aq EtOH, the solvent was then removed by
pipet, and the residue was dried under vacuum (1 Torr) for 2
h to give pale, semisolid material (1.59 g, 78%): ESI-MS 232.
Su p p or tin g In for m a tion Ava ila ble: General (analytical)
methods, 13C NMR spectra of compounds 1-7. Additional
kinetic data for the decomposition of compounds 1-7, Tables
S1-S9. Geometries in Cartesian coordinates, electronic ener-
gies, thermochemical corrections, and enthalpies of all struc-
tures computed and presented in Table 6 or in the text, Tables
S10-S13. This material is available free of charge via the
Internet at http://pubs.acs.org.
J O049927Y
3120 J . Org. Chem., Vol. 69, No. 9, 2004