Reaction of arylimodimagnesium with aromatic bifunctional nitriles: Competition and/or cooperation of functional groups affected by electron-accepting ability and location

Article abstract of DOI:10.1002/(SICI)1099-1395(199704)10:4<242::AID-POC901>3.0.CO;2-U

The reactions of aryliminodimagnesium [ArN(MgBr)₂, IDMg] with p′-substituted p-cyanobenzophenones, 1-cyano-9-fluorenone, o-, m- and p-dicyanobenzenes and o-, m- and p-nitrobenzonitriles were examined, and the relative yields of products were referred to the previous results of electron spin resonance studies. The products of condensation with carbonyl and nitro groups and of addition to cyano groups were formed. From the variations of the yields caused by the molar ratio of the magnesium reagent versus substrates and by the substituents of both reactants, the preferred groups were determined. The groups of p- and m-substrates are consistent with the position of highest density of free electrons in the anion radicals, whereas those of the o-substrates are inconsistent The consistency indicates competition of functional groups reflecting the relative ability of single electron acceptance from the reagent, whereas the inconsistency is ascribed to cooperation of neighbouring groups for σ-complexation with the Mg atom of the reagent. The categories of competition and cooperation are discussed in relation to fundamental features and general governing factors proposed for the IDMg reactions of monofunctional substrates.

Full text of DOI:10.1002/(SICI)1099-1395(199704)10:4<242::AID-POC901>3.0.CO;2-U

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)
REACTION OF ARYLIMODIMAGNESIUM WITH AROMATIC
BIFUNCTIONAL NITRILES: COMPETITION AND/OR COOPERATION OF
FUNCTIONAL GROUPS AFFECTED BY ELECTRON-ACCEPTING
ABILITY AND LOCATION†
MASAO OKUBO* MASAYO SAKATA, YUKO IWATSU, NOBUYUKI TSURUSAKI, SAWAKO NAKASHIMA,
YOSHIHITO IWAMOTO, HARUMI NONAKA, AKIRA YAMAUCHI AND KOJI MATSUO
Department of Chemistry, Faculty of Science and Engineering, Saga University, Honjo-machi, Saga 840, Japan
The reactions of aryliminodimagnesium [ArN(MgBr)2 , IDMg] with pЈ-substituted p-cyanobenzophenones, 1-cyano-
9-fluorenone, o-, m- and p-dicyanobenzenes and o-, m- and p-nitrobenzonitriles were examined, and the relative yields
of products were referred to the previous results of electron spin resonance studies. The products of condensation with
carbonyl and nitro groups and of addition to cyano groups were formed. From the variations of the yields caused by the
molar ratio of the magnesium reagent versus substrates and by the substituents of both reactants, the preferred groups
were determined. The groups of p- and m-substrates are consistent with the position of highest density of free electrons
in the anion radicals, whereas those of the o-substrates are inconsistent. The consistency indicates competition of
functional groups reflecting the relative ability of single electron acceptance from the reagent, whereas the inconsistency
is ascribed to cooperation of neighbouring groups for ␴-complexation with the Mg atom of the reagent. The categories
of competition and cooperation are discussed in relation to fundamental features and general governing factors
proposed for the IDMg reactions of monofunctional substrates. © 1997 by John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 10, 242–252 (1997) No. of Figures: 8 No. of Tables: 7 No. of References: 18
Keywords: aryliminodimagnesium; condensation with NO2 and CO groups; addition to CN group; single electron transfer
efficiency; ␴-complexation ability; inter- and intramolecular competition; bifunctional substrates; positional effect of functional
groups.
Received 29 January 1996; revised 7 December 1996; accepted 20 January 1997
INTRODUCTION
advantageous for elucidating factors governing the reactions
of magnesium reagents.1 In this study, IDMg was allowed to
react with some bifunctional aromatic nitriles having CN
and CO, CN and CN, and CN and NO2 groups in an
intramolecular competition manner. The relative yields of
products, indicating which of two groups is preferred, were
consistent or inconsistent with the reported distribution of
free electrons in anion radicals of the bifunctional sub-
strates. The reasons are discussed.
Reactions of aryliminodimagnesium [ArN(MgBr)2 , IDMg]
with aromatic monofunctional nitro, carbonyl and cyano
substrates are controlled by the efficiency of single electron
transfer (SET).1 The distribution of condensation (with NO2
and Ar1CO groups) or addition (to CN group) products
[–N(O)=N–Ar, >C=N–Ar and –C(NH2 )=N–Ar; normal]
and typical radical (dimerization or hydrogen abstraction)
products (abnormal) were correlated with the relative
electron-accepting ability (EAA) of functional groups
(NO2 >CO>CN) regardless of reaction manners. The IDMg
reagent (having an N–Mg bond) has a weak electron-
donating ability (EDA) in comparison with Grignard
reagents (having a C–Mg bond).2 According to the ‘less
reactive, more selective’ principle, the use of mild IDMg is
RESULTS
Intermolecular competition of two kinds of IDMgs with
a benzonitrile and vice versa (Schemes 1a and 1b)
As preliminary experiments, intermolecular competition
using IDMgs and benzonitriles having p-MeO, p-Me and
p-Cl substituents was carried out. Two kinds of nitriles (or
IDMgs) were allowed to react with a kind of IDMg (or
nitrile). Substituent effects on EDA (p-MeO>p-Me>p-Cl)
† Aryliminodimagnesium Reagents, Part XXIX. Part XXVIII; M.
Okubo and Y. Omote, J. Phys. Org. Chem. 9, 212 (1996).
* Correspondence to: M. Okubo, Present address: Grand Forme
Sumaura 401, Ichinotani, Suma-ku, Kobe 654, Japan.
© 1997 by John Wiley & Sons, Ltd.
CCC 0894–3230/97/040242–11 $17.50

M. OKUBO ET AL.
243
Scheme 1a
Table 1a. Yield ratio of two amidines in the reaction in Scheme 1a
Difference in EDA (R1/R2 )
———
EAA
EDA
Rg
Small ←
→ Large
Nit (R3 )
MeO/Me
Me/Cl
MeO/Cl
Weak
l
MeO
Me
Cl
65/35
47/17
63/32
57/42
73/26
61/30
67/30
53/16
55/31
Strong
Scheme 1b
Table 1b. Yield ratio of two amidines in the reaction in Scheme 1b
———
EDA
Rg (R1 )
Nit
Weak ←
→ Strong
Cl
Me
MeO
EAA
Small
Me/MeO 46/36
52/46
71/16
97/67
96/98
96/53
99/75
Difference
in EAA
Cl/Me
64/10
77/13
l
(RR33/R4 )
Large
Cl/MeO
are indicated by the oxidation potentials (Eox ) of the IDMg
and ArMgBr reagents, and the effects on EAA (p-Cl>p-
regardless of their EAA, react preferentially with the IDMg
of stronger EDA. The data of Table 1b indicate a similar
trend that the magnesium reagents, regardless of their EDA,
react preferentially with the nitrile of stronger EAA. The
yield variation caused by the EDA of IDMg (Table 1b)
seems greater than that caused by the EAA of the nitrile
(Table 1a). Even with dilution using 200 ml of THF in some
combinations, no lowering of the yields was observed,
which suggests tight binding of reagent and substrate.
The relation of terms for SET [higher efficiency↔
smaller ⌬E (difference2 between the potentials: ⌬E=
Eox ϪEred )↔stronger EDA and/or EAA] is noted. The non-
competition reaction of nitriles was characterized by the low
SET efficiency,3 i.e. weaker EDA and EAA. The character
[start of reaction from formation of tight ␴-complex
(coordination of N atom of CN group to the ␦+ charged Mg
Me>p-MeO) are indicated by the reduction potentials (Ered
)
of the NO2 and Ar1CO substrates.2 At a concentration of
5·0 mmol IDMg in 50 ml of THF, the reactions were carried
out at 55 °C for 1 h. Two moles of IDMg (in total) were
used in all cases in order to obtain the amidines in sufficient
yields for comparison of the effects of substituents, because
no effect was observable by the use of 3 mol, leading
always to excellent yields.3c The yields of the two amidines
obtained in the reactions of two kinds of IDMgs with a
nitrile are given in Table 1a and yields (total 200%)
obtained in the reactions of a substituted IDMg with two
kinds of nitriles in Table 1b. No abnormal product was
detected.
The data in Table 1a indicate a trend that the nitriles,
© 1997 by John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)

244
REACTION OF ARYLIMINODIMAGNESIUM WITH AROMATIC NITRILES
Scheme 2
Table 2. Yields of products 1–3 in the reaction in Scheme 2a
Yield/%
Run
No.
Molar Temperature
Recovery
(%)
Hydrolysis
(%)
R1
R2
ratio
(°C)
1
2
3
1
2
3
4
5
6
7
MeO MeO
MeO MeO
2·5
2·5
2·5
2·5
1·0
2·5
2·5
55
r.t.
55
r.t.
55
55
r.t.
Trace
Trace
—
6
—
38
58
38
41
—
33
—
58
19
55
Trace
—
54
—
4
—
—
—
—
—
—
—
Me
Me
Me
Cl
MeO
MeO
MeO
Meo
MeO
—
38
92
—
97
Trace
Trace
Cl
—
8
9
MeO Me
MeO Me
MeO Me
2·5
2·5
2·0
2·5
2·5
2·0
2·5
2·5
55
r.t.
55
55
r.t.
55
55
r.t.
—
16
4
—
Trace
—
32
56
9
36
65
42
52
54
13
72
62
15
56
47
Trace
—
13
10
—
—
—
—
78
6
—
—
—
—
—
—
—
10
11
12
13
14
15
Me
Me
Ne
Cl
Me
Me
Ne
Me
Me
—
Cl
Trace Trace
16
17
18
19
MeO Cl
MeO Cl
2·5
2·5
2·5
2·5
55
r.t.
55
r.t.
Trace
—
—
14
30
39
—
72
18
52
—
—
41
—
94
9
10
—
—
Cl
Cl
Cl
Cl
—
a Reaction mixture were stirred for 3 h in THF. Ered of substrate, R2 =MeOϪ1·646 V: R2 =MeϪ1·629 V:
R2 =ClϪ1·552 V.
atom, as governing step) followed by SET inside the
complex3 ] is supported by the trends given above, and will
be referred to in the following discussion.
and p-nitrobenzophenones6 have the highest free electron
(spin) density on the NO2 groups. Diphenyl diketone
(benzil) reacts via the generation of the anion radical of an
Mg chelate (very tight ␴-complex),7, 8 implying that two CO
groups cooperate. It should be noted that two groups
compete or cooperate depending on their relative EAA and
location.
Intramolecular competition and/or cooperation of
cyano, carbonyl and nitro groups
Some IDMg reactions with bifunctional nitro and carbonyl
compounds have been examined previously. The two NO2
groups of dinitrobenzenes compete and lead mainly to self-
decomposition via vigorous SET.4 The NO2 and CO groups
of p-nitrobenzophenone compete, and the former group of
stronger EAA is preferred.4 The involvement of SET is
supported by comparison with previous ESR results: the
anion radicals of m- and p-nitroacetophenones5 and o-, m-
The following discussion on the features of IDMg
reactions with bifunctional substrates having at least one
CN group will be based on the reported result5 that the anion
radicals of m-and p-cyanoacetophenones have the highest
spin density on the CO group. Undoubtedly from all the
ESR results cited,5, 6 the spin density in anion radicals of
nitrobenzonitriles, although it is not reported, must be
highest on the NO2 group.
© 1997 by John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)

M. OKUBO ET AL.
245
Scheme 3
Table 3. Yields of products in the reaction in Scheme 3
Yield (%)
The discussion will also be referred to the known facts
and features of general meaning on IDMg reactions:8–14
determination of relative EAA of substrates the Ered values
of which are unobservable,9 basic character of IDMg
reaction with diaryl ketones,10 nuclear substitution and/or
decomposition of substrates of strong EAA,8, 11, 12 general
need for aggregate excess moles of the magnesium reagents
(IDMg and RMgX) for product formation,13 action and role
of hexamethylphosphoramide (HMPA) (having a very
strong coordinating ability) added (in a calculated amount)
to IDMg solution13 and oxidation and decomposition of
IDMg molecule in reactions of efficient SET.14
Run
No.
Temperature
(°C)
Recovery
(%)
R1
R2
1
2
3
1
2
3
4
5
6
7
Me MeO
Me MeO
Cl MeO
Me Me
Me Me
Me Cl
55
r.t.
55
55
r.t.
55
r.t.
52 Trace 41
—
23
—
—
44
—
62
58
59
8
8
—
30
26 Trace 65
53
30
36
—
—
—
—
68
—
Me Cl
On the basis of consistency and/or inconsistency in the
position of highest spin density with the most reactive site,
a set of two categories of the positional effects of functional
a The reaction mixture was stirred for 3 h in THF.
Scheme 4
Table 4. Yields of products in the reaction in Scheme 4a
Yield (%)
Run
No.
Molar Temperature
Recovery
(%)
ratio
(°C)
1fl
2fl
3fl
4fl
1
2
3
2·5
2·5
1·5
3·0
3·0
55
r.t.
55
55
r.t.
31
—
—
53
10
—
—
10
6
36
6
—
18
trace
21
—
—
—
—
11
93
20
6
4 (NH)
5 (NH)
—
82
a The reaction mixture was stirred for 3 h.
© 1997 by John Wiley & Sons, Ltd. JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)

246
REACTION OF ARYLIMINODIMAGNESIUM WITH AROMATIC NITRILES
Scheme 5
Table 5. Yields of products in the reaction in Scheme 5.a
Yield (%)
Overall
yield
(%)
Run
No.
Ar=RC6H4 Molar
Substrate
R
ratio
5
6
7
8
1
2
3
4
5
o-
o-
o-
o-
o-
p-Meo
p-Me
p-Me
p-Me
p-Cl
4·0
4·0
2·0
1·0
4·0
—
—
—
—
—
—
—
—
—
—
—
—
46
64
78
100
92
49
13
17
100
92
95
77
95
6
7
8
9
10
11
m-
m-
m-
m-
m-
m-
p-MeO
p-MeO
p-Me
p-Me
p-Me
p-Cl
4·0
6·0
4·0
2·0
1·0
4·0
59 40
57 34
76 21
52 41
—
—
—
—
—
—
—
—
—
—
—
—
99
91
97
93
41
93
40
1
67 26
12
13
14
15
16
17
p-
p-
p-
p-
p-
p-
p-MeO
p-MeO
p-Me
p-Me
p-Me
p-Cl
4·0
6·0
4·0
2·0
1·0
4·0
44 40
19 37
54 46
62 27
—
—
—
—
—
—
—
—
—
—
—
—
84
56
100
89
95
85
94
1
45 40
a Concentration of IDMg=200 mmol 1. b The reaction mixture was stirred for 1 h at 55 °C
in THF. Ered: p-, Ϫ1·802 V; o-, Ϫ1·881 V; m-, unobservable.
Scheme 6
groups originating from fundamental nature of the SET
governance in IDMg reactions is proposed in the final
section.
R1 =R2 =MeO, Me, Cl) at 55 °C and at room temperature
(Scheme 2). The products are an amidine (1) formed via
IDMg addition to CN group, an anil (2) formed via IDMg
condensation with CO group and an anil-amidine (3)
formed via addition–condensation. As Table 2 (runs 1–19)
shows, the yields of 1–3 vary according to the substituents
of the reactants and the reaction conditions.
CO and CN groups (Schemes 2–4)
pЈ-R2-substituted p-cyanobenzophenones were allowed to
react with p-R1C6H4-IDMg (mainly a 2·5 molar amount;
The effects of molar ratio and temperature are shown in
© 1997 by John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)

M. OKUBO ET AL.
247
Scheme 7
Table 6. Yields of products in the reaction in Scheme 7
Yield (%)
Run
No.
Molar
ratio
Temperature
Recovery
(%)
R
HMPA
(°C)
9
10
11
12
13
1
2
3
4
MeO
MeO
MeO
MeO
2·0
2·0
4·0
3·0
0
0
0
0
63
68
38
34
3
9
9
8
9
6
17
—
—
—
—
—
4
6
6
9
4
—
5
0a
0
ϫ
r.t.
—
5
6
7
Me
Me
Me
2·0
4·0
3·0
0
0
ϫ
0
0
r.t.
75
54
69
Trace
Trace 14 Trace
5
11
—
3
1
10
—
—
—
—
—
8
9
10
11
12
Cl
Cl
Cl
Cl
Cl
2·0
4·0
2·0
4·0
3·0
0
0
0
0
ϫ
0
0
55
55
r.t.
17 Trace
15 Trace
3
2
8
10
13
5
4
—
1
69
62
—
—
—
Trace
63
54
23
Trace
Trace 12
19
2
2
5
—
a 0=Present; ϫ=absent.
b Equimolar amount of HMPA based on Mg.
runs 3–5, 8–10 and 11–13. Equimolar IDMg leads to no
product even at 55 °C, and more than 2 mol are needed for
a sufficient yield.13 With the use of 2·5 mol, the yield of 2
plus 3 predominates over 1 (formed in trace or small
amounts) while the formation of 3 is depressed at room
temperature. From comparison of the yields of 2 and 3 in
runs 1–18 (at 55 °C and room temperature), it is clear that
the CO group (more strongly electron-accepting although
sterically more crowded than the CN group) is preferred to
give 2. The generated product 2 having CN and arylimino
groups must have a weaker EAA than the starting CO–CN
substrate (suggested from the reported examples2b, 9 ), and
undergoes efficient IDMg addition to the CN group to give
3. The poor yield of 1 seems slightly improved at room
temperature (runs 4 and 9), probably because a lower
temperature promotes aggregation of IDMg molecules to
favour ␣-complexation with the CN group.
ambiguous manner by the substituent of the substrate, but in
a clear manner by the substituent of IDMg. The weaker p-Cl
IDMg2 (Runs 7, 15 and 19) leads to a high recovery at room
temperature, while the stronger p-MeO IDMg (runs 8, 16
and 17) gives benzhydrol (diarylmethanol, in a small
amount) via hydrogen abstraction from solvent molecules
by the ketyl radical, generated via efficient SET due to the
apparently small Ered (see footnote to Table 2).
Undoubtedly, the condensation with the CO group
precedes the addition to the CN group. Exclusive formation
of 1 is expected with the use of anilinomagnesium
(ArNHMgBr), which is capable of addition to CN group but
incapable of condensation with CO group.3, 10 By the use of
2·5 mol (Table 3, Scheme 3), the yield of 1 was improved in
the expected manner, but the formation of 2 and 3 was,
unexpectedly, not excluded. The reason is given later.
In the reaction of 1-cyano-9-fluorenone [Scheme 4
(R1 =Me), Table 4], the expected 1fl and 3fl were formed. In
As shown in Table 2, the relative yields are affected in an
© 1997 by John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)

248
REACTION OF ARYLIMINODIMAGNESIUM WITH AROMATIC NITRILES
addition, product 4fl of nuclear substitution at the 8-position
was formed, reflecting the efficient SET8, 11, 12 [due to the
apparently small Ered (Ϫ1·218V)]. Distinct from the data of
Table 2, the higher yield of 1fl and no formation of 2fl in run
1 (using 2·5 mol of IDMg) clearly implies a preference for
the CN group. The reason is discussed later.
spin densities than the CN group.5 This is not related to the
detailed distribution of products, but may represent the
weak EAA of the CN group responsible for the notably mild
characters of the reactions in Scheme 13 and also Scheme
5.
NO2 and CN groups; effect of added HMPA (Scheme 7)
Two CN groups (Scheme 5, Table 5)
The IDMg reaction of p- and m-dicyanobenzenes leads to
mono- and bisamidines (5 and 6). The o-isomer (phthalodi-
nitrile) gives the arylimino derivative of cyclic
benzopyrrolidindione (7) and its N-aryl derivative (8). The
formation of 7 and 8 (runs 1–5) may be initiated by IDMg
addition (single and/or double): cyclization accompanies
the release of NH3 unit and/or hydrolysis of the precursors
(Scheme 6). The formation of symmetrical azobenzenes via
oxidative coupling of IDMg is negligible, implying that the
reactions are mild.
As Table 5 shows, the product yields vary with the molar
ratio of the reactants: variations in the yields of 5 and 6
formed from the p- and m-substrates (Ered of the latter is
unobservable but may be larger than 2.0 V3 ) are first noted.
The data for runs 6, 8, 11, 12, 14 and 17 indicate that 4 mol
of IDMg (more than 2 mol of IDMg per one CN group) are
usually needed for efficient consumption of CN to give
sufficient overall yields. However, the data for run 13
indicate that the use of 6 mol of IDMg causes a lowering of
the yield. Therefore, a suitable excess of reagent is needed
to assist the formation of the normal product,13 but too large
an excess also causes an efficient SET giving the abnormal
product (frequently seen in Grignard and IDMg reac-
tions10, 14 ) via a process other than radical dimerization,
hydrogen abstraction from solvent or radical substitution on
the aromatic ring (cf. 4fl ). Most probably, the abnormal
product is an aromatic hydrocarbon formed via removal of
an anion (CNϪ , XϪ or NO2Ϫ ) from the substrate’s anion
radical (see Wittig reaction) to leave a neutral hydrocarbon
radical abstracting hydrogen from the solvent: the lower
boiling hydrocarbon is lost on treatment of mixture. The
term ‘abnormal’ means the product coming from one of the
pair of substrate and reagent.
p-Dinitrobenzene has a small negative Ered (Ϫ1·003 V) and
gives (via a too efficient SET) unidentifiable black materials
in large amounts and the expected bisazoxy and bisazo
products in poor yields.4, 14 However, in spite of the
apparently much smaller Ered (Ϫ0·886 V), the NO2 group of
p-nitrobenzonitrile reacts efficiently to give the expected
products in fair yields (probably the CN group is ␴-
complexed with IDMg, changing the electron distribution of
the anion radical): cyanoazoxy (9; condensation), cyanoazo
(10; condensation–deoxygenation), amidinoazoxy (11; con-
densation–addition), 3-anilino (12; nuclear substitution)
and symmetrical azo (13; oxidative coupling14 ) products.
Owing to interest in the physical properties of pЈ-alkoxy-p-
cyanoazoxybenzene,15 improvement in the synthesis of 9 by
using HMPA was examined.
By addition of HMPA (2 mol based on Mg) to IDMg
solution in THF, the reaction with nitrobenzenes of strong
EAA was modified. Although the coordination of HMPA
must cause elevation of the HOMO level of IDMg to make
the SET much more efficient, the yield of the normal
condensation product was improved.13 Therefore, it is
proposed that, by exchange of two THF ligands on Mg with
HMPA, its 3d-orbitals must facilitate an electronic field
feasible for the later step succeeding SET, i.e. transfer of
radicals. Therefore, the method was applied to the reaction
with the p-NO2–CN substrate, the yields in the presence and
absence of HMPA being given in Table 6 (runs 1–12). By
comparison of the yields in runs 1 and 2 with those in run
4, the yields in run 5 with those in run 7 and the yields in
runs 10 and 11 with those in run 12, it is found that, in the
presence of HMPA, the yields of 9 and 11 (normal products)
are considerably and slightly increased, respectively, and
the yields of 10 and 13 are slightly decreased.
Among structural effects of the reagent and substrate,
exclusive formation of 8 from the o-substrate and the p-
MeO and p-Me IDMg (runs 1 and 2) is notable (see later
discussion). The behaviours of the p- and o-substrates
towards p-MeC6H4NHMgBr (4 mol) are distinct from those
towards IDMg. The p-substrate gives the double addition
product 6 almost quantitatively (99%, with 1% of 5),
implying that ArNHMgBr of stronger EDA is preferable to
IDMg.2a The o-substrate gave 7 and 8 in lower and
comparable yields (42% and 46%), implying that, in
comparison with IDMg, ArNHMgBr is less favourable for
aggregation (and ␴-complexation)13 because of the half
equivalent of the Mg atom.
In the reactions with the o-substrate (Ererd =Ϫ0·994 V),
the azoxy compound corresponding to 9 is the major
product (32–50% yields) in the absence of HMPA. Even in
the presence of HMPA, the yields are only slightly
improved (5–8%); the reason for the difference from the
case of p-substrate is discussed in the next section.
SURVEY OF POSITIONAL EFFECTS
The positional effects of functional groups observed in the
reactions in Schemes 2–5 and 7, are surveyed below.
The IDMg reaction in Scheme 2 with the p-CO–CN
substrate takes place predominantly at the CO group,
reflecting its stronger EAA compared with the CN group. In
In free anion radicals of benzonitrile and o-, m- and p-
dicyanobenzenes, ring carbon atoms have slightly higher
© 1997 by John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)

M. OKUBO ET AL.
249
Figure 1. Illustration: relative degree of participation of efficiency of SET and ␴-complexation for IDMg and Grignard reactions with
monofunctional substrates. Given bifunctional compounds accompanied by short arrows indicate, for comparison, approximate ⌬E values for
the present IDMg reactions
the reaction in Scheme 3 using ArNHMgBr, part of it
condenses with a CO group. The reagent has a stronger
EDA than IDMg,2 and probably disproportionates via SET
induction to generate ArN(MgBr)2 and ArNH2 species: the
former species is responsible for the formation of 2 and 3.
In the reaction in Scheme 4 using a sufficient excess of
IDMg, in spite of the apparently smaller Ered of the 9-CO–
1-CN substrate, addition to the 1-CN group is favoured over
condensation with the 9-CO group. The reason for the
apparent deviation from the order of relative EAA of groups
(see Scheme 2) is as follows. With use of the likely oriented
lone pairs of electrons, the two groups cooperate to gather
IDMg molecules to aggregate and assist ␴-complexation
with the less crowded CN group. Thus, the ␴-complexation
must be fairly tight and, accordingly, perturbs the HOMO–
LUMO energy levels to make possible the inner-sphere
SET. The complexation plays an essential role,3, 13 i.e.
overcomes the weak resonance demand of the CN group
notable that the o-substrate requires only 2 mol of IDMg
(except for the weak p-Cl reagent) per CN group for perfect
double addition; this is distinct from the reaction with
ordinary benzonitrile, requiring 3 mol.3 The apparently
strong EAA seems responsible, but cooperation of neigh-
bouring CN groups similar to the case in Scheme 4 must be
more effective. The ArNHMgBr reacts almost perfectly
with the p-CN–CN substrate but inefficiently with the
o-CN–CN substrate; a separate dependence on the stronger
EDA or on the half equivalent of Mg atoms is noted.
In the IDMg reactions in Scheme 7 with the NO2–CN
substrates, the NO2 group predominates over the CN group
not only at p- but also at o-positions, i.e. both depend simply
on the relative EAA: this arises from the great difference in
the EAA of the groups. By comparison with the known
effect of added HMPA,13 the reaction with the p-substrate is
modified as expected but, unexpectedly, the reaction with
the o-isomer is poorly modified. The Mg atom of IDMg,
reacting with the NO2 group, is tightly flanked by bulky
molecules of HMPA: the 3d-orbitals must be crowded with
the o-CN group, and may be in an orientation less available
for radical transfer.
(see the small ␴ value13c ). The energetically less favourable
␲
reaction with the CN group (than that with the CO group)
becomes predominant.
In the reaction in Scheme 5 of CN–CN substrates, it is
© 1997 by John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)

250
REACTION OF ARYLIMINODIMAGNESIUM WITH AROMATIC NITRILES
Table 7. Melting points and 1H NMR data for products
Product
No.a
M.p. (°C)
␦ (ppm) in CDCl3
MeO
1
1
1
1
1
1
1
177·0–178·9 7·89–7·62 (4H, m, ArH), 7·20 (1H, d, ArH), 7·01–6·60 (7H, m, ArH), 4·65 (2H, s), 3·89 (6H, d)
150·0–151·8 8·01–7·80 (6H, M, ArH), 7·20 (2H, d, ArH), 6·99–6·89 (4H, M, ArH), 4·90 (2H, s), 3·90 (3H, s), 2·35 (3H, s)
190·0–193·0 7·95, 7·35 (4H, ABq, J=8·79 Hz), 7·85 (4H, d, ArH), 7·01–6·94 (4H, m, ArH), 4·91 (2H, s), 3·90 (3H, s)
MeO
MeO
Me
MeO
Cl
Me
Oil
8·11 (2H, s), 7·76–7·63 (3H, m, ArH), 7·54 (1H, d, ArH), 7·28–6·60 (8H, m, ArH), 3·76 (3H, s), 2·44 (3H, s)
MeO
Me
150·0–151·8 7·96–7·68 (6H, m, ArH), 7·29–7·15 (4H, m, ArH), 6·87 (2H, d, ArH), 4·98 (2H, s), 2·44 (3H, s)
190·0–193·0 7·95–7·82 (4H, m, ArH), 7·70 (2H, d, ArH), 7·31 (4H, m, ArH), 6·95 (2H, d, ArH), 3·66 (2H, s), 2·45 (3H, s)
155·0–156·9 8·00 (2H, m, ArH), 7·82–7·74 (4H, m, ArH), 7·50 (2H, m, ArH), 7·18 (2H, m, ArH), 6·92 (2H, d, ArH), 4·91
(2H, s), 2·35 (3H, s)
128·5–130·0 7·84–7·56 (4H, m, ArH), 7·25 (1H, d, ArH), 7·03–6·57 (7H, m, ArH), 3·85 (6H, q)
164·0–165·0 7·84–7·54 (4H, m, ArH), 7·25–6·53 (8H, m, ArH), 3·90 (3H, t), 2·20 (3H, d)
124·8–126·5 7·85–7·64 (4H, m, ArH), 7·23–6·56 (8H, m, ArH), 3·80 (3H, d)
146·0–147·3 7·85–6·56 (8H, m, ArH), 6·71–6·60 (4H, m, ArH), 3·75 (3H, d), 2·41 (3H, d)
164·0–165·0 7·84–7·54 (3H, m, ArH), 7·27–6·89 (6H, m, ArH), 6·62 (3H, m, ArH), 2·41 (6H, d)
124·8–126·5 7·58–7·54 (4H, m, ArH), 7·25–7·10 (5H, m, ArH), 7·96 (1H, d, ArH), 6·60 (2H, m, ArH), 2·34 (3H, d)
167·0–169·0 7·87–7·22 (8H, m, ArH), 7·06 (1H, d, ArH), 6·75–6·58 (3H, m, ArH), 3·76 (3H, d)
177·0–178·3 7·84–7·02 (12H, m, ArH)
Me
Me
Cl
Cl
Me
MeO
2
2
2
2
2
2
2
2
MeO
MeO
Me
MeO
Cl
Me
MeO
Me
Me
Me
Cl
Cl
Meo
Cl
Cl
MeO
3
Oil
7·84–7·64 (4H, m, ArH), 7·20 (2H, d, ArH), 7·05 (1H, d, ArH), 6·91–6·88 (5H, m, ArH), 6·78 (1H, d, ArH), 6·71
(1H, s, ArH), 6·67 (2H, s, ArH), 3·80 (9H, q)
MeO
MeO
3
3
3
3
153·0–154·5 7·93–7·64 (4H, m, ArH), 7·21–6·58 (12H, m, ArH), 4·84 (2H, s), 3·89 (3H, d), 2·30 (6H, t)
176·0–177·3 7·85–7·64 (3H, m, ArH), 7·33–6·61 (9H, m, ArH), 4·87 (2H, s), 3·38 (3H, d)
149·8–151·6 7·83–7·56 (12H, m, ArH), 7·18–6·65 (4H, m, ArH), 4·92 (2H, s), 3·76 (6H, t), 2·38 (3H, d)
153·0–154·5 7·91–7·76 (3H, m, ArH), 7·61 (1H, d, ArH), 7·21–6·85 (10H, m, ArH), 6·63 (2H, m, ArH), 4·89 (2H, s), 2·33
(9H, t)
Me
MeO
Cl
Me
MeO
Me
Me
Me
Cl
3
176·0–177·3 7·90–7·70 (3H, m, ArH), 7·60 (1H, d, ArH), 7·34–6·87 (10H, m, ArH), 6·65 (2H, m, ArH), 4·90 (2H, s), 2·40
(3H, d)
Cl
3
3
3
159·8–160·4 7·88–7·64 (4H, m, ArH), 7·38–6·91 (8H, m, ArH), 6·74–6·65 (4H, m, ArH), 4·85 (2H, s), 3·80 (6H, q)
151·0–152·0 7·98–6·84 (14H, m, ArH), 6·62 (2H, d, ArH), 4·83 (2H, s), 2·30 (6H, t)
181·0–183·5 7·92–7·65 (4H, m, ArH), 7·39–6·89 (10H, m, ArH), 6·63 (2H, d, ArH), 4·84 (2H, s)
182·0–184·1 8·26 (1H, s, ArH), 7·67–7·53 (3H, m, ArH), 7·35–7·26 (5H, m, ArH), 6·95 (2H, s, ArH), 6·46 (2H, s), 3·82 (3H, s)
169·0–169·5 8·30 (1H, s, ArH), 7·67–7·52 (5H, m, ArH), 7·34–7·24 (1H, m, ArH), 7·20 (2H, d, ArH), 6·92 (2H, d, ArH), 6·46
(2H, s), 2·34 (3H, s)
MeO
Cl
Me
Cl
Cl
1flMeO
1flMeO
1flCl
Oil
Oil
8·20 (1H, s, ArH), 7·67–7·53 (4H, m, ArH), 7·34–7·30 (4H, m, ArH), 6·95 (2H, d, ArH)
8·24 (1H, d, ArH), 7·78–7·28 (6H, m, ArH), 7·08–6·93 (4H, m, ArH), 3·81 (3H, d)
2flMeO
2flMe
2flCl
187·0–189·3 7·81 (1H, d, ArH), 7·66–7·21 (4H, m, ArH), 7·07 & 6·85 (4H, Abq, J=7·8 Hz), 6·93 (1H, d, ArH), 2·42 (3H, d)
225·0–227·8 7·82 (1H, d, ArH), 7·67–7·51 (3H, m, ArH), 7·44–7·37 (3H, m, ArH), 7·10–6·95 (3H, m, ArH), 6·81 (1H, d, ArH)
186·0–187·8 8·06 (1H, s, ArH), 7·67–7·31 (4H, m, ArH), 6·97–6·83 (8H, m, ArH), 6·70 (2H, d, ArH), 3·86 (3H, s), 3·76 (3H, s)
3flMeO
3flMe
Oil
8·17 (1H, s, ArH), 7·70–7·50 (3H, m, ArH), 7·37–7·23 (3H, m, ArH), 7·20 (2H, d, ArH), 7·11 (2H, d, ArH),
6·97–6·84 (2H, m, ArH), 6·65 (2H, d, ArH), 2·41 (3H, s), 2·30 (3H, s)
3flCl
196·0–197·5 7·87 (1H, s, ArH), 7·69–7·49 (3H, m, ArH), 7·41–7·21 (5H, m, ArH), 7·10–6·87 (5H, m, ArH), 6·65 (1H, d, ArH)
129·0–130·0 7·84–7·36 (6H, m, ArH), 7·06–6·93 (4H, m, ArH), 3·87 (3H, s)
162·5–163·8 7·84 (1H, d, ArH), 7·68 (1H, d, ArH), 7·57 (1H, d, ArH), 7·24 (1H, d, ArH), 7·10 (1H, d, ArH), 6·95 (1H, m, ArH),
6·63 & 6·53 (4H, Abq, J=8·3 Hz), 2·33 (3H, q)
4flMeO
4flMe
m-
MeO
5
Oil
Oil
Oil
8·18 (1H, s, ArH), 8·09 (1H, d, ArH), 7·74 (1H, d, ArH), 7·55 (1H, t, ArH), 6·91 (4H, s, ArH), 4·94 (2H, b), 3·81
(3H, s)
8·19 (1H, s, ArH), 8·10 (1H, d, ArH), 7·73 (1H, d, ArH), 7·55 (1H, t, ArH), 7·18, 6·86 (4H, Abq, J=8·3 Hz), 4·92
(2H, b), 2·34 (3H, s)
m-
Me
5
m-
Cl
5
8·16 (1H, s, ArH), 8·08 (1H, d, ArH), 7·74 (1H, d, ArH), 7·54 (1H, t, ArH), 7·32, 6·89 (4H, Abq, J=8·3 Hz), 5·05
(2H, s)
p-
5
5
5
151·5–152·0 8·00, 7·74 (4H, ABq, J=8·3 Hz), 6·92 (4H, s, ArH), 4·92 (2H, b), 3·81 (3H, s)
144·0–145·5 8·01, 7·75 (4H, ABq, J=8·79 Hz), 7·19 & 6·87 (4H, ABq, J=8·3 Hz), 4·87 (2H, b), 2·34 (3H, s)
159·0–160·0 7·99, 7·76 (4H, ABq, J=8·3 Hz), 7·35 & 6·91 (4H, ABq, J=8·79 Hz), 4·89 (2H, b)
175·0–176·5 8·67 (1H, s, ArH), 8·21 (2H, d, ArH), 7·75 (1H, t, ArH), 7·25 & 7·02 (8H, ABq, J=8·79 Hz), 3·84 (6H, s),
4·76–2·54 (4H, b)
MeO
p-
Me
p-
Cl
m-
b
6
6
6
MeO
m-
Me
92·0–94·0
8·36 (1H, s, ArH), 7·95 (2H, d, ArH), 7·45 (1H, t, ArH), 7·17, 6·90 (8H, ABq, J=8·3 Hz), 5·49–4·51 (4H, b), 2·34
(6H, s)
m-b
243·5–244·5 8·39 (1H, s, ArH), 8·04 (2H, d, ArH), 7·53 (1H, t, ArH), 7·34, 6·94 (8H, ABq, J=8·3 Hz), 5·33 (4H, b)
223·0–224·5 8·00 (4H, s, ArH), 6·92 (8H, s, ArH), 6·68–6·14 (4H, b), 3·57 (6H, s)
Cl
b
6pMeO
© 1997 by John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)

M. OKUBO ET AL.
251
Table 7. Continued
Product
No.a
M.p. (°C)
␦ (ppm) in CDCl3
p-
6
266·0–268·0 7·95 (4H, s, ArH), 7·18 & 6·90 (8H, Abq, J=8·3 Hz), 4·88 (4H, b), 2·34 (6H, s)
269·0–270·0 7·36 (4H, s, ArH), 7·32, 6·95 (8H, ABq, J=8·3 Hz), 5·40 (4H, b)
181·0–186·0 8·40 (1H, s), 7·85 (1H, d, ArH), 7·78 (1H, d, ArH), 7·55 (2H, t, ArH), 7·22, 7·09 (4H, ABq, J=8·3 Hz), 2·29
Me
p-b
6
Cl
o-
b
7
Me
(3H, s)
o-
Cl
o-
c
7
8
8
8
—
7·97 (1H, m, ArH), 7·62 (2H, t, ArH), 7·50 (1H, d, ArH), 7·36 (1H, s), 7·38, 7·02 (4H, ABq, J=8·79 Hz)
8·03, 7·67 (4H, ABq, J=8·79 Hz), 6·93 (8H, m, ArH), 3·83 (6H, t)
80·0–81·5
MeO
o-
141·1–141·5 8·03 (2H, d, ArH), 7·68 (2H, d, ArH), 7·14, 6·91 (8H, ABq, J=8·3 Hz), 2·36 (6H, s)
159·0–160·0 7·99 (2H, t, ArH), 7·63 (2H, m, ArH), 7·37, 7·03 (8H, ABq, J=8·3 Hz)
203·0–204·0 8·34, 7·83 (4H, Abq, J=9·28 Hz), 8·42, 7·02 (4H, ABq, J=8·79 Hz), 3·90 (3H, s)
154·0–154·5 8·43, 7·85 (4H, Abq, J=9·28 Hz), 8·17, 7·33 (4H, ABq, J=8·3 Hz), 2·44 (3H, s)
Me
o-
CL
9MeO
9Me
9Cl
168·0–169·0 8·21, 7·86 (4H, Abq, J=9·27 Hz), 8·43, 7·46 (4H, Abq, J=8·79 Hz)
c
10MeO
11MeO
11Me
—
7·93, 7·81 (4H, Abq, J=8·79 Hz), 7·93, 7·01 (4H, Abq, J=8·79 Hz), 3·91 (3H, s)
170·5–171·0 8·39 (4H, d, ArH), 8·04 (2H, d, ArH), 6·99 (2H, d, ArH), 6·95 (4H, s, ArH), 4·92 (2H, b), 3·90 (3H, s), 3·82 (3H, s)
167·0–168·0 8·37 (2H, d, ArH), 8·17 (2H, d, ArH), 8·05 (2H, d, ArH), 7·19 (2H, d, ArH), 7·32, 6·92 (4H, Abq, J=8·3 Hz), 4·89
(2H, b), 2·43 (3H, s), 2·35 (3H, s)
11Cl
12Cl
185·0–186·0 8·37 (2H, d, ArH), 8·04 (2H, d, ArH), 7·49 (2H, d, ArH), 7·36 (2H, d, ArH), 8·20, 6·96 (4H, ABq, J=8·79 Hz),
4·92 (2H, b)
168·5–169·5 9·41 (1H, s), 8·29 (1H, d, ArH), 7·38 (1H, s, ArH), 7·45, 7·26 (4H, ABq, J=8·79 Hz), 7·03 (1H, d, ArH)
a Superscripts indicate substrate; fl=1-cyano-9-fluorenone, o-, m-, p-=o-dicyano-, m-dicyano, p-dicyanobenzene. Subscripts indicate IDMg and anilinomagnesium
(Ar-NHMgBr).
b in DMSO-d6 .
c Undetermined.
GENERAL DISCUSSION AND CONCLUSION
of functional groups. Scheme numbers are given in square
brackets below.
The Mg2+ ion has a small ionic radius, and thus has a very
strong polarizing ability. This is reflected in the relative
EDA of magnesium reagents (CMg>NMg>OMg)2 accord-
ing to the electronegativities, the known donative effects of
lone pair electrons of N and O atoms (combined with carbon
atoms) being eliminated. This specific nature of magnesium
reagents is responsible for the SET-governed features of
reactions in general, i.e. the product distribution depends
principally on the combination of EDA and EAA (relative
value of ⌬E). Moreover, the polarized NϪ —Mg+ bond
allows lone pair electrons of functional groups to coordinate
with the Mg atom (i.e. ␴-complexation). Participation of
this factor causes modification of the reaction.1
The concept of relative participation of SET and ␴-
complexation applied to reactions of IDMg and Grignard
reagents with monofunctional substrates (concerning the
classification into B and C1 ) is illustrated in Figure 1. As a
reference, the apparent ⌬E values for bifunctional cases
[estimated from Eox and Ered (of the whole molecule)] are
indicated by short and broad arrows. However, the role of
the relative EAA of individual groups of the substrates used
is of great significance (see below).
The category of competition represents the ordinary
contribution of the EAA of individual groups and the EDA
of reagents when the groups are located apart from each
other. The category includes the reported case of the p-
NO2–CO substrate4 and the present cases of p-CO–CN [2]
and p-NO2–CN substrates [7]. The EAA–EDA dependence
of the behaviour of the p-CO–CN substrate [3] causing
suggested disproportionation of ArNHMgBr is an example
of the same category. Also included is the behaviour of the
p-CN–CN substrate [5] reflecting the EDAs of two
reagents,2 i.e. reacting sluggishly with IDMg and efficiently
with ArNHMgBr.
The category of cooperation implies an extraordinary
participation of ␴-complexation, overwhelming the effect of
the difference (not so great) in the EAA of two groups
located as neighbours. This category, caused by access of
reagent, includes the reported case of an ␣-CO–CO
substrate,7 and the present case of the 9-CO–1-CN substrate
[4]. In contrast to the p-CN–CN substrate, the o-isomer [5]
is converted easily by IDMg but sluggishly with
ArNHMgBr, the behaviour depending on the number of Mg
atoms as the accepting site of ␴-complexation is of the same
category. The behaviour of the o-NO2–CN substrate [7] is
outside this category owing to the too great difference in
EAA. This o-substrate undergoes a novel type of steric
hindrance causing depression of the HMPA effect.
The consistency in the positions of highest spin density
with the most reactive site implies that the SET process is
involved in the IDMg reaction with bifunctional nitriles.
Owing to the mild natures of the CN group and IDMg
reagent (in spite of the apparently strong EAA of the whole
molecule of bifunctional nitriles), positional effects are
observed and categorized into competition and cooperation
In conclusion, the concept of the relative participation of
SET and ␴-complexation1 can be reasonably applied, with a
small modification, to cases of bifunctional substrates. It is
© 1997 by John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)

252
REACTION OF ARYLIMINODIMAGNESIUM WITH AROMATIC NITRILES
and K. Matsuo, Rev. Heteroatom Chem. 10, 213–249 (1994).
2. (a) M. Okubo, T. Tsutsumi, A. Ichimura and T. Kitagawa, Bull.
Chem. Soc. Jpn. 57, 2679 (1984); (b) M. Okubo, T. Tsutsumi
and K. Matsuo, Bull. Chem. Soc. Jpn. 60, 2085 (1987).
3. M. Okubo, K. Matsuo, N. Tsurusaki, K. Niwaki and M. Tanaka,
J. Phys. Org. Chem. 6 509 (1993). The Ered of Ar1CN is located
outside the potential window of Bu4NClO4–THF, and an
approximate value was suggested from the value in DMF.
4. M. Okubo, H. Aratani, T. Gondo and K. Koga, Bull. Chem. Soc.
Jpn. 56, 788 (1983).
5. P. H. Rieger and G. K. Fraenkel, J. Chem. Phys. 40, 2795, 2811
(1963). No anion radical of nitrobenzonitrile was measured,
but the highest free spin density on the NO2 group is certain.
6. K. Maruyama and T. Otsuki, Bull. Chem. Soc. Jpn. 41, 444
(1968).
worthwhile using this concept as a general guiding principle
for the study of reactions of magnesium reagents.
EXPERIMENTAL
Materials. p-Cyanobenzophenones were prepared by
Friedel–Crafts reaction of p-cyanobenzoyl chloride and
substituted benzenes and/or reaction of p-bromobenzo-
phenones with copper cyanide in pyridine. 1-Cyano-
9-fluorenone was prepared by Sandmeyer reaction of
1-amino-9-fluorenone18 using copper cyanide or dehydra-
tion of the 1-NH2CO derivative. Other substituted (p-MeO,
p-Me, p-Cl; o-, m-, p-CN; o-, m-, p-NO2 ) benzonitriles were
commercially available.
7. M. Okubo, Y. Fukuyama, M. Sato, K. Matsuo, T. Kitahara and
M. Nakashima, J. Phys. Org. Chem. 3, 379 (1990).
8. K. Matsuo, R. Shiraki and M. Okubo, J. Phys. Org. Chem. 7,
567 (1994).
9. M. Okubo, Y. Nakagawa, M. Yoshida and H. Yanase, Bull.
Chem. Soc. Jpn. 58, 2707 (1983).
10. M. Okubo, Bull. Chem. Soc. Jpn. 58, 3108 (1985), and
references cited therein.
11. M. Okubo, C. Sugimori, M. Tokisada and T. Tsutsumi, Bull.
Chem. Soc. Jpn. 59, 1644 (1986).
Procedures. The preparation of aryl-IDMg2a and the
reaction with various substrates were carried out in THF
according to the reported method. The molar concentrations
of IDMg and the molar ratios with substrates are given
under Tables 1–6. The reaction mixture was quenched with
saturated aqueous NH4Cl and extracted with diethyl ether
and/or dichloromethane, dried over MgSO4 , rotoevaporated
and separated by flash chromatography on silica gel (Wako
Gel FC-40).
12. M. Okubo, Y. Inatomi, N. Taniguchi and K. Imamura, Bull.
Chem. Soc. Jpn. 61, 3581 (1988).
Products. During the course of the chromatographic
separation of products of the reactions in Scheme 1, two
types of amidines are isolated: a highly soluble one having
a lower melting point and simple 1H NMR spectrum, and a
poorly soluble one having a higher melting point and more
complex spectrum. The latter amidine is completely con-
verted into the former on the treatment with water, and is
assigned to a hydrogen-bonded dimeric compound consist-
ing of the (amine-type) amidine and its (imine-type)
tautomer. In Table 7, summarizing the melting points and 1H
NMR data for amidines, those for the dimeric compounds
are omitted.
13. (a) K. Maruyama and T. Katagiri, J. Phys. Org. Chem. 1, 21
(1989); 4, 381 (1991); (b) M. Okubo, M. Tanaka, H. Shiku, A.
Yamauchi and K. Matsuo, J. Phys. Org. Chem. 4, 693 (1991);
(c) Y. Tsuno, M. Fujio and Y. Yukawa, Bull. Chem. Soc. Jpn.
48, 3324 (1975).
14. M. Okubo, S. Yoshida, Y. Egami, K. Matsuo and S. Nagase,
Bull. Chem. Soc. Jpn. 60, 1741 (1987).
15. M. Sawada, Y. Takai, T. Tanaka, T. Hanafusa, M. Okubo and Y.
Tsuno, Bull. Chem. Soc. Jpn. 62, 915 (1989). Liquid crystalline
nature is also known.
16. M. Okubo and Y. Omote, J. Phys. Org. Chem. 9, 212 (1996).
17. (a) M. Okubo, T. Takahashi and K. Koga, Bull. Chem. Soc. Jpn.
56, 199 (1983); (b) M. Okubo and K. Koga, Bull. Chem. Soc.
Jpn. 56, 203 (1983).
18. M. Okubo and Y. Uematsu, Bull. Chem. Soc. Jpn. 55, 1121
(1982).
REFERENCES
1. (a) M. Okubo and K. Matsuo, Yukigosei Kagaku Kyokaishi
(Synth. Org. Chem. Jpn.) 50, 682–691 (1992); (b) M. Okubo
© 1997 by John Wiley & Sons, Ltd.
JOURNAL OF PHYSICAL ORGANIC CHEMISTRY, VOL. 10, 242–252 (1997)