Multinuclear NMR Studies and Spin System Analyses on Dibenzo[c.e][1,2]Oxaphosphorine 2-Oxides

Article abstract of DOI:10.1002/hc.21313

2H-Dibenzo[c.e][1,2]oxaphosphorine 2-oxide (HDOPO), 2-(N,N-diethylamino)-dibenzo[c.e][1,2]oxaphosphorine 2-oxide (DEADOPO), and 2-ethoxy-dibenzo[c.e][1,2]oxaphosphorine 2-oxide (EtODOPO) are fully characterized in CDCl₃ by ¹H, ¹³C, ³¹P, and ¹⁵N NMR spectroscopy on 800- and 500-MHz instruments. A strategy enabling their unambiguous signal assignment is presented, with special emphasis on 2D ¹H,¹³C HMBC spectra. Additional line-shape iterations of the aromatic ¹H multiplets (ABCDX and ABCDEX spin systems) provided all long-range ⁿJ_H,H and ⁿJ_P,H coupling constants with utmost precision. The experimental results augmented with those of the model compound phenylphosphonous acid clearly demonstrate that ⁿJ_H,H couplings of the PH proton as well as the ⁿJ_P,C values do not decrease in a monotonic manner with the number of intervening bonds from the phosphorus atom. This fact may potentially lead signal misassignments, if the analysis starts out from the coupling constants, as it occurred for HDOPO in the recent publication by Wagner et al. (Phosphorus, Sulfur and Silicon, 187, 2012, 781-798). The corrected assignment will be given in the present paper. Finally, the A₂M₃X or ABM₃X type ¹H spectral patterns of ethyl groups are also analyzed and explicit equations are derived to evaluate the strongly coupled ABM₃X multiplets in EtODOPO.

Full text of DOI:10.1002/hc.21313

Heteroatom Chemistry
Volume 27, Number 3, 2016
M
ultinuclear NMR Studies and Spin System
Analyses on Dibenzo[c.e][1,2]Oxaphosphorine
-Oxides
2
1
2
1
Zolt a´ n Szak a´ cs, Gerhard H a¨ gele, Mikl o´ s D e´ k a´ ny,
3
3
´
P e´ ter Abr a´ nyi-Balogh, and Gy o¨ rgy Keglevich
1
Spectroscopic Research, Gedeon Richter Plc., Gy o¨ mr o˝ i u´ t 19–21, H-1103, Budapest, Hungary
2Institut f u¨ r Anorganische Chemie und Strukturchemie, Heinrich Heine Universit a¨ t D u¨ sseldorf,
Universit a¨ tsstrasse 1, D-40225, D u¨ sseldorf, Germany
3Department of Organic Chemistry and Technology, Budapest University of Technology and
Economics, H-1521, Budapest, Hungary
Received 7 June 2015; revised 13 December 2015; accepted 19 January 2016
1
87, 2012, 781–798). The corrected assignment will
ABSTRACT: 2H-Dibenzo[c.e][1,2]oxaphosphorine
-oxide (HDOPO), 2-(N,N-diethylamino)-dibenzo[c.e]
1,2]oxaphosphorine 2-oxide (DEADOPO), and 2-
ethoxy-dibenzo[c.e][1,2]oxaphosphorine 2-oxide
EtODOPO) are fully characterized in CDCl3 by H,
be given in the present paper. Finally, the A2M3X
or ABM3X type H spectral patterns of ethyl groups
are also analyzed and explicit equations are derived
to evaluate the strongly coupled ABM3X multiplets
in EtODOPO. ꢀC 2016 Wiley Periodicals, Inc. Het-
eroatom Chem. 27:158–172, 2016; View this article
online at wileyonlinelibrary.com. DOI 10.1002/hc.
1 23 13
2
[
1
1
(
1
3
31
15
C, P, and N NMR spectroscopy on 800- and
00-MHz instruments. A strategy enabling their una-
5
mbiguous signal assignment is presented, with special
emphasis on 2D 1H, C HMBC spectra. Additional
13
1
line-shape iterations of the aromatic H multiplets
(
ABCDX and ABCDEX spin systems) provided all
long-range JH,H and nJP,H coupling constants with
utmost precision. The experimental results augmented
with those of the model compound phenylphospho-
n
INTRODUCTION
P-substituted derivatives of the phenanthrene ana-
logue dibenzo [c.e] [1,2]oxaphosphorines 1 and their
2-oxides 2 (Fig. 1) attracted considerable attention
in synthetic, coordination, and applied chemistry.
A few representative examples employed as precur-
sors and building units for coordination compounds,
polymers, flame retardants, and other practical ap-
plications are listed in Table 1.
n
nous acid clearly demonstrate that JH,H couplings
of the PH proton as well as the JP,C values do not
n
decrease in a monotonic manner with the number of
intervening bonds from the phosphorus atom. This
fact may potentially lead signal misassignments, if
the analysis starts out from the coupling constants,
as it occurred for HDOPO in the recent publication
by Wagner et al. (Phosphorus, Sulfur and Silicon,
The identity and purity of such compounds
have been previously characterized using NMR spec-
1
13
31
troscopic methods by observing H, C, and
P
1
Correspondence to: Zolt a´ n Szak a´ cs; e-mail: z.szakacs@richter.
hu; Gerhard H a¨ gele; e-mail: haegele@uni-duesseldorf.de
Supporting Information is available in the online issue at
www.wileyonlinelibrary.com.
nuclei. H NMR studies often revealed crowded,
overlapping spectral patterns for the aromatic pro-
tons, which quite frequently were described as “com-
plex multiplets,” without attempting a more detailed
ꢀC
2016 Wiley Periodicals, Inc.
158

Multinuclear NMR Studies and Spin System Analyses of DOPO Compounds 159
FIGURE 1 General formula for dibenzo[c.e] [1, 2]oxaphosphorines 1 and their 2-oxides 2. For the latter type, the skeletal
atoms are numbered (and the benzene rings are color-coded in the online publication) to aid comprehension.
TABLE 1 Literature Overview of Some Representative
Dibenzo[c.e] [1,2]Oxaphosphorines 1 and their 2-Oxides 2
of this type (2a, 2b, and 2c in Scheme 1) at higher
magnetic field strengths in order to determine pos-
sibly all long-range coupling constants, to identify
the pitfalls of previous assignment strategies, and to
substitute them with more robust approaches.
The present article describes our complete
R Group in Compounds 1
R Group in Compounds 2
Cl [1–5]
–
–
H [2,7]
Ph [4,6], iPr
3
Ph [6], Me [8]
3
Ph [4,6], iPr Ph [6], Me [8],
1
multinuclear assignment and H spin analysis of
Et [8], Pr [8]
NHCHMePh [4,9]
NEt [3,4,9]
NHCHMePh [4,9]
NEt [3,4,9]
2
OH [10]
OEt [3,4,10,11],
OMe [4,11],OiPr [4,11],
OiBu [4,11],
2a, 2-N,N-diethylamino-dibenzo[c.e] [1, 2]oxaphos-
phorine 2-oxide (DEADOPO) 2b [3, 4], and
2
–
2-ethoxy-dibenzo[c.e] [1, 2]oxaphosphorine 2-oxide
OEt [4,11], OMe [4,8,11],
OiPr [4,11], OiBu [4,11],
OMenth [4,9], OPh [1],
OoXyl [12], OMes [1]
(EtODOPO) 2c [3, 4, 9–11]. These substances were
synthesized in good yields from the commercially
available 2a according to procedures summarized
in Scheme 1.
OMenth [4,9]
spectral analysis. Systematic determination and in-
terpretation of long-range homo- and heteronuclear
RESULTS AND DISCUSSION
n
n
n
coupling constants ( JH,H, JP,H, and/or JP,C) were
undertaken in a few cases only [13,14], revealing that
magnitudes of heteronuclear couplings do not nec-
essarily decrease with increasing distance from the
phosphorus atom, hence their utilization for assign-
ment purposes should be exercised with caution. A
The Complete Multinuclear NMR Assignment of
DEADOPO 2b
1
The H NMR spectrum of 2b comprises seven groups
of signals for the eight aromatic protons H7–H14
(Fig. 2), since the two multiplets near 7.22 ppm
overlap even at 800 MHz measuring frequency (see
Fig. S1 in the Supporting Information). The
following shorthand notation will be used in the
forthcoming discussion: H(7.79) denotes the multi-
plet centered at 7.79 ppm.
1
13
closer inspection of the published H/ C NMR data
of 2H-dibenzo[c.e] [1, 2]oxaphosphorine 2-oxide 2a
(HDOPO; see Scheme 1) by Wagner et al. [14] in-
deed revealed some inconsistencies and mistakes.
This inspired us to reinvestigate three compounds
SCHEME 1 Synthesis of DEADOPO 2b and EtODOPO 2c.
He te ro at om Ch em i s try DOI 10.1002/hc

ꢀ
1
60 Szakacs et al.
FIGURE 2 Aromatic region of the 500 MHz 1H NMR spectrum of DEADOPO 2b. (a) Experimental spectrum (* denotes
residual CHCl from the CDCl solvent), (b) the iterated spectrum with signal assignments (see text for their derivation).
3
3
FIGURE 3 Phase-sensitive double-quantum filtered COSY spectrum of DEADOPO 2b.
In the COSY [15] spectrum shown in Fig. 3, the
intense cross-peaks identify vicinal protons on the
“backwards,” H10 to H7) or H11 to H14 (or “back-
wards”) on the DEADOPO skeleton (Fig. 1). Belong-
ing to a separate spin system, the signals H(7.95),
H(7.22), H(7.35), and H(7.22) are due to adjacent
protons on the other aromatic ring, with assign-
ments again being possible in two ways. Thus, there
are altogether eight possible assignments of the
3
benzene ring, based on the well-known rule JH,H >
4
5
JH,H > JH,H. The signals H(7.99), H(7.65), H(7.48),
and H(7.79) in this order arise from adjacent pro-
tons on one benzene ring and they may be in prin-
ciple assigned to either the positions H7 to H10 (or
H e t e r o a t o m C h e m i sD t rO yI 10.1002/hc

Multinuclear NMR Studies and Spin System Analyses of DOPO Compounds 161
1
FIGURE 4 (a) Reference 500 MHz H NMR spectrum of DEADOPO 2b and its selective 1D NOESY spectra (b) by exciting
the multiplet at 7.99 ppm or (c) the multiplet at 7.95 ppm.
aromatic protons, deliberately not considering at
this point the signal multiplicities, since the mag-
nitudes of coupling constants can be misleading re-
garding their covalent distance from the phosphorus
atom, as will be detailed below.
of oxygen and phosphoryl substituents [14]; observ-
ing δH trends in related structures; chemical shift
prediction by suitable computer programs or analyz-
ing the experimental long-range phosphorus-proton
n
coupling constants ( JP,H) [14]. The mentioned ap-
1
The next straightforward step towards the full
H assignment may be the observation of the nuclear
proaches are commonly applied for H NMR assign-
1
ment of organophosphorus compounds, but may oc-
casionally lead to mistakes [14]. We thus followed a
more general and robust strategy by turning our at-
Overhauser enhancement (NOE) between H10 and
H11, since their distance is estimated to be less than
1
3
˚
5
A from previous structural studies on biphenyls or
phenanthrenes. The selective 1D NOESY experiment
16] is a particularly useful and time-efficient tech-
tention to the experimental C chemical shifts and
their one-bond (HSQC) and multiple-bond (HMBC)
connectivities to protons [15].
[
nique for this purpose, but it holds the prerequisite
that the multiplet in question has to be selectively ex-
cited. From the four possible “candidate” signals for
the H10 position, only H(7.99), H(7.95), or H(7.79)
can be selectively excited by a shaped radiofrequency
pulse, since irradiating the multiplet near 7.22 ppm
would hit two protons. When performing the three
respective selective NOESY1D experiments, a NOE
was detected on H(7.99) upon inversion of the mul-
tiplet H(7.95) and vice versa (see Fig. 4). A NOE
between two so closely spaced signals would have
been difficult to detect near the “thick” diagonal of a
The proton-decoupled 13C NMR spectrum of 2b
consists of twelve aromatic signals, most of them are
n
doublets due to the JP,C coupling (see Table 2 and
Fig. S2 in the Supporting Information). Instead of
interpreting these coupling constants to guide the
carbon signal assignment, only two basic presump-
tions are made at this point. First, an HMBC spec-
n
trum optimized to JC,H ࣈ 8 Hz coupling constant
yields the most intense correlations for C,H pairs
separated by three (or, less probably, two) covalent
bonds. In our practice of organic structure elucida-
tion, this has proven to be a very robust criterion.
The second assumption is the actual starting point of
the assignment: the most deshielded carbon signal at
150.3 ppm belongs to C6 bonded to the oxygen atom.
This chemical shift value is characteristic of ethers
and phosphate esters of phenols [17]. Inspecting the
bottom row of the HMBC spectrum in Fig. 5, this
C(150.3) shows correlations with H(7.95), H(7.35),
and one of the H(7.22) signals. Combining this with
2D NOESY spectrum. Hence, these two signals be-
long to the structural positions H10 and H11, reduc-
ing the eight possibilities to two alternative assign-
ments. More information cannot be extracted from
solely proton-based NMR experiments, hence fur-
ther presumptions are required to resolve this ambi-
1
guity. The interpretation of H chemical shifts may
involve, for example, speculating on inductive effects
He te ro at om Ch em i s try DOI 10.1002/hc

ꢀ
1
62 Szakacs et al.
13
1
n
TABLE 2
C and H NMR Assignments and Magnitudes of the Heteronuclear JP, C Coupling Constants of DEADOPO 2b
Carbon #
δ
C
(ppm)
nJP, C (Hz)
At 126 MHz
At 201 MHz
Proton #
δ
H
(ppm)a
1
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
Ethyl CH
Ethyl CH
124.7
137.4
121.8
150.3
120.6
130.1
124.0
124.8
123.5
132.5
128.1
129.9
38.7
J
J
J
J
J
J
J
J
J
J
J
J
J
J
166.4
7.0
11.6
7.7
166.5
7.1
11.6
7.8
6.3
–
–
–
–
–
–
–
7.22
7.35
7.22
7.95
7.99
7.65
7.48
7.79
2
3
2
–
3
6.4
H7
H8
H9
H10
H11
H12
H13
H14
4
Not resolved
0.7
Not resolved
Not resolved
1.2
5
4
1.1
11.4
2.6
14.6
9.3
5.1
3
11.3
2.5
14.5
9.3
5.2
2.3
4
3
2
2
2
Ethyl CH
Ethyl CH
3
2
ca. 3.15
1.15
3
3
14.1
2.3
a
1
Determined from multiplet positions in the H or the HSQC spectrum; for more accurate values by line-shape iteration of the proton spectrum,
see Tables 3 and 4.
The proton-coupled 31P spectrum of DEADOPO
(Fig. S3 in the Supporting Information) reveals its
the COSY information, these signals are assigned
to H10, H8, and H7, respectively. The HSQC spec-
trum in Fig. 6 gives δC values of these CH moieties
3
1
15
1
P chemical shift δ to be 16.4 ppm, while a N, H
P
(
Table 2). From COSY, the fourth proton in this spin
HMBC spectrum (Fig. S4 in the Supporting Informa-
1
5
system also resonates at 7.22 ppm; it is attached to
C(124.0) from HSQC and its HMBC correlation over
three bonds identifies C5 at 121.8 ppm. The HMBC
correlation between C(121.8) and H(7.99) clearly
identifies H11 of the P-substituted ring. The assign-
ment of this latter proton signal is also available from
the already discussed 1D NOESY experiments, but
HMBC represents a more general and robust tool
tion) yields the N chemical shift δ of −324.8 ppm.
N
1
Splitting of the latter correlation peak due to J
P,N
could not be observed, since this coupling constant
is expected to be less than 2 Hz for tetrahedral phos-
phorus atoms [18].
Line-Shape Iterations on the Aromatic Proton
Signals of DEADOPO 2b
1
for H signal assignment around biphenyl junctions
(the applicability of 1D NOESY critically depends on
1
n
n
spectral overlap, determined in part by the H NMR
measurement frequency). By using H11 as an an-
chor point and combining information from COSY
and HSQC spectra, δH and δC are determined for
CHs in ring positions 12, 13, and 14 (see Table 2).
The quaternary carbon signal at 137.4 ppm shows
correlations to H(7.95), H(7.79), and H(7.65), hence
it belongs to C4. Finally, C(124.7) shows HMBC cor-
relations to H(7.99) and H(7.48), thus attributed to
C3.
Most of the JH,H and JP,H coupling constants ex-
ceeding 1 Hz can be evaluated directly as differences
1
of the appropriate line frequencies in the H NMR
multiplets. However, couplings smaller than the line
width cannot be determined in this way and, in addi-
tion, the frequency difference of H9 and H7 is small
4
compared to their JH,H coupling, rendering this pair
of protons strongly coupled. The ABCDX spin sys-
3
1
tem (X = P) in the O-substituted ring thus approx-
imates ABCC’X, the evaluation of which requires a
total line shape fitting approach by computer. Before
The phosphorus-carbon coupling constants of
b, calculated from its 1D 13C spectra are also
1
2
submitting the 500 and 800 MHz H spectra to com-
summarized in Table 2. Their magnitude does not
decrease monotonically with the covalent distance
from the phosphorus atom, so their order is in-
formative for future or retrospective studies on
dibenzo[c.e] [1,2]oxaphosphorine 2-oxides:
puter evaluations, it was necessary to judge whether
the two rings can be treated separately (as isolated
ABCDX and ABCC’X five-spin systems), or a more
time-consuming full 9-spin system (ABCDKLMNX)
iteration should be carried out. The H10, H11, H7,
or H14 signals are not appreciably more broadened
than the others (as justified a posteriori by our total
line-shape analyses at both 500 and 800 MHz, see
Table S1 in the Supporting Information), so the
inter-ring coupling constants are unresolved in the
1
3
3
3
JP,C3 >> JP,C13 > JP,C5 ∼ JP,C11 >
2
4
2
2
3
JP,C14 > JP,C6 ∼ JP,C4 ∼ JP,C7 >>
4
4
5
JP,C12 > JP,C10 ∼ JP,C9 ∼ JP,C8
H
e
t
e
r
o
a
t
o
m
C
h
e
m
i
s
t
r
y
D
O
I
1
0
.
1
0
0
2
/
h
c

Multinuclear NMR Studies and Spin System Analyses of DOPO Compounds 163
1
13
FIGURE 5 Band-selective H, C HMBC spectrum of 2b at 500 MHz.
1
13
FIGURE 6 500 MHz H, C HSQC spectrum of 2b, aromatic region.
√
zCOSY spectra (see Figs. S5 and S6 in the Support-
ing Information) indeed reveal a weak coupling be-
tween the signals of H10 and H11, but its magnitude
could not be extracted. Hence, the two sets of CH
1
D spectrum (J < HW / 3). For structurally re-
lated biphenyl derivatives, none of the literature re-
ported on long-range inter-ring coupling constants
19]. 2D correlation spectra are more sensitive to de-
[
3
1
protons together with the P nucleus can safely be
tect long-range couplings. COSY and (E-COSY type)
He te ro at om Ch em i s try DOI 10.1002/hc

ꢀ
1
64 Szakacs et al.
TABLE 3 Iterated NMR Parameters of Aromatic Protons in
DEADOPO 2b
3.16 ppm, in which 23 lines from a theoretical max-
imum of 32 are resolved at 800 MHz. The M-part
of this ABM3X system gives rise to a simple triplet
At 500 At 800
At 500 At 800
4
Parametera MHz
MHz
Parameter MHz
MHz
(not shown), where JM,X ( JP,H) remains unresolved.
Exact spectral parameters for this methyl group
are extracted by deconvolution, while the methy-
lene group is dealt with by iteration. The calculated
spectra are depicted in Fig. 7, while the resulting
parameters are listed in Table 4 (see Table S2 in
the Supporting Information for complete iteration
results).
δ
δ
δ
H10
H9
7.9482 7.9488
7.2234 7.2249
7.3501 7.3513
7.2198 7.2206
δ
δ
δ
H11
H12
H13
H14
7.9931 7.9936
7.6534 7.6550
7.4775 7.4787
7.7875 7.7859
H8
δ
H7
δ
3
3
J
J
J
J
J
J
J
J
J
J
H7,H8
8.16
7.30
7.98
1.27
1.58
0.42
0.44
1.12
0.53
0.47
8.08
7.34
7.86
1.16
1.51
0.41
0.35
1.04
0.49
0.45
J
J
J
J
J
J
J
J
J
J
H13,H14
7.61
7.37
8.16
1.45
1.06
0.63
14.55
3.02
5.80
1.11
7.58
7.33
8.08
1.39
0.96
0.55
14.53
3.03
5.79
1.02
3
3
4
4
5
4
5
5
6
3
3
4
4
5
3
4
4
5
H8,H9
H9,H10
H7,H9
H8,H10
H7,H10
P, H 7
H12,H13
H11,H12
H12,H14
H11,H13
H11,H14
P, H 1 4
From the aspects of symmetry and molecu-
1
lar dynamics, it is instructive to compare the
H
NMR properties of ethyl protons in the phospho-
namidate 2b with those in N,N-diethylbenzamide
(3 in Fig. 8; [20–23]) and its ortho (di)substituted
derivatives [21,24,25] measured in various solvents
and at various temperatures. These studies con-
tributed profoundly to the understanding of decep-
tively simple spectra of ethyl groups that often have
anisochronous protons: the ethyl moiety in N,N-
diethylbenzamides exhibit either A B type signals
P, H 8
P, H 1 3
P, H 1 0
P, H 9
P, H 1 1
P, H 1 4
aChemical shifts δ in ppm, coupling constants J in Hz.
treated as two disjunct 5-spin systems in our calcu-
lations.
2
3
Line shape iterations require rather close initial
estimates for the parameters to be refined. The chem-
ical shifts and the manually derived J > 1 Hz cou-
pling constants served as good starting points. Total
line shape fittings by the WINDAISY module of the
TopSpin software yielded the calculated spectrum
2 3 2 3
or overlays from A B and C D systems, distin-
guished according to the Z or E position of the
ethyl group with respect to the oxygen atom (see
Fig. 8). Obviously, the polarized mesomeric struc-
–
+
ture of O–C=N contributes significantly to the
carbon-nitrogen bond order in the hindered rota-
tional equilibrium. Methylene protons within the A2
1
shown in Fig. 2b and the H parameters summarized
in Table 3 (for complete results including line widths
and statistical descriptors, see Table S1 in the Sup-
porting Information). NMR parameters from the 500
(or C ) groups are found to be magnetically equiva-
2
3
lent with typical J
H,H
coupling constants JA,B (JC,D)
of 7.1 Hz.
1
and 800 MHz H spectra show excellent agreement.
In contrast to carboxamides, in the proton spec-
The magnitudes of the heteronuclear coupling con-
trum of the P-chiral DEADOPO 2b only one ABC X
3
3
4
5
6
stants follow the order: JP,H >> JP,H > JP,H ࣈ JP,H,
spin system is detected at ambient temperature and
4
the single exception being the low JP,H7 = 0.4 Hz
not the overlay of different ABC X and DEF X sys-
3
3
value, presumably due to electronic effects of the in-
tervening oxygen atom.
tems. This fact implies fast rotation around the P–N,
N–C, and C–C bonds of the P(O)N(CH CH ) part.
2
3 2
It seems justified to assume that the polarized struc-
–
+
ture II O–P=N is less significant with respect to I or
III in Fig. 8. Nevertheless, the solid-state structure [9]
of DEADOPO 2b features different ethyl groups oc-
cupying the E- and Z-positions.
Line-Shape Analysis of the N,N-Diethyl Group
Signals in DEADOPO 2b and Comparison with
Spectral Patterns of Other Ethylamides
The 1H NMR multiplet pattern of the P(O)
N(CH2CH3)2 moiety in DEADOPO approximates
that of an ABM3X spin system. In the forthcom-
ing discussion, capital letters indicate the AB part
of the total spin system, while lower case letters ab
are defined for sub-systems or sub-sub-systems (for-
mally indicating the AB-character of spectral parts)
in NMR sub-spectral analysis (see Supplementary in-
formation, Pages 11–15 for details). Figure 7 shows
the superposition of 8 ab type sub-subspectra con-
stituting the symmetrical CH2 multiplet centered at
A final comparison includes the P-chiral phos-
phinic N,N-diethylamide RR’P(O)N(CH2CH3)2
4
1
with R=C6H5 and R’=2-HO-C6H4. Its H NMR spec-
trum was described [26] as a doublet of quartets at
3
.08 ppm (CH2; JA,M = 7.1 Hz, JA,X = 11.4 Hz) and
as a triplet at 1.11 ppm (CH3; JA,M = 7.1 Hz, JM,X not
resolved). Hence, an A2B3X/A2M3X type spectrum
is observed for 4 instead of ABC3X/ABM3X as found
for DEADOPO 2b, which indicates a stronger “chiral
induction” effect of the methylene non-equivalence
at work in the latter compound.
H
e
t
e
r
o
a
t
o
m
C
h
e
m
i
s
t
r
y
D
O
I
1
0
.
1
0
0
2
/
h
c
.

Multinuclear NMR Studies and Spin System Analyses of DOPO Compounds 165
1
FIGURE 7 The CH
2
multiplet of DEADOPO 2b in the following H NMR spectra: (a) 500 MHz experimental, (b) 500 MHz
simulated, (c) 800 MHz experimental, and (d) 800 MHz simulated.
TABLE 4 Iterated Parameters for the Methylene (A,B) and
0.2 ppm, forming thus a congested multiplet with
H11. H10 resonates within 0.03 ppm from the latter
signal, so the NOESY1D-based distinction of H10
and H11 is not practicable here even at 800 MHz.
Methyl (M
3
) Protons of DEADOPO 2b
At 500 At 800
At 500
At 800
MHz
Parametera MHz
MHz Parameter MHz
Nevertheless, HSQC still offers a reliable identifica-
2
δ
δ
δ
HA
HB
HC
3.1853 3.1873
3.1304 3.1306
1.1494 1.1510
11.75 11.73
J
J
J
J
HA,HB
HA,HC
HB,HC
P, H C
−14.43 −14.46
13
tion of these protons by utilizing their different
C
3
3
4
7.09
7.05
7.06
7.03
0.40
shifts, and HMBC corroborates the COSY-based se-
quence of protons on both rings: the ipso carbon
C5 at 122.6 ppm is correlated via 3 bonds with H7
(7.25 ppm), H9 (7.28 ppm), and H11 (7.97 ppm),
while three correlation peaks of H10 (7.94 ppm),
H12 (7.72 ppm), and H14 (7.98 ppm) define C4 at
3JP, H A
Not
resolved
–
3JP, H B
12.24 12.23
–
–
aChemical shifts δ in ppm, coupling constants J in Hz.
1
13
1
37.0 ppm. The complete H and C assignment is
The Complete Multinuclear NMR Assignment
and Aromatic Line-Shape Analysis of EtODOPO
summarized in Table 5. Most of these NMR param-
eters are highly similar to those in 2b (δH agrees
n
2c
within 0.07 ppm, δ within 1 ppm, and J within
C
P,C
1
1 Hz), the exceptions being located near the phos-
phorus atom: the chemical shifts of the ethyl group;
JP,C is increased to 181.7 Hz and the already men-
The aromatic multiplets in the H NMR spectrum
of 2c (Fig. 9) are assigned on the basis of 2D
COSY, HSQC, and HMBC spectra (see Figs. S7–
S9 in the Supporting Information). In contrast to
DEADOPO 2b, H7 and H9 show no signal overlap
here. Instead, the H14 signal is shifted downfield by
1
tioned δH of H14. To complete the multinuclear as-
3
1
signment of 2c, its P chemical shift is determined
to be 10.1 ppm.
FIGURE 8 Plausible mesomeric structures in N,N-diethylbenzamides 3 (upper row) and in DEADOPO 2b (lower row).
H et ero at om C he mi s try DOI 10.1002/hc.

ꢀ
1
66 Szakacs et al.
1
3
FIGURE 9 Aromatic region of the 800 MHz H NMR spectrum of EtODOPO 2c. (a) Experimental spectrum (*CHCl ), (b)
iterated spectrum with signal assignments.
1
3
1
TABLE 5
C and H NMR Assignments and Magnitudes of
TABLE 6 Iterated NMR Parameters of Aromatic Protons in
EtODOPO 2c
n
the Heteronuclear JP, C Coupling Constants (from 201 MHz
C Spectrum) of EtODOPO 2c
13
At 500 At 800
Parametera MHz
MHz Parameter
At 500 At 800
MHz MHz
nJP, C
value
(Hz)
δ
C
Carbon # (ppm) nJP, C
(ppm)a
δ
H
δ
δ
δ
H10
7.9443 7.9454 δH11
7.2788 7.2811 δH12
7.3926 7.3745 δH13
7.9735 7.9744
7.7238 7.7265
7.5254 7.5275
7.9788 7.9794
Proton #
H9
1
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
Ethyl CH
Ethyl CH
122.4
137.0
122.6
149.9
120.2
130.5
124.7
125.2
124.0
133.4
128.3
130.2
63.0
J
J
J
J
J
J
J
J
J
J
J
J
J
J
181.7
–
–
–
–
–
–
–
7.25
7.39
7.28
7.94
7.97
7.72
7.53
7.98
H8
2
3
2
3
4
5
4
3
4
3
2
2
3
δ
7.2434 7.2447 δ
H14
6.9
12.0
7.8
H7
3
3
JH7,H8
JH8,H9
JH9,H10
JH7,H9
JH8,H10
8.19
7.31
7.99
1.32
1.63
0.40
0.32
0.58
1.19
0.64
8.12
7.28
7.92
1.25
1.55
0.43
0.39
0.63
1.09
0.57
JH13,H14
JH12,H13
JH11,H12
JH12,H14
JH11,H13
7.58
7.43
8.16
1.52
1.08
0.69
14.66
6.21
3.59
1.07
7.53
7.42
8.06
1.42
1.00
0.59
14.65
6.27
3.56
0.98
3
3
4
4
5
4
5
5
6
3
3
4
4
5
3
4
4
5
–
6.5
H7
Not resolved H8
Not resolved H9
JH7,H10
JH11,H14
1.3
12.2
2.5
15.5
9.2
H10
H11
H12
H13
JP, H 7
JP, H 1 4
JP, H 1 0
JP, H 8
JP, H 9
JP, H 1 1
JP, H 1 3
JP, H 1 4
H14
2
3
6.5
5.9
Ethyl CH
Ethyl CH
2
ca. 4.23
1.28
aChemical shifts δ in ppm, coupling constants J in Hz.
16.3
3
aDetermined from multiplet positions in the 1H or the HSQC spec-
trum; for more accurate values by line-shape iteration of the proton
spectrum, see Tables 6 and 7.
The Ethoxy Group in EtODOPO 2c: Analyzing a
Strongly Coupled ABM X Spectrum and
3
Comparison with Related Ethylesters
1
Line shape iterations of the H spectrum yielded
The OCH2CH3 groups of 2c at 500 MHz give rise
to a deceptively simple A2M3X spectral pattern.
Figure 10a shows a doublet of quartets for the A2
methylene protons, so these spins appear to be mag-
netically equivalent at 500 MHz. The M3 signal of the
methyl protons is a simple triplet (not shown). The
manual or computer analysis of the A2M3X multi-
plets is straightforward and the results are summa-
rized in Table 7, while the fitted spectrum is depicted
in Fig. 10b.
the fitted spectrum shown in Fig. 9b (see Fig. S11
in the Supporting Information for the experimen-
tal and calculated 500 MHz H spectra) and the re-
sulting chemical shifts and coupling constants are
summarized in Table 6 (see Table S3 in the Support-
ing Information for more detailed iteration results).
The 500 MHz zCOSY spectrum of 2c (Fig. S11 in
the Supporting Information) confirms that coupling
1
n
constants JP,H have identical signs.
H e te ro a to m C h e m is try DOI 10.1002/hc.

Multinuclear NMR Studies and Spin System Analyses of DOPO Compounds 167
FIGURE 10 The CH
2
multiplet of EtODOPO 2c in the (a) 500 MHz experimental spectrum, (b) 500 MHz simulated spectrum,
(
c) 800 MHz experimental spectrum, and (d) 800 MHz simulated spectrum.
TABLE 7 NMR Parameters of the A
served at 500 MHz and the ABM X Spin System Observed
Protons in EtODOPO 2c
2
M
3
X Spin System Ob-
for these auxiliary parameters and further calcula-
tions to arrive at the targeted chemical shift and
homo- and heteronuclear coupling constant values
are given on pages 11–15 of the Supporting Infor-
mation. The NMR parameters emerging from this
manual evaluation are listed in Table 7, and served
as a starting point of the subsequent computer iter-
ations.
3
2 3
at 800 MHz for the OCH CH
At 500
MHz
DAISY
At 800
MHz
Manual
Calc.
At 800
MHz
DAISY
Parametera
Evaluated
by
Parametera
Evaluated
by
b
c
Iterations
Iterations
δ
δ
HA
4.2252
–
1.2812
–
7.09
–
9.03
–
δ
δ
δ
HA
HB
HM
4.2284
4.2246
1.2816
−10.24
7.09
4.2286
4.2245
1.2816
–
–
Since essential information is hidden in weak
lines in this particular problem, initial parameter
estimates reasonably close to the true solution are
essential for the convergence of the intensity-based
line shape iterator in DAISY. Otherwise calculations
may stop in local minima instead of locating the
global minimum of the error hypersurface. This lat-
ter target function is the root-mean-square deviation
of the experimental and calculated spectra, spanned
over the multidimensional space of the iterated pa-
rameters. In principle, an auxiliary frequency-based
iterator might be a suitable tool in such situations,
but the missing lines in our particular case would
pose difficulties even for those programs. After some
time-consuming calculations, DAISY produced the
NMR parameters listed in Table 7 (see Table S5 in
the Supporting Information for more details) and
the calculated spectrum shown in Fig. 10d. In com-
parison with the results of manual evaluation, only
HM
2
J
J
J
J
J
J
HA,HB
−10.24
3JHA,HM
3
3
3
3
4
7.09
7.15
9.49
8.48
0.30
HA,HM
HB,HM
P, H A
–
7.19
7.37
3JP, H A
–
P, H B
10.39
0.30
4JP, H M
0.30
d
d
P, H M
a
Chemical shifts δ in ppm, coupling constants J in Hz.
Global line width used in the simulations: 1.06 Hz.
Line widths used in the simulations: 0.95 Hz for CH2 and 1.38 Hz
b
c
for CH3 (by deconvolution).
d
Estimated from the CH3 line width.
Nevertheless, the 800 MHz proton spectrum
of 2c already reveals the non-equivalence of
methylene protons, since an ABC3X pattern approx-
imating the ABM3X case is observed (Fig. 10c). In
principle, the sub-spectral breakdown of an ABM3X
spectrum leads to a maximum of 32 lines due to 8 ab-
type sub-sub-spectra which may be analyzed manu-
ally without support of suitable computer programs.
But the situation is unfavorable for the ethyl groups
in EtODOPO, since only the two most intense in-
ner lines are observable for both ab sub-sub spectra,
most of the outer lines are obscured or very weak.
This intriguing problem of NMR spectral analysis
could be solved in a two-step approach, a manual
evaluation followed by computer line-shape analy-
sis.
3
3
the values of J
P,HA
and J
P,HB
have changed signifi-
cantly by computer iterations.
As an a posteriori corroboration of our results,
highly similar homo- and heteronuclear coupling
constants have been found for the P(O)(OCH CH )
2
3 2
groups of the C-chiral α-aminophosphonate es-
ters RNH-R’CH-P(O)(OCH CH ) (5: R = p-C H F,
2
3 2
6
4
R’ = 1-naphthyl; 6: R = 1-naphthyl, R’ = p-
C H F; see Table S6 in the Supporting Informa-
6
4
tion citing their iterated spectral parameters from
First, equations were applied to calculate key
auxiliary parameters of this spin system (Z, NM, NX,
Ref.27) and the P-chiral phosphinates Cl C = CH-
2
CH -PR(O)OCH CH (7: R = CH ; 8: R = C H ; see
2
2
3
3
6
5
ꢀ, LM, LX, and J) from the line frequencies observed
Table S7 in the Supporting Information citing their
spectral parameters from Ref.28).
in the ab multiplets. The lengthy defining equations
H e te ro at om Ch em i s try DOI 10.1002/hc

ꢀ
1688 Szakacs et al.
1
FIGURE 11 Aromatic region of the 800 MHz H NMR spectrum of HDOPO 2a. (a) Experimental spectrum (*residual CHCl
b) the iterated spectrum with signal assignments.
3
),
(
1
13
FIGURE 12 Band-selective H, C HMBC spectrum of 2a at 800 MHz.
Revision of the Multinuclear NMR Assignment
of HDOPO 2a
to C6 using the same arguments as in the case of
2b. C(148.2) is, in turn, correlated over two or three
bonds to H(7.29), H(7.42), and H(7.94), identifying
H7, H8, and H10. Three types of COSY spectra (in
Figs. S13–S15 in the Supporting Information) con-
firm this sequence of protons and HSQC (Fig. S16
in the Supporting Information) reveals the chemi-
HDOPO 2a exhibits mostly separated multiplets at
800 MHz proton frequency (Fig. 11), signal over-
laps occur only near 7.94 and 7.30 ppm. The obvious
starting point of the analysis is the PH doublet cen-
tered at 8.07 ppm. Following the robust heteronu-
clear assignment strategy detailed for 2b above, the
HMBC spectrum of 2a in Fig. 12 reveals correla-
tions of this PH proton over three bonds to C(135.9)
and C(148.2). The latter chemical shift is attributed
1
3
cal shifts of the direct bonded C nuclei, also en-
abling distinction of the overlapping signals of H7
and H9 near 7.30 ppm by virtue of their differing
1
3
C chemical shift. HMBC correlations from these
protons identify C5 at 122.0 ppm, the latter giving
H e te ro a to m C h e m is try DOI 10.1002/hc.

Multinuclear NMR Studies and Spin System Analyses of DOPO Compounds 169
TABLE 8 1H and 13C NMR Assignments and Magnitudes of the Heteronuclear nJP, C Coupling Constants of HDOPO 2a
Our Assignment
Assignment in Ref.14
δ
C
(ppm) nJP, C nJP, C Value (Hz) nJP, C Calc’d.c
Position
δ
H
(ppm)a
δ
C
(ppm) nJP, C
nJP, C Value (Hz)b
δ
H
(ppm)
1
1
2
3
2
3
4
5
4
3
4
3
2
C3
C4
C5
C6
CH7
CH8
CH9
CH10
CH11
CH12
CH13
CH14
PH
–
–
–
–
7.29
7.42
7.30
7.94
7.99
7.76
7.57
7.92
8.07
123.6
J
J
J
J
J
J
J
J
J
J
J
J
122.3
6.5
12.2
8.4
–
–
–
–
7.23
7.33
7.21
7.84
7.81
7.47
7.66
7.89
8.04
123.6
J
J
J
J
J
J
J
J
J
J
J
J
122.1
6.5
12.2
8.4
126.5
137.9
124.0
152.2
120.1
131.3
123.9
125.0
123.8
134.6
128.5
130.3
–
2
3
2
3
4
5
4
3
4
3
2
135.9
122.0
148.2
120.7
130.9
125.1
125.2
124.0
134.1
128.6
130.6
–
135.9
122.1
148.2
120.7
131.0
125.3
125.2
130.7
128.8
134.2
124.1
–
6.5
6.3
d
n. r.
n. r.
n. r.
n. r.
12.7
14.3
2.4
n. r.
1.0
9.7
2.5
14.5
12.8
–
9.7
–
–
–
aDetermined from multiplet positions in the 800 MHz 1H or the HSQC spectrum; for more accurate values by line-shape iteration of the proton
spectrum, see Table 9.
b
Determined from the 201 MHz 13C spectrum.
c
Calculated values from Ref.14, but assigned to 13C nucleus by our analysis.
d
Not resolved.
a further HMBC correlation to H(7.99), which must
correspond to H11. The spatial proximity of H11 and
H10 is also corroborated by a 1D NOESY spectrum
attributed to the “nearest” ring proton H14 over
three bonds. This misleading correlation was also
observed in our gCOSY and zCOSY spectra (but not
in the DQF-COSY; cf. Figs. S13–S15 in the Support-
ing Information) and its origin is clearly identified in
our detailed spin analysis below. Further interpre-
tations of the COSY and HSQC correlations of the
P-substituted ring were otherwise logically correct
in Ref.14. This case is an instructive example how
a less robust element in the interpretational strat-
egy (specifically, a misinterpreted long-range COSY
(Fig. S17 in the Supporting Information), similarly
to 2b. Then starting off from H(7.99), its COSY corre-
lations readily identify the neighboring protons H12,
H13, and H14 at 7.76, 7.57, and 7.92 ppm, respec-
1
3
tively, while HSQC yields their corresponding
C
chemical shifts. Finally, C3 is attributed to the sig-
1
nal at 123.6 ppm with the largest JP,C splitting, thus
1
13
completing the H and C assignment of HDOPO,
1
3
1
13
as summarized in Table 8. The observed C chem-
ical shifts and the JP,C values are highly similar to
correlation) may scramble the correct H and
C
n
assignment in a polyaromatic compound. It is inter-
1
3
those observed for 2b.
esting to note that Wagner’s predicted C chemical
shifts by IGLO-based calculations [14] show a much
better agreement with our corrected assignment (es-
pecially for C11 to C14), than with the erroneously
assigned experimental values given in Ref.14.
The assignment of 2a published by Wagner
et al. [14] is also listed in the last four columns of
1
3
Table 8 for comparison. It is obvious that the
C
chemical shifts at positions 11 to 14 are reversed with
1
3
respect to our assignment. The remaining C chemi-
cal shifts coincide with ours within 0.2 ppm, while δH
are consequently smaller by 0.03 to 0.10 ppm. This
latter discrepancy may stem from a small difference
in referencing (the residual CHCl3 signal was set to
Proton Line-Shape Analysis of HDOPO 2a
1
The JP,H value of H2 was determined by deconvolu-
tion and listed in Table 9. All the other signals were
subjected to total line shape fitting in WINDAISY,
yielding the calculated spectrum shown in Fig. 11b
7
.24 in Ref.14 and due to a possible C-D . . . O=P in-
teraction, TMS is a more suitable reference material
for phosphorine oxides) or owing to a concentration
effect (our measurements were performed on a sat-
urated solution). More importantly, in retrospect it
is easy to recognize the possible source of the mis-
assignment of the C-H pairs at ring positions 11 to
1
and the H parameters summarized in Table 9 (for
complete results including line widths and statis-
tical descriptors, see Table S8 in the Supporting
Information). While the intra-ring nJH,H coupling
constants fall monotonically with the number of in-
tervening covalent bonds, this was found not to be
the case for couplings with the PH proton, showing a
significantly larger coupling over five bonds to H11
14. The spectral analysis in Ref.14 was started by ob-
serving the COSY correlation of the P-bonded pro-
ton to the multiplet at 7.89 ppm, which was falsely
H
e
t
e
r
o
a
t
o
m
C
h
e
m
i
s
t
r
y
D
O
I
1
0
.
1
0
0
2
/
h
c

ꢀ
1
70 Szakacs et al.
1
TABLE 9 Iterated H NMR Parameters from the 800 MHz
Spectrum of HDOPO 2a
the analysis started out from a COSY spectrum. In
the present paper, a more reliable and robust strat-
1
13
Parametera Value Parameter Value Parameter Value
egy is proposed on the basis of H, C HMBC and
NOESY spectra, which furnished the correct sig-
nal assignments for 2a and its two ethyl deriva-
tives. In addition, computer-assisted spin system
analysis of their 500 and/or 800 MHz proton spec-
3
3
3
4
4
4
4
4
5
5
5
5
6
1
3
4
4
4
5
5
5
6
δ
δ
δ
δ
δ
δ
δ
δ
H2
8.0648
7.2889
7.4145
7.2976
7.9373
7.9849
7.7583
7.5636
7.9135
8.13
J
J
J
J
J
J
J
J
J
J
J
J
H11,H12
H12,H13
H13,H14
H2,H14
H7,H9
8.07
7.41
7.50
0.19
1.22
1.56
0.95
1.40
0.57
0.32
0.34
0.60
J
J
J
J
J
J
J
J
J
J
H2,H12
0.32
592.64
14.77
0.59
5.11
3.02
1.05
0.59
1.08
0.58
–
H7
P, H 2
H8
P, H 1 4
P, H 7
H9
tra yielded reliable values for all of homo- and het-
H10
H11
H12
H13
H14
P, H 1 1
P, H 1 3
P, H 8
5
eronuclear coupling constants, including the J
H,H
H8,H10
H11,H13
H12,H14
H2,H11
H2,H13
H7,H10
H11,H14
6
and JP,H ones. This complete set of spectral parame-
ters may aid the correct assignment of further DOPO
derivatives. Finally, the chemical non-equivalence of
methylene protons owing to “chiral induction” was
also investigated, giving insights into the side-chain
flexibility.
P, H 1 0
P, H 1 2
δ
3
J
J
J
H7,H8
P, H 9
3
3
H8,H9
7.29
7.91
–
–
H9,H10
–
aChemical shifts δ in ppm, magnitudes of the coupling constants J in
Hz.
EXPERIMENTAL
General
(
0.57 Hz) than over four bonds to H14 (0.19 Hz).
This phenomenon may root in the unique elec-
tronic distribution of 2a and can clearly be studied
separately in the model compound phenylphospho-
nous acid 9, in which we have observed the follow-
ing coupling constants of phenyl protons with PH:
Compounds 2a and 9 were purchased from commer-
cial sources, while 2b and 2c were synthesized at the
Budapest University of Technology and Economics
by optimizing previously published syntheses [3].
Experimental NMR and high-resolution MS
spectra were recorded at the Spectroscopic Re-
search Department of Gedeon Richter Plc. Budapest,
Hungary. HR-MS analyses were performed on a
Finnigan MAT 95 XP system. The ionization method
was EI, operated in positive ion mode. Electron en-
ergy was 70 eV and the source temperature was set
at 220°C. Data acquisition and analysis were accom-
plished with Xcalibur software version 2.0 (Thermo
Electron Corp., San Jos e` , Ca, USA).
5
6
4
JH,H = 0.34 Hz > JH,H = 0.15 Hz > JH,H = 0.01 Hz
(to save journal space, the pertinent spectra and
all further details are given on pages 23–27 of the
the Supporting Information). Anyway, this non-
monotonic behavior of the JH,H coupling constants
may have been the reason why the deceivingly in-
tense COSY cross peak observed between H2 and
H11 provided a false premise for the assignment in
Ref.14 regarding the position of the most strongly
coupled proton. Case studies describing similar
n
“mental traps” with a potential danger to “derail” the
NMR spectra were collected at 25°C for (nearly)
correct NMR signal assignment were recently pub-
lished by our group [29]. Finally, the magnitudes
of the JP,H values in 2a follow the trend discussed
saturated solutions of the compounds in CDCl on
either of the following Varian (now Agilent Tech-
nologies, Palo Alto, CA, USA) instruments: VNMRS
3
n
1
13
for 2b above. Observable splittings of the proton-
800 ( H: 799.7 MHz, C: 201.1 MHz) equipped with
3
1
1
13 15
13
coupled P spectrum of 2a (Fig. S18 in the Support-
ing Information) are consistent with the derived cou-
pling constants and the 31P chemical shift of HDOPO
is 14.66 ppm.
a H{ C/ N} Triple Resonance C Enhanced Salt
1
Tolerant Cold Probe, VNMRS 500 ( H: 499.9 MHz,
C: 125.7 MHz) equipped with a H{ C/ N} 5 mm
PFG Triple Resonance C Enhanced Cold Probe
and a VNMRS 400 ( P: 161.8 MHz) equipped with
a OneNMR PFG Probe, using standard pulse se-
quences available in VnmrJ 3.2A software. H and
C chemical shifts were referenced to internal TMS
at 0.00 ppm, while the P referencing was accom-
13
1
13 15
1
3
3
1
CONCLUSIONS
1
The unambiguous H and 13C signal assignment of
oxaphosphaphenanthrene compounds can become
problematic, since our results here also illustrate
1
13
3
1
plished on the basis of ꢁ values according to the
n
1
that the magnitudes of homonuclear JPH,CH and het-
latest IUPAC guidelines [30]. H NMR spectra were
n
eronuclear JP,C coupling constants do not vary in a
measured with a relaxation delay of 50 s, a 90° read-
ing pulse, by recording 64K data points in one tran-
sient (scan) over a spectral width of 16 ppm, pro-
viding a digital resolution (DR) of 0.05 Hz/point or
monotonic manner with the number of intervening
covalent bonds n. These facts may easily lead to mis-
assignments, as for example occurred in the case
of four C-H pairs of HDOPO 2a in Ref.14, where
1
3
better. Proton-decoupled C spectra were run with
H
e
t
e
r
o
a
t
o
m
C
h
e
m
i
s
t
r
y
D
O
I
1
0
.
1
0
0
2
/
h
c
.

Multinuclear NMR Studies and Spin System Analyses of DOPO Compounds 171
a relaxation delay of 2 s, a 90° reading pulse and DR
mixture was extracted with 2 × 5 mL water, and
the organic layer was dried (Na2SO4). The solvent
was removed under vacuum, and the crude product
so obtained was purified by flash chromatography
(silica gel, gradient elution from dichloromethane
to dichloromethane–methanol 97:3) to furnish the
product as a colorless oil (0.62 g). The crude product
was purified by repeated column chromatography to
give 0.39 g (65%) of pure 2b. P NMR (161.8 MHz,
CDCl3), δ 16.4. HRMS: 287.10633 (C16H18O2NP;
calc. 287.10697); m/z (rel. int. %): 287(13), 272(71),
215(100), 168(45), 139(13), 72(27).
3
1
0.2 Hz/point (number of transients, nt > 256).
P
spectra were collected both with and without broad-
band (WALTZ) proton decoupling, with a relaxation
delay of 10 s and a 60° reading pulse (nt 128). For al-
most all 2D spectra, spectral windows were confined
to the aromatic region to enhance DR.
Parameters of the acquired 2D gradient-
enhanced spectra: absolute-value COSY: nt 1, num-
ber of increments (ni) 200, DR 1.9 Hz/pt (F1) and
3
1
7.6 Hz/pt (F2); double quantum filtered gradient-
enhanced COSY and the zCOSY (ECOSY type exper-
iment): nt 2, ni 128, DR 1.5 Hz/pt (F1) and 0.4 Hz/pt
1
13
(
F2); H, C HSQC with adiabatic 180° pulse: 32
2
-Ethoxy-Dibenzo[c.e] [1, 2]Oxaphosphorine-2-
Oxide 2c. Approximately 2.3 mmol of the crude
-chloro-dibenzo[c.e] [1, 2]oxaphosphorine [3] was
dissolved in 10 mL of dry toluene and 0.322 mL
dummy scans, nt 2, ni 200, DR 1 Hz/pt (F1) and
1
13
2
–9 Hz/pt (F2); H, C HMBC: band-selective
n
2
wurst2i pulse (3.2 ms, 42 db), JCH 8.5 Hz, nt 8, ni
28, DR 12 Hz/pt (F1) and 0.3 Hz/pt (F2); H, N
HMBC: JNH 8 Hz, nt 128, ni 180, DR 30 Hz/pt (F1)
and 4 Hz/pt (F2); 1D selective NOESY with zero-
quantum suppression: with a selective q3 pulse of
1
15
1
n
(2.31 mmol) of triethylamine and 0.135 mL
(2.31 mmol) of ethanol was added at 26°C on in-
tensive stirring. The mixture was stirred further for
h in nitrogen atmosphere. The contents of the flask
4
27 Hz bandwidth and a mixing time of 0.8 s. Af-
were filtered in nitrogen atmosphere, and the sol-
vent was removed under vacuum to give the crude
product.
ter converting the 1D spectra to Bruker format, line
shape iterations were performed using computer
programs available at D u¨ sseldorf [31].
The crude 2-ethoxy-dibenzo[c.e] [1, 2]oxaphos-
phorine (ꢀ2.3 mmol) was dissolved in 15 mL of
dry chloroform, and 3 mL of 30% H2O2 (14 eq.,
NMR spectra of C6H5PO2H2 9 (see supplement)
were measured at 500 MHz by Dr. Michael Acker-
mann at BRUKER BioSpin, Rheinstetten, Germany.
29.75 mmol) was added at 0°C on intensive stir-
ring. Then the mixture was allowed to warm up
to 26°C, and stirred for 1 h. The mixture was ex-
tracted with 2 × 5 mL water, and the organic layer
was dried (Na2SO4). The solvent was removed un-
der vacuum and the crude product so obtained was
purified by flash chromatography (silica gel, gradi-
ent elution from hexane–ethyl acetate 7:3 to ethyl
acetate) to furnish 0.26 g (47%) of the product as
a yellowish oil. The crude product was not crystal-
2
-Chloro-Dibenzo[c.e] [1, 2]Oxaphosphorine.
0
.50 g (2.31 mmol) of 2H-dibenzo [c.e] [1,2]oxaphos-
phorine 2-oxide 2a [2, 7] was dissolved in 20 mL of
dry (distilled from P2O5) chloroform, and 0.3 mL
1.5 eq., 3.47 mmol) of phosphorus trichloride
was added in nitrogen atmosphere. The mixture
was stirred at the boiling point for 4 h. The solvent
was removed under vacuum to give the crude prod-
uct quantitatively as a yellowish oil.
(
3
1
lized. P NMR (161.8 MHz, CDCl3), δ 10.1. HRMS:
60.05946 (C14H13O3P; calc. 260.05968); m/z (rel.
int. %): 260(44), 232(100), 216(20), 186(8), 168(29),
39(23).
2
2-N,N-Diethylamino-Dibenzo[c.e] [1, 2]Oxaphos-
phorine 2-Oxide 2b. Approximately 2.3 mmol of
the crude 2-chloro-dibenzo[c.e] [1, 2]oxaphospho-
rine [3] was dissolved in 10 mL of dry toluene and
1
0.5 mL (2 eq., 4.62 mmol) of diethylamine was added
at 26°C on intensive stirring. The mixture was stirred
further for 4 h in nitrogen atmosphere. The contents
of the flask were filtered in nitrogen atmosphere, and
the solvent was removed under vacuum to give the
crude product as a yellowish oil.
Acknowledgments
The authors are indebted to Dr. Istv a´ n Greiner, the
research director of Gedeon Richter Plc. for sup-
porting this study, Dr. Csaba Sz a´ ntay Jr. for in-
spiring discussions during the preparation of the
manuscript and Erika Pallag for her technical as-
sistance in the HR-MS measurements. Michael Ack-
ermann from Bruker Biospin is kindly acknowl-
edged for the measurements of phenylphosphonous
acid.
The crude 2-diethlyamino-dibenzo [c.e] [1, 2]
oxaphosphorine (ꢀ2.3 mmol) was dissolved in
15 mL of dry chloroform, and 3 mL of 30% H2O2
(14 eq., 29.75 mmol) was added at 0°C on inten-
sive stirring. Then the mixture was allowed to warm
up to 26°C and stirred for an additional hour. The
He te ro at om Ch em i s try DOI 10.1002/hc

ꢀ
1
72 Szakacs et al.
REFERENCES
1982, 19, 98–101; (c)Nevalainen, T.; Kohlemainen,
E.; Vil e´ n, E. Magn Res Chem 1995, 33, 355–366.
[18] McFarlane, W.; Wrackmeyer, B. J Chem Soc, Dalton
Trans 1976, 2351–2355.
[19] (a) Tarpley, A. R. Jr.; Goldstein, J. H. J Phys Chem
1971, 75, 421–430; (b) Yanagisawa, M.; Hayamizu,
K.; Yamamoto, O. Magn Res Chem 1987, 25, 184–
186.
[
[
[
[
[
1] Pastor, S. D.; Spivack, J. D.; Steinhuebel, L. P. Phos-
phorus Sulfur Silicon Relat Elem 1987, 31, 71–76.
2] Wang, C.-S.; Shieh, J.-Y. Polymer 1998, 39, 5819–
5
826.
´
3] Abr a´ nyi-Balogh, P.; Keglevich, G. Synth Comm 2011,
4
1, 1421–1426.
4] Keglevich, G.; Ker e´ nyi, A. Trends Org Chem 2008, 12,
3–77.
[20] Gryff-Keller, A.; Szczecinski, P. Org Magn Reson
1978, 11, 258–261.
7
5] Cherynshev, E. A.; Aksenov, V. I.; Ponomarev, V. V.;
Golubtsov, S. A.; Bugerenko, E. F. J Gen Chem USSR
[21] Gryff-Keller, A.; Szczecinski, P.; Koziel, H.
Organomet Chem 1989, 372, 225–230.
J
1
972, 42, 88–90.
[22] Turnbull, M. M.; Nelson, D. J.; Lekouses, W.; Sarnov,
M. L.; Tartarini, K. A.; Huang, T.-K. Tetrahedron
1990, 46, 6613–6622.
[23] Karlsen, H.; Kolsaker, P.; Romming, C.; Uggerud, E.
J Chem Soc, Perkin Trans, 2002, 2, 404–409.
[24] D u¨ rst, T.; Gryff-Keller, A.; Terpinski, J. Org Magn
Reson 1983, 21, 657–661.
[25] Kawski, P.; Kochel, A.; Perezkovina, M. G.; Fi-
larowski, A. J Mol Struct 2006, 790, 65–73.
[26] de Bod, H.; Williams, D. B. G.; Roodt, A.; Muller, A.
Acta Cryst 2004, E60, o1241–o1243.
[27] (a) Gruß, U. PhD Thesis, 1997, Heinrich Heine Uni-
versit a¨ t D u¨ sseldorf. (b) H a¨ gele, G. Aminophosphonic
and Aminophosphinic acids; John Wiley and Sons:
Chichester, 2000; Ch. 8.
[28] Prior, U.; PhD Thesis, 1992, Heinrich Heine Univer-
sit a¨ t D u¨ sseldorf.
[29] Sz a´ ntay, Cs. Jr. Anthropic Awareness, The Human
Aspects of Scientific Thinking in NMR Spectroscopy
and Mass Spectrometry; Elsevier: Amsterdam, 2015;
Ch. 9, pp. 328–352.
[30] Harris, R.; Becker, E. D.; De Menezes, S. M. C.; Good-
fellow, R.; Granger, P. Pure Appl Chem 2001, 73,
1795–1818.
[31] Programs from BRUKER BIOSPIN, Rheinstetten,
Germany. Spectral evaluation with WINNMR and
TOPSPIN, and Spectral Iteration and Simulation
with WINDAISY. Basic principles of WINDAISY
were developed in the D u¨ sseldorf group and contin-
ued in cooperation with BRUKER; e. g.: (a) Weber,
U.; Spiske, R., H o¨ ffken, H.-W., H a¨ gele, G., Thiele, H.
Manual and Programsystem; Bruker Manual 1993.
(b) Zinin, V. N.; Il’yasov, A. V., Weber, U., H a¨ gele,
G., Thiele, H. J Fluorine Chem 1995, 70, 289–292. (c)
Weber, U.; Thiele, H. NMR Spectroscopy: Modern
Spectral Analysis, Wiley-VCH: Weinheim, 1998.
[
6] Keglevich, G.; Szelke, H.; Ker e´ nyi, A.; Imre, T.;
Lud a´ nyi, K.; Dukai, J.; Nagy, F.; Ar a´ nyi, P. Het-
eroatom Chem 2004, 15, 459–463.
[
7] Ding, J.; Tao, Z.; Zuo, X.; Fan, L.; Yang, S. Polym Bull
2
009, 62, 829–841.
[
8] (a) Artner, J.; Ciesielski, M.; Ahlmann, M.; Walter,
O.; D o¨ ring, M.; Perez, R. M.; Altst a¨ dt, V.; Sandler, J.
K. W.; Schartel, B. Phosphorus Sulfur Silicon Relat
Elem 2007, 182, 2131–2148 (b) K o¨ nig, A.; Kroke, E.
Polym Adv Technol 2011, 22, 5–13.
[
9] Keglevich, G.; Szelke, H.; Ker e´ nyi, A.; Kudar, V.;
Hanusz, M.; Simon, K.; Imre, T.; Lud a´ nyi, K. Tetra-
hedron Asymm 2005, 16, 4015–4021. Cambridge
data file code: szp8ox. 6-Diethylamino-dibenzo[c.e]
[
5,6]oxaphosphorine-6-oxide.
[
[
[
[
10] Gloede, J.; Pieper, U.; Costisella, B.; Kr u¨ ger, R. P. Z
Anorg Allg Chem 2003, 629, 998–1000.
11] Keglevich, G.; Ker e´ nyi, A.; Mayer, B.; K o¨ rtv e´ lyesi, T.;
Lud a´ nyi, K. Trans Met Chem 2008, 33, 505–510.
12] Prakasha, T. K.; Day, R. O.; Holmes, R. H. J Am Chem
Soc 1994, 116, 8005–8104.
13] Gholivand, K.; Ghadimi, S.; Naderimanesh, H.;
Forouzanfar, A. Magn Res Chem 2001, 39, 684–
6
88.
[
[
14] Wagner, S.; Rakotomalala, M.; Chesneau, F.; Zevaco,
T.; D o¨ ring, M. Phosphorus Sulfur Silicon Relat Elem
2
012, 187, 781–798.
15] Berger, S.; Braun, S. 200 and More NMR Exper-
iments. A Practical Guide; Wiley-VCH: Weinheim,
2
004.
[
[
16] Stott, K.; Keeler, J.; Van, Q. N.; Shaka, A. J. J Magn
Res 1997, 125, 302–324.
17] (a)Nord e´ n, B.; Edlund, U.; Nilsson, C.-A. Magn
Res Chem 1981, 13, 391–395; (b)Buchanan, G. W.;
Wightman, R. H.; Malaiyandi, M. Magn Res Chem
H e te ro a to m C h e m is try DOI 10.1002/hc.