Selective fluorination by halogen exchange of chlorodiazines and chloropyridines promoted by the 'proton sponge' - Triethylamine tris(hydrogen fluoride) system

Article abstract of DOI:10.1016/S0040-4020(00)01060-7

The 'proton sponge' - triethylamine tris(hydrogen fluoride) mixtures provide a mild and efficient fluorinating reagent to introduce selectively fluorine atoms by halogen exchange into chlorodiazines and chloronitropyridine series.

Full text of DOI:10.1016/S0040-4020(00)01060-7

TETRAHEDRON
Pergamon
Tetrahedron 57 (2001) 739±750
Selective ¯uorination by halogen exchange of chlorodiazines and
chloropyridines promoted by the `proton sponge'Ðtriethylamine
tris(hydrogen ¯uoride) system
a
a
a,p
Mircea Darabantu, Thierry Lequeux, Jean-Claude Pommelet, Nelly Ple and Alain Turck
b
Âb
a
  Â
Laboratoire de Chimie Moleculaire et Thio-organique, Universite de Caen- ISMRA, 6 Boulevard du Marechal Juin, 14050 Caen, France
b
Â
Universite de Rouen, IRCOF, 76131 Mont Saint-Aignan, France
Received 28 July 2000; revised 6 November 2000; accepted 7 November 2000
AbstractÐThe `proton sponge'Ðtriethylamine tris(hydrogen ¯uoride) mixtures provide a mild and ef®cient ¯uorinating reagent to intro-
duce selectively ¯uorine atoms by halogen exchange into chlorodiazines and chloronitropyridine series. q 2001 Elsevier Science Ltd. All
rights reserved.
1. Introduction
tion of chlorine atoms in (poly)chloro(benzo)diazines and
pyridines. Thus, the mixtures of Et3N´3HF with PS appeared
to us appropriate for this purpose and our initial results have
been described in a preliminary communication.8
The combinations of hydrogen ¯uoride with organic bases
(the so called `onium hydrogen ¯uorides') as classical
¯uorinating agents were recently reviewed.1 In this context,
special interest has recently focused on triethylamine
tris(hydrogen ¯uoride) (Et3N´3HF)2 for its mild and ef®cient
reactivity. On the other hand, the strain effects on basicity of
`proton sponge' [PS, 1,8-bis(dimethylamino)naphthalene]
have been exhaustively documented as well.3 Nevertheless,
relatively little has been reported to describe in detail the
ability of PS to bind the proton from hydrogen ¯uoride (to
form the `PS' hydrogen ¯uoride, PS´1HF but in a different
way to a simple `onium' structure) and to release ¯uoride
ion (Scheme 1).4,5
We now report the extension of these investigations de®ning
more fully the following topics: (a) the combinations of `PS'
with hydrogen ¯uoride as ¯uorinating reagents; (b) the
ability of mixtures of Et3N´3HF with PS as effective ¯uori-
nating reagents; (c) the quantitative and qualitative results
of the ¯uorinations performed by using these systems. To
our knowledge, this basic synthetic approach has not been
considered so far.
2. Results and discussion
Our ongoing research in the chemistry of diazines6,7
prompted us to look for a mild ¯uorinating reagent, capable
of both selective halogen exchange and complete substitu-
2.1. The ¯uorinating reagents and test experiments
2.1.1. The combinations of `PS' with hydrogen ¯uoride.
Our ®rst objective was to compare different ratios of PS and
HF to identify the best formulation to achieve the ¯uorina-
tion of the title compounds. In order to access PS´1HF (equi-
molar ratio PS:HF, Scheme 1) without the need to handle
hazardous hydrogen ¯uoride4,5 we reacted commercially
available Et3N´3HF with PS (1:3 molar ratio) and com-
pletely removed the triethylamine under vacuum at 1008C;
the reaction was monitored by NMR (Eq. (1)):
3PS 1 Et3N´3HF ! 3ꢀPS´1HF† 1 Et3N
ꢀ1†
Scheme 1.
Keywords: ¯uorine and compounds; pyrimidines; pyridazines; phthala-
zines; quinoxalines; pyridines.
* Corresponding author. Tel.: 133-2-31-45-28-53; fax: 133-2-31-45-28-
77; e-mail: pommelet@ismra.fr
The resulting solid residue was obtained in almost quanti-
tative yield and its NMR spectrum exhibited signals at
chemical shifts close to those previously reported for
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)01060-7

740
M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
promoted by temperature and/or solvents (Eqs. (2) and (3)),
it is similar to some other classical `onium' salts of hydro-
gen ¯uoride such as melamine, pyridine, etc.1
ꢀEther or Pentane†
3ꢀPS´1HF†
!
2PS ꢀas ethereal or pentane solution† 1 PS´3HF #
(2)
ꢀ508C,t,1238C†
3ꢀPS´1HF†
!
2PSꢀmelted mass† 1 PS´3HFꢀsolid†
ꢀ3†
Indeed, supplementary evidence provided by spectral data is
in agreement with the existence of two distinct compounds,
of different stoichiometry, PS´1HF and PS´3HF at least in
the solid state. In the IR spectra (KBr), the previously
reported characteristic bands4,5,9,10 were completely and
identically displayed by the two compounds, with the
exception of the large peak located at 1800 cm21, which is
assigned as combined band of R3NH1 units, strong in
PS´3HF but very weak in PS´1HF.10 PS´1HF and PS´3HF
1
Figure 1. H NMR spectra of PS´1HF.
PS´1HF by Chambers and co-workers4,5 and gave satisfac-
tory elemental analysis. On this basis we assigned this
compound as PS´1HF. When this solid was heated, it was
converted into a mixture of PS and PS´3HF as shown by the
similarity between the temperature (48(9)8C) at which this
change occurs and the literature melting point of 49±508C
quoted for commercially available PS itself. This tempera-
ture differs from the literature value4 (117±88C) given for
PS´1HF. Moreover, the work up at room temperature of our
product with ether or pentane, afforded a new crystalline
solid corresponding to the PS´3HF stoichiometry and the
mother liquor yielded the theoretical amount of starting
free base PS (Eq. (2)). This product PS´3HF melts at 125±
68C. This behaviour is not altogether unexpected since, as
exhibited the strong nNH stretch band (located at 3500 cm21
)
which is completely absent in the PS spectrum.
The NMR spectra showed rather interesting differences
between the two types of salts: only PS´3HF displayed all
the expected sharp splitting in the aromatic region (see
Experimental) (Figs. 1 and 2).
This peculiar behaviour of PS´1HF (Eqs. (2) and (3)) led us
to assume that the structure should be, from thermodynamic
point of view, less stable than PS´3HF, and that only PS´3HF
exhibits enough stability to develop crystals for X-ray
analysis.8 The solid state structure gives the formulation
Figure 2. Relevant NMR spectra of PS´3HF; (a) 1H spectrum in the aromatic region (<0.070 g/mL), from down®eld to up®eld (d, ppm): 7.93 (H-para, d), 7.83
1
(d, H-ortho), 7.66 (dd as t, H-meta); (b) H spectrum in the aromatic region (<0.140 g/mL) from down®eld to up®eld (d ppm): 7.94 (H-para, d), 7.92 (d,
H-ortho),7.68 (dd as t, H-meta).

M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
741
supplementary test, provided by 2-chloro-3,5-dinitropyri-
dine, was considered.
Surprisingly, the NMR monitoring of the reactions (Scheme
2) unambiguously revealed the presence of unreacted
PS´3HF in both cases I, II for both compounds (Fig. 4).
In our opinion these facts support the below preliminary
conclusions:
(a) the real ¯uorinating reagent, in both cases, is PS´3HF;
consequently, a structure having the PS´1HF stoichiometry
is valid in solid state only. In solution one has rather to
consider an equilibrium involving the generation of another
stoichiometry assigned as (PS)2(PS´3HF) in which the two
`free' PS units are designed to bind the resulting HCl (Eq.
(4)):
3ꢀPS´1HF† Y ꢀPS†2ꢀPS´3HF†
ꢀ4†
(b) the lower reactivity of PS´3HF alone (Eqs. (5) and (6)),
compared with that of PS´1HF (Eq. (7)) (in identical condi-
tions, Scheme 2) can be explained by simple acid-base
considerations:
Figure 3. The ORTEP diagram of PS´3HF.
(PSH1) (HF22) (HF) (Fig. 3) whereas PS´1HF revealed slow
decomposition as hygroscopicity and oxidation.
R±Cl 1 PS´3HF ! R±F 1 PS´2HF 1 HCl
PS´2HF 1 HCl ! PS´HCl # 12HF
R±Cl 1 PS´1HF ! R±F 1 PS´1HCl #
ꢀ5†
ꢀ6†
ꢀ7†
A comparative test of reactivity of these two reagents
(PS´1HF and PS´3HF) against 2,4,6-trichloropyrimidine
and 2-chloro-3,5-dinitropyridine was straightforward and
supported this assumption (Scheme 2).
As shown by the conversions, PS´1HF exhibited a higher
reactivity than PS´3HF. We note however that, in the case of
2,4,6-trichloropyrimidine, in contrast to the results
previously reported by Chambers and co-workers,4,5 no
reaction occurred at room temperature with any of the
above reagents; no side product (`PS' hydrochloride
PS´1HCl) precipitated from the reaction mixture with aceto-
nitrile solution. It is also worth mentioning that, although
the presence of tri¯uoropyrimidine was unambiguously
revealed by 19F NMR spectroscopy (e.g. two signals
located at 240.6 and 252.3 ppm in 1:2 intensity ratio,
respectively), its effective isolation in the conditions
depicted in Scheme 2 was irrelevant due to its high
volatility. Therefore, the conversion was calculated based
on the effective amounts of the isolated PS´1HCl and a
In the case of PS´3HF, there is no free base to neutralise the
resulting HCl. Next, one should take into account the
normal higher stability of PS´1HCl vs. all the analogous
PS´xHF (e.g. xˆ2, Eq. (6)); that is, in conditions depicted
in Scheme 2, just one of the three HF units in PS´3HF is
`active'.
2.1.2. The `PS'Ðtriethylamine tris(hydrogen ¯uoride)
(Et3N´3HF) system. The idea to test the above mixtures
as ¯uorinating reagents was evidently inspired by the possi-
bility to generate in situ the PS hydrogen ¯uorides in
connection with the presence of another base, triethylamine,
assigned a priori as HCl acceptor.
Scheme 2.

742
M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
Figure 4. 1H NMR spectra of the crude reaction mixtures of 2-chloro-3,5-dinitropyridine with `PS' hydrogen ¯uorides, after 24 h (details); (a) Route I
(Scheme 2, ¯uorinating reagent PS´1HF, 69% conversion), from down®eld to up®eld (d ppm, J Hz): 9.43 (1H, d, 2.5, chloro derivative); 9.36 (1H, dd, 3.2, 2.5,
¯uoro derivative); 9.31 (1H, dd, 7.5, 2.5, ¯uoro derivative); 9.03 (1H, d, 2.5, chloro derivative); 8.00±7.50, PS´3HF (aromatic protons); (b) Route II (Scheme
2, ¯uorinating reagent PS´3HF, conversion38%).
Since the above mixtures are, clearly, acid-base equilibria-
systems and based on the difference between pKa values
of the two bases (12.10 PS and 10.75 Et3N),3 in a pre-
liminary experiment we attempted to prepare PS´3HF in a
similar pathway as PS´1HF (equimolar ratio PS:Et3N´3HF,
Eq. (8)).
support the spectral data (Eqs. (9) and (10)).
PS´1HF 1 2ꢀEt3N´3HF† Y PS´3HF 1 2ꢀEt3N´2HF†
ꢀ10†
Thus, we attempted at estimating the compositions but
restricted to the same type of species: derived from triethyl-
amine (Et3N´3HF and Et3N´2HF, Eqs. (9) and (10)) and
derived from PS (in Eq. (9): PS and PS´1HF; in Eq. (10):
PS´1HF and PS´3HF). We selected as relevant signals
the most separated and in¯uenced ones belonging to the
same basic unit: 13C d value of methyl groups in PS,
PS 1 Et3N´3HF Y PS´3HF 1 Et3N
ꢀ8†
This experiment failed, indicating that the difference
between pKa values it is too small for complete `transfer'
of all three HF units (but one, Eq. (9)) to the stronger base
from the weaker one even by attempting to remove the latter
(to be compared with Eq. (1)).
1
PS´1HF, PS´3HF and H d value of methylene groups in
Et3N´3HF and Et3N´2HF. The estimated compositions,
according to the selected equilibria 9 and 10, are collected
in Table 1.
PS 1 Et3N´3HF X PS´1HF 1 Et3N´2HF
ꢀ9†
Data summarised in Table 1 prompted us the following
remarks:
On the other hand, as we recently reported,8 the behaviour of
these mixtures against (poly)chloro(benzo)diazines was like
authentic ¯uorinating reagents on a large scale of molar
ratio between PS and Et3N´3HF and very different than
Et3N´3HF alone. Therefore, we investigated from the
NMR point of view several mixtures de®ned as
xPS1y(Et3N´3HF) by varying the molar ratio between
(a) the option to restrict the acid-base process to the equi-
libria 9 and 10 only as dominant seems to be satisfactory
since the observed averaged d values of the mixtures
xPS1y(Et3N´3HF) ranged strictly between chemical shifts
of the pure involved species (e.g. no free Et3N is present
since the d values of methylene group ranged only between
2.91±3.00 ppm).
²
components x:y from 4:1 to 1:4 The appearance of these
spectra revealed unambiguously the presence of PS´1HF
and PS´3HF environments, depending on x:y ratio.
(b) in the domain of excess of PS vs. Et3N´3HF (x:y from 4:1
to 1:1, respectively) the reagents can be described by the
Eq. (9): the content of PS´1HF species increases rapidly with
increasing the ratio of Et3N´3HF. In turn, the variation of the
d values for F2 suggests rather the Et3N´2HF environment.
Anyhow, its ability to halogen exchange should increase
from left to right. It was expected that this type of reagent
should be useful for compounds possessing themselves a
good reactivity (see Scheme 2).
Obviously, the intimate nature of all possible terms involved
in the above equilibria is quite impossible to elucidate
averaged d values by means of NMR. Nevertheless, in our
opinion, only two successive equilibria are rather relevant to
²
This domain of variation of x:y molar ratio was currently used in our
syntheses; see later.

M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
Table 1. Relevant NMR data for the equilibria (9) and (10) (solvent CDCl3)
743
da (ppm) in pure compound
Reagent (Eq. (9))
Reagent (Eq. (10))
d (ppm) in pure compound
x:y in xPS1y(Et3N´3HF)
x:y in xPS1y(Et3N´3HF)
4:1
1:1
1:2
46.0
1:4
46.1
38%
44.8 (d 13C CH3) 100% PS
44.9b
80%d
20%
45.1c
50%
50%
46.6 (d 13C CH3) 100% PS´3HF
PS
PS´1HF
50%
50%
PS´1HF
PS´3HF
62%
2.85c (d 1H CH2) 100%
Et3N´2HF
2.91
2.96
2.96
3.00
3.13 (d 1H CH2) 100%
Et3N´3HF
80%
20%
Et3N´2HF
Et3N´3HF
60%
40%
40%
60%
Et3N´3HF
Et3N´2HF
54%
46%
2154.4 (d 19F) 100% Et3N´2HF
2158.0
2158.7
2159.5
2160.5
2167.6 (d 19F) 100% PS´3HF
a d All signals were measured against the solvent peak, throughout located at 78.0 ppm (13C NMR) and 7.27 ppm (1H NMR).
b d Minimal value in the successive mixtures, by varying x:y.
c d Maximal value in the successive mixtures, by varying x:y; d value for CH3 group in isolated PS´1HF is 45.4 ppm
d All percentages were calculated starting from the averaged d values applied to Eqs. (9) and (10) by using the known relationship dˆd1x11d2x2 where x1,2 are
the molar fractions of the considered environments: PS/PS´1HF (Eq. (9)) PS´1HF/PS´3HF (Eq. (10)) and Et3N´2HF/Et3N´3HF (Eqs. (9) and (10)).
(c) in the domain of excess of Et3N´3HF vs. PS (x:y from 1:1
to 4:1, respectively) the reagents should obey to Eq. (10):
one can see only a slight increasing of PS´3HF species.
Indeed, the HF acceptor in these mixtures is now PS´1HF
and not the free base PS (to compare with Eq. (9)). Never-
theless, the appearance of PS´3HF environment was clearly
displayed in the mixture xˆ1, yˆ4.
Based on the test experiments, we assumed that these
reagents should be appropriate for less reactive substrates.
Scheme 3.

744
M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
Table 2. Relevant results of the ¯uorination of the diazines 1±4
a
Total conversion (% molar) of the starting materials 1±4 towards all products 1a±c, 2a, 3a, b and 4, respectively.
Yields are calculated as isolated materials; the amount of the recyclable starting material was not taken into account.
See Experimental for this result as Method A.
b
c
d
e
f
1
Relative molar ratios between reaction products in the crude reaction mixture as issued from H correlated with 19F NMR spectra.
See Experimental for this result as Method B.
See Experimental for this result as Method C.
g For the compounds 3a, b Fukuhara, T. and Yoneda, N. reported a 92% yield (as 56% di¯uro- 3b vs. 44% mono ¯uoroderivative 3a) by using anhydrous HF in
autoclave (2 h at 1008C); no structural assignment accompanied these data.11 However, if the complete substitution of chlorine is desired, much stronger
conditions are required, in two step synthesis.13,17
Indeed, the greater stability of PS´3HF allows its use at
higher temperatures and hard conditions. However, a base
should be present to bind the resulting HCl and this can be
readily accomplished (Eq. (6)) by the excess of Et3N´xHF
(xˆ2, 3) to yield the more thermodynamically stable
triethylamine hydrochloride. The free hydrogen ¯uoride
could either activate the substrate towards nucleophilic
substitution, in agreement with the recent Yoneda and
co-workers assignments.11
As usually observed,14±16 2,4-dichloropyrimidine 1 (Table
2, entry 1) exhibited higher reactivity at C-4 than at C-2
position and regioselectivity was not found to be dependent
on the use of a solvent or type of reagent, if global equimolar
ratio (dichloroderivative:F2) was used. If the reaction was
carried out in acetonitrile, a lower yield was observed as the
consequence of the loss of the product during the work-up
(promoted by its high volatility), in order to remove the
2.2. Selective ¯uorinations by using the PS±Et3N´3HF
system
In this section, attention is dedicated to the selective nucleo-
philic replacement of chlorine atoms in some p-de®cient
heteroaromatic compounds4,5,11±13 according to Scheme 3.
2.2.1. Fluorination of the chlorodiazines 1±4 (Table 2,
Scheme 4). Chloropyrimidines were suitable substrates to
undergo either selective or complete substitution of the
chlorine atoms.
Scheme 4.

M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
Table 3. Relevant results of the ¯uorination of the dichlorobenzodiazines 5 and 6
745
a
Total conversion (% molar) of the starting material 5, 6 towards all products 5a, b and 6a, b, respectively.
b Yields are calculated as isolated materials; the amount of the recyclable starting materials was not taken into account.
See Experimental for this result as Method B.
c
d Relative molar ratios between reaction products in the crude reaction mixture as issued from 1H correlated with 19F NMR spectra.
See Experimental for this result as Method D.
e
solvent. Apparently, the use of the mixture 1PS11/
3(Et3N´3HF) without solvent had been more convenient
but the isolated product was strongly contaminated with
triethylamine during the isolation by vacuum distillation.
Consequently, the use of PS´1HF alone was the best
pathway to access 1a.
Preliminary attempts performed on 3,6-dichloropyridazine
3 at 808C with Et3N´3HF alone shown that it required more
than 20 equiv. of F2 for the detection of 3a in small traces
only. Hence 3 exhibited lower reactivity than the chloro-
pyrimidines 1, 2 and we selected stronger reagents (Table 1,
xˆ2, yˆ4). Thus (Table 2, entry 3), we found convenient
conditions for a chemoselective process, as a compromise
between conversion and selectivity. This option was then
supported by the results obtained when the reaction was
attempted with single PS´1HF (too weak: 3a was detected in
traces after 111 h/808C, 3:1 molar ratio reagent:3) or PS´3HF/
Et3N (best selectivity as 30:1 3a:3b but only 23% yield).
Complete ¯uorination to access 1b (seen as trivial case) was
not attempted.6
4-Chloro-2-methylsulfanylpyrimidine 2 (Table 2, entry 2)
required stronger conditions, presumably because of the
electron-donating substituent attached at C-2. Previous
`classical' ¯uorination of this compound by us6 (1508C,
KF, 6 h in tetraglyme) afforded 2a in 67% yield. With
the present reagent (3PS13Et3N´3HF) quantitative
conversion was observed; due to losses of product
during the work-up to remove triethylamine, the isolated
yield was reduced to 70%. In turn, step by step 1H
NMR monitoring of the same ¯uorination but promoted
by single PS´1HF or PS´3HF showed a faster decomposition
of the reaction mixture than a satisfactory conversion of 2
into 2a.
In the case of 3,6-dichloropyridazine, the high selectivity
generally observed could be correlated with the difference
between the PM3 calculated stability of the corresponding
Meisenheimer complex: M-3a was found more stable than
M-3b as precursor of 3b (Scheme 4).
It is of interest to note that the presence of phenyl substituent
at C-4 (compound 4) signi®cantly decreased the reactivity
(entry 4) since it acts like a releasing group; accordingly, no
reaction occurred with 4-methoxy derivative.
Scheme 5.

746
M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
Scheme 6.
2.2.2. Fluorination of the dichlorobenzodiazines 5 and 6
(Table 3, Schemes 5 and 6). As the type of ¯uorinating
reagents used indicate, the dichlorobenzodiazines 5 and 6
were found with middle reactivity between the correspond-
ing pyrimidines and pyridazines.
by the greater stability of the Meisenheimer complex
M-6a against M-6b (precursor of di¯uoro derivative 6b,
Scheme 6).
2.2.3. Fluorination of the chloronitropyridines 7 and 8
(Table 4). The test experiments revealed as appropriate
substrates to be ¯uorinated by the present reagents only
chloronitropyridines among chloropyridines: no reaction
occurred with 2,3-dichloropyridine.
For 2,3-dichloroquinoxaline 5, preliminary examination of
the results of PM3 calculations led us to anticipate that
selectivity should be poorer since the DDHf value between
the two Meisenheimer complex M-5a vs. M-5b (Scheme 5)
was found much smaller as compared to 3,6-dichloro-
pyridazine (Scheme 4).
We have already seen the high reactivity of 2-chloro-3,5-
dinitro-pyridine (Scheme 3). 2-Chloro-3-nitropyridine 7
yielded 7a comparably with already reported results if
potassium ¯uoride had been used (1508C, DMF, 6 h, 76%
yield).20
Accordingly, we successfully performed either monochloro
(5a) or dichloro (5b) substitution by changing molar ratio of
the reagents to the substrate (entry 1). We outline that
complete separation of the reaction products was easily
achieved by classical protocols (see Experimental). No
peculiar susceptibility of di¯uoro derivative 5b towards
hydrolysis, during the work-up, was observed.18
Attempts at regioselective mono¯uorination of 2,6-
dichloro-3-nitropyridine 8 failed. The regioisomers 8b, c
were detected in almost equal proportion (19F NMR
monitoring) on a large scale of molar ratio (x vs. y), to afford
the di¯uoro derivative 8a (entry 2) as the major product.
Thus, total substitution was found to be the dominant
process. Despite excellent conversions, yields were poor
due to the high volatility of 8a (vacuum distillation or
¯ash column chromatography).
Unexpected problems were encountered concerning the
¯uorination of 1,4-dichlorophthalazine 6 (entry 2) since
the target product, the mono¯uoro derivative 6a, exhibited
instability in aqueous media. It seemed to us the behaviour
of these 1,4-dihalophthalazines as similar to that of their
perchloro- and per¯uoro-analogues which were also
reported as very sensitive to nucleophilic replacement of
halogen atoms.19 The selectivity was very high, in mild
conditions, with reasonable time reaction and supported
3. Conclusion
`PS' hydrogen ¯uorides or reagents generated in situ from
Table 4. Relevant results of the ¯uorination of the chloronitropyridines 7 and 8
a
Total conversion (% molar) of the starting materials 7, 8 towards all products 7a, 8a±c.
Yields are calculated as isolated materials; the amount of the recyclable starting materials was not taken into account.
See Experimental for this result as Method B.
b
c
d
e
See Experimental for this result as Method E.
1
Relative molar ratios between reaction products in the crude reaction mixture as issued from H correlated with 19F NMR spectra.

M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
747
mixtures of triethylamine tris(hydrogen ¯uoride) and PS
provide versatile and ef®cient ¯uorinating reagents. Their
use is compatible with wide ranges of temperature and
concentration. For the ®rst time, complete or selective
¯uorinations of certain dichloro(benzo)diazines can be
conveniently carried out using this type of reagent with
satisfactory yields. In some cases, the selectivity can be
explained based on the difference between thermodynamic
stability of the Meisenheimer complexes.
Method B: the substrate and the entire amount of reagent
xPS1y(Et3N´3HF) are intimately mixed and then heated at
the required temperature; no solvent is used.
Method C: the reagent xPS1y(Et3N´3HF) is added portion-
wise to the substrate heated at the required temperature; no
solvent is used.
Method D: the reagent xPS1y(Et3N´3HF) is added portion-
wise to the substrate (the latter as acetonitrile solution)
under heating at 808C.
4. Experimental
4.1. General
Method E: the substrate and the entire amount of reagent
xPS1y(Et3N´3HF) are dissolved in acetonitrile and heated
at 808C.
All reagents were purchased from Aldrichw and used with-
out supplementary puri®cation. All syntheses were carried
out under dry nitrogen. Melting points were determined on a
Reichert Austria instrument and are not corrected. NMR
spectra were recorded on Brucker 250 instrument operating
at 250, 62.5 and 235 MHz for 1H, 13C and 19F nuclei, respec-
tively. No SiMe4 was added; chemical shifts (ppm) were
measured against the solvent peak (throughout CDCl3)
except d values for 19F nuclei (CFCl3 as external standard).
Samples were directly prepared in NMR tubes by using
concentrations ranging between 0.015±0.28 g/mL and
measured after 128 scans (for 1H NMR spectra). High
resolution mass spectra were performed on JEOL AX500
instrument operating at IEˆ70 ev. IR spectra were recorded
in solid state (KBr) on a Perkin±Elmer 16PC FT-IR instru-
ment. TLC was performed by using aluminium sheets with
silica gel 60 F254 (Merckw); ¯ash column chromatography
was conducted on Silica gel Si 60 (40±63 mm) (Merckw).
4.3. Preparation of 1,8-bis(dimethylamino)naphthalene
hydrogen ¯uoride, PS´1HF
`PS' PS, 0.5 g, (2.33 mmol) was dissolved in dry dichloro-
methane (10 mL) then triethylamine tris(hydrogen ¯uoride)
Et3N´3HF (0.125 g, 0.126 mL, 0.77 mmol) was added. The
resulting solution was evaporated under vacuum at room
temperature to give a crystalline solid that was heated for
1.5 h at 90±58C (under vacuum, 10±15 mmHg) on a steam
bath. The crystalline residue was cooled at room tempera-
ture, to yield a sticky mass. Yield 98% (0.535 g).
4.3.1. 1,8-Bis(dimethylamino)naphthalene hydrogen
¯uoride, PS´1HF. Grey crystalline powder; slow thermal
dissociation up to 498C. Careful inspection of the crystals
and the behaviour of the solid between 45±508C did not
reveal the presence of two distinct types of crystals, but
only one which converted into melted mass and a similar
type of crystal; the latter exhibited a higher melting point, at
1
125±68C. H NMR (d ppm, J Hz): 2.95 (bs, 12H, ±CH3),
4.2. Special remarks
7.48, 7.39, 7.23 (2H, 2H, 2H as three broad singlets); 13C
NMR (d ppm): 45.4 (4C, ±CH3), 126.3 and 120.5; 19F
NMR: 2167.2 (s) Anal. calcd for C14H19FN2: C 71.76%,
H 8.17%, N 11.95%; found C 72.41%, H 8.50%, N 12.35%.
IR (KBr, s-strong, m-medim, w-weak, cm21): 3450 (s, nNH),
2000 (w, nNH1), 1800 (w, nN1 comb.), 1500 (s, naphthalene
Several factors should be outlined as common for the entire
series: the difference between essential physical data (vola-
tility, polarity, basicity, etc.) regarding the starting materials
and products is minor for complete analytical separation.
Thus, the option for a certain ¯uorinating reagent
[PS´1HF, PS´3HF or xPS1y(Et3N´3HF)] was crucial since
it had to consider not only the reactivity of the substrate but
also the volatility of the desired products. Accordingly, ®ve
different synthetic methods (A±E) were used. PS´1HF,
PS´3HF, 2-¯uoro-3,5-dinitropyridine and compounds 2a,
5a, 5b, 8a were isolated as pure analytical samples and
exhibited satisfactory elemental analysis and/or HR-MS
results; for the rest of the series, the desired compounds
(1a, 3a, 4a, 6a, 7a) were isolated as inseparable mixtures.
Their identity was fully con®rmed by NMR spectra and
HR-MS.
ring9), 1600±1400 (s, dNH), 600 (m, nNH N).
1
4.4. Preparation of 1,8-bis(dimethylamino)naphthalene
tris(hydrogen ¯uoride), PS´3HF
The PS´1HF (0.535 g, as the solid residue, see above) was
taken with dry diethyl ether (or pentane) (10 mL) and stirred
at room temperature to give a ®ne suspension. After ®lter-
ing, very well washing with diethyl ether (or pentane) and
drying at room temperature, PS´3HF (0.200 g, 96% yield)
was obtained. Evaporation of the mother liquor afforded
0.310 g (95% recovered) PS as free base. These syntheses
allowed a scale up to 10 g of the starting PS, with the same
yields.
The methods A±E are summarised below; for the reagent
xPS1y(Et3N´3HF), the parameters x and y indicate the
number of mole of each component vs. 1 mol of substrate.
Vigorous stirring was required in all cases. Although all
experiments were carried out at least twice, yields were
not optimised.
4.4.1. Bis(dimethylamino)naphthalene tris(hydrogen
¯uoride). White crystalline powder; mpˆ125±68C. 1H
NMR (d ppm, J Hz): 3.25 (s, 12H, ±CH3), 7.93 (d, 2H,
7.9), 7.83 (d, 2H, 7.8), 7.66 (dd as t, 2H 7.8); 13C NMR (d
ppm): 46.6 (4C, ±CH3), 119.2 (2C), 121.6 (2C), 127.4 (2C),
129.5 (2C), 135.8 (1C), 144.9 (1C); 19F NMR: 2167.6 (s).
Method A: PS´1HF is added portionwise to the substrate
heated at the required temperature; no solvent is used.

748
M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
Anal. calcd for C14H21F3N2: C 61.29%, H 7.71%, N 10.21%;
found C 61.38%, H 7.79%, N 10.23%. IR (KBr, s-strong, m-
medium, w-weak, cm21): 3450 (s, nNH), 2000 (w, nNH1),
mixed together and heated with stirring, at 1208C for 24 h.
After cooling at room temperature, diethyl ether (3£25 mL)
was used to extract the crude product from the reaction
mixture (the unsoluble solids were kept to recycle the PS).
The ethereal solution was washed with water (3£25 mL),
dried over MgSO4, then evaporated in vacuo without heat-
ing to yield 0.870 g crude product. After distillation of the
latter in a kugelrohr instrument (1008C, 15 mmHg), 0.610 g
(70% yield) 2a was obtained as yellow liquid.
1800 (s, nN comb.), 1500 (s, naphthalene ring9), 1600±
1
1
1400 (s, dNH), 600 (m, nNH N).
4.5. NMR-monitoring of the equilibria (9) and (10)
The mixtures xPS1y(Et3N´3HF) were prepared directly in
the NMR tube and x and y were selected as absolute values
to give about 0.28±0.29 g/mL; all tubes were measured
consequently by using the same shimming values, at room
temperature. Pure compounds PS, PS´1HF, PS´3HF,
Et3N´3HF were measured individually as 0.280, 0.150,
0.080 and 0.015 g/mL. Et3N´2HF was prepared from
Et3N´3HF and the corresponding amount of triethylamine
free base by using the same method, as described in the
literature.2
4.7.1. 4-Fluoro-2-methylsulfanylpyrimidine, 2a. Yellow
liquid, bpˆ1008C/15 mmHg. Elemental analysis and
NMR data were listed elswhere.6 HR MS for C5H5FN2S
144.0157; found 144.0157.
4.8. Preparation of 6-chloro-3-¯uoropyridazine, 3a
(Method C, Table 2)
3,6-Dichloropyridazine (7.22 g, 46.9 mmol) and 50% from
the required amount of the reagent 2PS14(Et3N´3HF) were
mixed together and heated with stirring at 808C. The rest of
the reagent was added as follows: 33% (after 30 h) and 17%
(after 36 h). After additional 30 h, the reaction mixture was
cooled at room temperature, dissolved in dichloromethane
(200 mL) and the organic solution was washed with water
(200 mL). The aqueous layer was kept to recycle the PS.
The dichloromethane layer was washed with water
(3£50 mL), dried over MgSO4 then concentrated under
vacuum to yield 6.15 g crude product (see composition in
Table 2). The distillation in a kugelrorh instrument (1508C,
0.5 mmHg) gave 4.25 g mixture (3, 3a, b) isolated as
colourless liquid (about the same composition as the
crude, 53% yield with respect to the main product, 3a).
4.6. Preparation of 2-chloro-4-¯uoropyrimidine, 1a
(Method A, Table 2)
2,4-Dichloropyrimidine (1.50 g, 10.00 mmol) and PS´1HF
(15% from the equivalent amount required by the 1:1
stoichiometry) were intimately mixed and heated at 658C
with stirring. The rest of the amount of PS´1HF was added
portionwise as follows: 20% (after 3 h), 30% (after 10 h)
1
and 35% (after 24 h). After additional 24 h at 658C, the H
and 19F NMR monitoring of the crude reaction mixture
indicated the composition (as molar ratios) 1a:b:c as
depicted in Table 2. The crude reaction mixture was directly
distilled under vacuum (1258C, 15 mmHg) in a kugelrohr
instrument to afford 1.11 g mixture of 1a±c (about the same
composition as the crude). Yield 40%, with respect to the
main product 1a. The solid residue was kept to recycle the
PS.
4.8.1. 6-Chloro-3-¯uoropyridazine, 3a. Yellowish liquid;
1H NMR (d ppm, J Hz): 7.91(1H, dd, 6.4, 9.1), 7.54 (1H, dd,
2.1, 9.1); 13C NMR (d ppm, J Hz): 166.4 (1C, d, 244.4),
155.3 (1C, d, 2.5), 134.2 (1C, d, 7.5), 119.2(1C, d, 35.0); 19F
NMR (d ppm, J Hz): 281.6 (1F, d, 5.7). HR MS for
C4H2ClFN2 131.9891 and 133.9861; found 131.9903 and
133.9872.
4.6.1. 2-Chloro-4-¯uoropyrimidine, 1a. Colourless liquid;
1H NMR (d ppm, J Hz): 8.72 (1H, dd, 5.5, 10.9), 7.05 (1H,
dd, 2.7, 5.5); 13C NMR (d ppm, J Hz): 168.2 (1C, d, 258.4),
163.3 (1C, d, 6.9), 161.9 (1C, d, 12.5), 106.5 (1C, d, 28.1);
19F NMR (d ppm, J Hz): 256.5 (1F, d, 10.9). HR MS for
C4H2N2ClF 131.9891 and 133.9861; found 131.9890 and
133.9876.
1
4.8.2. 3,6-Di¯uoropyridazine, 3b. Yellowish liquid; H
NMR (d ppm, J Hz): 7.67(2H, dd as t, 1.5); 13C NMR (d
ppm, J Hz): 165.5 (2C, d, 244.6), 122.2(2C, t, 22.5); 19F
NMR (d ppm, J Hz): 282.7 (2F, s). Only distinct peaks
were listed.
4.6.2. 2,4-Di¯uoropyrimidine, 1b. 1H NMR (d ppm, J Hz):
8.71 (1H, dd, 5.5, 11.0), 7.05 (1H, dd, 2.7, 5.5); 13C NMR (d
ppm, J Hz): 170.2 (1C, d, 258.8), 170.1 (1C, dd, 14.1,
258.4), 158.7 (1C, dd, 9.7, 24.1), 103.8 (1C, dd, 6.25,
28.1); 19F NMR (d ppm, J Hz): 256.0 (1F, d, 9.2), 243.6
(1F, s).
4.9. Preparation of 3-¯uoro-6-phenylpyridazine, 4a
(Table 2)
Method B was used, starting from 0.750 g (3.93 mmol)
3-chloro-6-phenylpyridazine and the corresponding amount
of the reagent 4PS16(Et3N´3HF); in the work up, diethyl
ether was replaced by the corresponding amount dichloro-
methane. The crude product (0.950 g) was puri®ed by
column chromatography (15 g silica gel, eluent Et2O/Pen-
tane 3:1 v/v, typical fraction 8 mL) to give 0.527 g product
as mixture 84% 4a116% 4 in a single fraction (63% yield
with respect to 4a).
4.6.3. 4-Chloro-2-¯uoropyrimidine, 1c. 1H NMR (d ppm,
J Hz): 7.40 (1H, dd, 3.0, 5.2); 13C NMR (d ppm, J Hz):
170.0 (1C, d, 258.1), 162.7 (1C, d, 6.3), 160.1 (1C, d, 12.5),
118.4(1C, d, 5.0); 19F NMR (d ppm, J Hz): 243.8 (1F, s).
Only distinct signals were listed.
4.7. Preparation of 4-¯uoro-2-methylsulfanyl-
pyrimidine, 2a (Method B, Table 2)
4-Chloro-2-thiomethylpyrimidine (1.00 g, 6.10 mmol), and
the required amount of the reagent 3PS13(Et3N´3HF) were
4.9.1. 3-Fluoro-6-phenylpyridazine, 4a. White crystalline
powder, mpˆ123±48C (pentane). H NMR (d ppm, J Hz):
1

M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
749
8.02±7.96 (3H, m), 7.54±7.50 (3H, m), 7.29 (1H, dd, 6.6,
7.2);13C NMR (d ppm, J Hz): 166.5 (1C, d, 243.8), 159.7
(1C, s), 135.5 (1C, s), 130.9(1C, s), 130.0 (1C, d, 7.5), 129.5
(2C, s), 127.5 (2C, s), 116.5 (1C, d, 33.1); 19F NMR (d ppm,
J Hz): 282.6 (1F, d, 6.6). HR MS for C10H7FN2 174.0593;
found 174.0587.
858C. After additional 48 h the crude reaction mixture was
directly submitted to column chromatography (eluent
dichloromethane/ethyl acetate 30:1 v/v) to afford: 0.016 g
(as the ®rst fraction: 6b 24% and PS´3HF 76%) and 0.275 g
(as the second fraction: 6a 62%, 6 38%, 37% yield with
respect to 6a).
4.12.1. 4-Chloro-1-¯uorophthalazine, 6a. 1H NMR (d
ppm, J Hz): 8.24±8.13 (4H, m); 13C NMR (d ppm, J Hz):
161.3 (1C, d, 251.9), 159.3 (1C, d, 6.9), 135.0 (1C, d, 21.3),
135.0 (1C, s), 134.8 (1C, d, 10.6), 134.9 (1C, s), 123.0 (2C,
s); 19F NMR (d ppm, J Hz): 287.5 (1F, s). HR MS for
C8H4ClFN2 182.0047 and 184.0019, found 184.0040 and
182.0057.
4.10. Preparation of 2,3-di¯uoroquinoxaline, 5b
(Table 3)
Method B was used, starting from 1.00 g (5.02 mmol) 2,3-
dichloroquinoxaline and the corresponding amount of
reagent 4PS14/3(Et3N´3HF). In the work up, diethyl ether
was replaced by the corresponding amount dichloro-
methane. The crude product (0.770 g) was puri®ed by
column chromatography (45 g silica gel, eluent chloro-
form/pentane 1:5 v/v, typical fraction 8 mL) to give
0.420 g di¯uoroderivative 5b and 0.070 g mono¯uoro
derivative 5a as two distinct fractions. Yield 50% with
respect to the desired product 5b.
1
4.12.2. 1,4-Di¯uorophthalazine, 6b. H NMR (d ppm, J
Hz): 7.64 (2H, dd, 5.7, 3.3), 7.46 (2H, dd, 5.7, 3.3); 19F
NMR (d ppm, J Hz): 284.2 (2F, s); only distinct peaks
were listed.
4.13. Preparation of 2-¯uoro-3-nitropyridine, 7a
(Table 4)
4.10.1. 3-Chloro-2-¯uoroquinoxaline, 5a. (White crystal-
line powder, mpˆ88±98C); 1H NMR (d ppm, J Hz): 8.11±
8.05 (1H, m), 8.03±7.92 (1H, m), 7.86±7.53 (2H, m); 13C
NMR (d ppm, J Hz): 152.1(1C, d, 256.3), 141.2 (1C, d, 2.5),
139.0(1C, d, 8.8), 137.1 (1C, d, 40.6), 131.7 (1C, s), 130.6
(1C, d, 3.1), 128.6 (1C, s), 128.2 (1C, d, 1.9); 19F NMR (d
ppm, J Hz): 282.7 (1F, s). HR MS for C8H4ClFN2 182.0047
and 184.0019, found 182.0052 and 184.0007. Anal. calcd
for C8H4ClFN2: C 52.63%, H 2.20%, N 15.33%; found C
52.96%, H 2.29%, N 15.06%.
Method B was used starting from 2-nitropyridine (0.635 g,
4 mmol) and the corresponding amount of the reagent
2PS16(Et3N´3HF). Diethyl ether was replaced by dichloro-
methane in the work-up. The crude product (0.578 g 60%
7a, 40% 7 according to 1H NMR spectrum) was distilled in a
kugelrohr instrument (bpˆ758C/0.6 mmHg) to give 0.426 g
as a single fraction (89% 7a, 58% yield and 11% 7).
4.13.1. 2-Fluoro-3-nitropyridine, 7a. Bright yellow oil; 1H
NMR (d ppm, J Hz): 8. 63±8.55 (2H, m), 7.54(1H, ddd, 7.9,
4.9, 0.7); 13C NMR (d ppm, J Hz): 155.5 (1C, d, 248.1),
153.3 (1C, d, 24.8), 153.0 (1C, d, 15.0), 137.3 (1C, s), 122.8
(1C, d, 5.6); 19F NMR (d ppm, J Hz): 268.4 (1F, d, 7.5). HR
MS for C5H3FN2O2 142.0179; found 142.0167.
4.10.2. 2,3-Di¯uoroquinoxaline, 5b. (Yellowish crystalline
1
powder, mpˆ92±38C); H NMR (d ppm, J Hz): 7.88 (2H,
dd, 6.3, 3.5), 7.69 (2H, dd, 6.3, 3.5); 13C NMR (d ppm, J
Hz): 146.5 (2C, dd, 261.3, 39.4), 138.8 (2C, dd as t, 5.6),
130.7(2C, s), 128.2(2C, s); 19F NMR (d ppm, J Hz): 282.7
(1F, s). HR MSfor C8H4N2F2 166.0342; found 166.0347.
Anal. calcd for C8H4F2N2: C 57.83%, H 2.42%, N
16.86%; found C 57.93%, H 2.54%, N 16.61%.
4.14. Preparation of 2,6-di¯uoro-3-nitropyridine, 8a
(Method E, Table 4)
4.11. Preparation of 3-chloro-2-¯uoroquinoxaline, 5a
(Method D, Table 3)
2,6-Dichloro-3-nitropyridine (1.11 g, 5.18 mmol), and the
required amount of the reagent 2.5PS12.5/3(Et3N´3HF) in
acetonitrile (3 mL) were heated at 808C for 72 h. After cool-
ing at room temperature, the dark red reaction mixture was
diluted with dichloromethane (15 mL) then very well
washed with water (15 mL) to colourless aqueous layer.
The aqueous layers were kept for recycling the PS. After
drying over MgSO4, the dichloromethane solution was
evaporated under vacuum to afford 1.17 g crude reaction
mixture as a red oil that was puri®ed by column chromato-
graphy on 40 g silica gel (eluent pentane/diethyl ether 3:1
v/v). Yield 20% (0.155 g 8a) as a red liquid.
2,3-Dichloroquinoxaline (1.00 g, 5.02 mmol) was dissolved
in acetonitrile (5 mL); to this solution, heated at 808C, PS
(1.08 g, 5.02 mmol) and Et3N´3HF (0.300 mL, 1.80 mmol)
were added during 96 h, as three equal portions (each 48 h).
After an additional 24 h the reaction mixture was cooled at
room temperature, taken up with 15 mL dichloromethane
and the organic layer was washed several times with
water (15 mL). After drying and removing the solvent,
¯ash column chromatography (silica gel, eluent pentane/
chloroform 5:1 v/v) was performed to yield 0.100 g 5b
0.250 g 5a (35% yield) and 0.260 g unreacted 5.
1
4.14.1. 2,6-Di¯uoro-3-nitropyridine, 8a. Red liquid. H
NMR (d ppm, J Hz): 8.69 (1H, ddd, 8.7, 8.7, 6.5),
7.06(1H, dd, 8.6, 3.2); 13C NMR (d ppm, J Hz): 163.1
(1C, dd, 255.6, 13.5), 154.9 (1C, dd, 258.1, 16.6), 142.4
(1C, d, 10.0), 108.3 (2C, dd, 36.3, 6.3); 19F NMR (d ppm,
J Hz): 256.4 (1F, s), 265.2 (1F, s). HR MS for C5H2F2N2O2
160.0798; found 160.0775.
4.12. Preparation of 4-chloro-1-¯uorophthalazine, 6a
(Table 3)
Method D was used, starting from 1,4-dichlorophthalazine
(0.500 g, 2.51 mmol) in acetonitrile (3 mL). The total
amount of the reagent 1PS11/3(Et3N´3HF) was added
portionwise (50, 30 and 20%, respectively, each 24 h) at
1
4.14.2. 6-Chloro-2-¯uoro-3-nitropyridine, 8b. H NMR

750
M. Darabantu et al. / Tetrahedron 57 (2001) 739±750
1
(d ppm, J Hz): 8.44 (1H, dd, 8.9, 6.5), 7.30(1H, dd, 8.1, 2.2);
19F NMR (d ppm, J Hz): 266.6 (1F, s).
Aldrichw product). H NMR (d ppm, J Hz): 2.93 (s, 12H,
±CH3), 7.41 (dd, 2H, 8.8, 1.4), 7.29 (dd, 2H, 8.8, 7.5), 6.98
(dd, 2H, 7.5, 1.4); 13C NMR (d ppm): 44.8 (4C, ±CH3),
113.2 (2C), 121.0 (1C), 122.2 (2C), 125.9 (2C), 138.3
(1C), 151.2 (2 C).
4.14.3. 2-Chloro-6-¯uoro-3-nitropyridine, 8c. 1H NMR (d
ppm, J Hz): 8.54 (1H, dd, 9.3, 8.4), 7.64 (1H, dd, 8.4, 1.0);
19F NMR (d ppm, J Hz): 259.2 (1F, d, 9.4). Only the
distinct peaks were given, as revealed by the crude NMR
spectra.
Acknowledgements
4.15. Preparation of 2-¯uoro-3,5-dinitropyridine
(Scheme 2)
We would like to thank Dr T. C. G. Bird for useful discus-
Â
Â
sion and M. D. thanks the Reseau Interregional Normand de
Chimie Organique Fine (RINCOF) (Contrat de Plan Etat
2-Chloro-3,5-dinitropyridine (0.300 g, 1.47 mmol) was
dissolved in acetonitrile (2 mL) and PS´1HF (0.344 g,
1.47 mmol) was added. The reaction mixture was stirred
Â
Bassin ParisienÐRegions Haute Normandie, Basse
Normandie) for the post-doctoral fellowship.
1
at room temperature for 48 h. H NMR monitoring of the
crude reaction mixture indicated the presence of the starting
material as about 20%. The reaction mixture was diluted
with acetonitrile (2 mL), cooled at 2208C for 12 h then
the solid PS´1HCl was ®ltered off. The organic solution
was evaporated in vacuo without heating to afford 0.250 g
crude product. This was puri®ed by column chromato-
graphy on silica gel (20 g, eluent pentane/acetone 4:1 v/v)
to give the desired product as a red liquid (0.137 g, 50%
yield).
References
1. Yoneda, N. Tetrahedron 1991, 47, 5329±5365.
2. Mc. Clinton, A. M. Aldrichimica Acta 1995, 28, 31±34.
3. Alder, W. R. Chem. Rev. 1989, 89, 1215±1223.
4. Chambers, D. R.; Holmes, F. T.; Korn, R. S.; Sandford, G.
J. Chem. Soc., Chem. Commun. 1993, 855±856.
5. Chambers, D. R.; Korn, R. S.; Sandford, G. J. Fluorine Chem.
1994, 69, 103±108.
4.15.1. 2-Fluoro-3,5-dinitropyridine. Red liquid. 1H NMR
(d ppm, J Hz): 9.36 (1H, dd, 3.2, 2.5), 9.31(1H, dd, 7.5, 2.5);
13C NMR (d ppm, J Hz): 155.5 (1C, d, 258.1), 148.6 (1C, d,
16.9), 147.5 (1C, s), 133.3 (1C, s), 130.1 (1C, s); 19F NMR
(d ppm, J Hz): 257.4 (1F, s). HR MS for C5H2FN3O4
187.0869; found 187.0899.
 Â
6. Ple, N.; Turck, A.; Heynderickx, A.; Queguiner, G.
J. Heterocycl. Chem. 1994, 31, 1311±1315.
 Â
7. Ple, N.; Turck, A.; Couture, K.; Queguiner, G. J. Org. Chem.
1995, 60, 378±3786.
8. Darabantu, M.; Lequeux, T.; Pommelet, J. C.; Ple, N.;
Â
Turck, A.; Toupet, L. Tetrahedron Lett. 2000, 41, 6763±6767.
9. Brzezinsky, B.; Grech, E.; Malarsky, Z.; Sobczyk, L. J. Chem.
Soc., Faraday Trans. 1990, 86, 1777±1780.
1
4.15.2. 2,4,6-Tri¯uoropyrimidine. Colourless liquid. H
NMR (d ppm, J Hz): 6.61 (1H, m); 13C NMR (d ppm, J
Hz): 174.2 (2C, dd, 261.2, 17.2), 174.4 (1C, dd, 243.1,
17.2); 90.9 (1C, ddd, 34.4, 34.4, 7.5) 19F NMR (d ppm, J
Hz): 240.6 (1F, s); 252.3 (2F, s). HR MS for C4HF3N2
134.0605; found 134.0622.
10. Prestch, E.; Seibl, J.; Simon, W. In Tables of Spectral Data for
Structure Determination of Organic Compounds; Springer:
Berlin, Heidelberg, New York, Tokyo, 1983; pp 1±100.
11. Fukuhara, T.; Yoneda, N. Chem. Lett. 1993, 509±512.
12. The intrinsic problems of separation di¯uoro- from chloro-
¯uorodiazines are described since the period of 1970's; see,
for example, Cheeseman, G. W. H.; Godwin, R. A. J. Chem.
Soc. C 1971, 2973±2976.
4.16. Recycling of 1,8-bis(dimethylamino)naphthalene,
`PS'
13. Uchibori, Y.; Umeno, M.; Yoshioka, H. Heterocycles 1992,
38, 1507±1510.
According to Method A, the solid residue after distillation
was dissolved in water (about 10% solid as aqueous solu-
tion). According to Methods B±D, the ®rst aqueous layer
was used; according to Method E, the recovered PS´1HCl
was dissolved in water as about 10% aqueous solution. All
the above aqueous solutions were made alkaline (pH up to
8) with solid Na2CO3 and the deposited solid (crude PS) was
let to stand at room temperature for several hours. After
®ltering, washing with cold water to neutrality and drying
at room temperature, the recovered PS was distilled in vacuo
(1508C/0.5 mmHg) in a kugelrorh instrument.Yield 60±
90%, white crystalline solid; mpˆ50±18C (49±508C
14. Chapman, N. B.; Rees, C. W. J. Chem. Soc. 1954, 1190±1196.
15. Banks, C. K. J. Am. Chem. Soc. 1944, 66, 1127±1130.
16. Maggiolo, A.; Phillips, A. P. J. Org. Chem. 1951, 16, 376±
382.
17. Ishikawa, N.; Kuroda, K.; Onodera, N. Kogyo Kagaku Zasshi
1971, 74, 1490±1491; Chem. Abstr. 1971, 75, p 76709g.
18. Hamer, J. J. Heterocycl. Chem. 1966, 3, 244.
19. Chambers, R. D.; Mc Bride, J. A. H.; Musgrave, W. K. R.;
Reilly, I. S. Tetrahedron Lett. 1970, 1, 57±60.
20. Finger, G. C.; Starr, D. L. J. Am. Chem. Soc. 1959, 81, 2674±
2675.