A proposed standard sample for nuclear overhauser effect measurements

Article abstract of DOI:10.1002/(SICI)1097-458X(199806)36:6<403::AID-OMR285>3.0.CO;2-A

A standard sample for the measurement of proton-proton nuclear Overhauser effects (NOE) is proposed. 1,5-Dichloro-2,4-dimethoxybenzene (DCDMB) shows an essentially full (50%) enhancement of the proton H-3 when the protons of the flanking methoxyl groups a

Full text of DOI:10.1002/(SICI)1097-458X(199806)36:6<403::AID-OMR285>3.0.CO;2-A

MAGNETIC RESONANCE IN CHEMISTRY, VOL. 36, 403È406 (1998)
A Proposed Standard Sample for Nuclear Overhauser Eþect
Measurements
Alex D. Bain,1* Eugene P. Mazzola2* and Samuel W. Page2
1
2
Department of Chemistry, McMaster University, 1280 Main St. W., Hamilton, Ontario, L8S 4M1, Canada
Joint Institute for Food Safety and Applied Nutrition, US Food and Drug Administration, 200 C St., S.W., Washington, DC 20204, USA
Received 14 July 1997; revised 29 December 1997; accepted 29 December 1997
ABSTRACT: A standard sample for the measurement of protonÈproton nuclear Overhauser e†ects (NOE) is pro-
posed. 1,5-Dichloro-2,4-dimethoxybenzene (DCDMB) shows an essentially full (50%) enhancement of the proton
H-3 when the protons of the Ñanking methoxyl groups are saturated. The molecule is appropriate as a standard for
the following reasons. The molecule is stable and easily synthesized. All the lines in the spectrum appear as singlets,
so selective population transfer e†ects cannot obscure the NOE. The proton at position 6 provides a monitor of
other relaxation mechanisms that would reduce the NOE. The NOE was veriÐed by two complementary experi-
ments. One is the standard technique of measuring the intensity with and without irradiation. A second method,
based on spinÈlattice relaxation rate experiments, provides unbiased data and gives an statistical estimate of the
errors. Within experimental error, the NOE for DCDMB is 50%. ( 1998 John Wiley & Sons, Ltd.
KEYWORDS: nuclear Overhauser measurements; standard sample; 1,5-dichloro-2,4-dimethoxybenzene
INTRODUCTION
the distance between spins from T measurements.10 A
1
reliable NOE reference material could also serve as a
The nuclear Overhauser e†ect (NOE) is an essential
tool in NMR.1h4 The Overhauser e†ect is caused by the
coupled relaxation of two di†erent spins, i.e. when the
relaxation rate of one spin depends on the state of the
other. One of the consequences is that saturation of one
of the spins changes the intensity of the signal of the
other. The most familiar case occurs when the coupled
relaxation is caused by dipoleÈdipole interaction. Since
the rate of dipolar relaxation is proportional to the
inverse sixth power of distance, a combination of relax-
ation and NOE measurements can provide a direct
measure of the distance between these two spins. This
information can be crucial in the structural determi-
nation of small molecules where the stereochemistry
must be established.5,6 In peptide and protein investiga-
tions, one of the essential experiments is a multi-
dimensional extension of the basic Overhauser e†ect.7,8
Much NOE work is qualitative or semi-quantitative.
check that a spectrometer is functioning, and being
operated, properly.
A candidate for a quantitative NOE standard is pro-
posed in this paper. The material, 1,5-dichloro-2,4-dime-
thoxybenzene (DCDMB), shown in Fig. 1, gives an
essentially full NOE to the proton at position 3 when
the methoxyl protons are saturated. To establish the
value of the NOE, two types of experiment were per-
formed. One was the traditional, steady-state NOE
experiment. The other, a spinÈlattice relaxation experi-
ment, measures the Overhauser e†ect in a more
unbiased fashion and gives realistic error bounds on the
observed value.
BASIC THEORY
The Overhauser e†ect can be derived from the following
basic equation governing coupled spinÈlattice relax-
ation of two spins, I and S:11
For instance, one proton of a CH group may exhibit
2
an NOE to a neighbouring spin whereas the other does
L
A
I [ I B AR (I) R (IS)BAI [ I B
not. This observation can permit di†erentiation of
pro-R from pro-S protons.9 In protein studies, NOESY
cross peaks are classiÐed as weak, medium or strong,
but further quantitation is difficult. However, there are
many situations when an exact value of the Overhauser
e†ect would be very useful. If the correlation time is
known, the value of the NOE can be used to estimate
=
=
\ [
1
1
1
1
=
=
(1)
Lt S [ S
R (IS) R (S) S [ S
where R (I) is the spinÈlattice relaxation rate of spin I
1
and R (S) is that of S. R (IS) is the coupling term.
1
1
*
Correspondence to: A. D. Bain, Department of Chemistry, McMas-
ter University, 1280 Main St. W., Hamilton, Ontario L8S 4M1,
Canada, E-mail address: bain=mcmaster.ca; E. P. Mazzola, Joint
Institute for Food Safety and Applied Nutrition, US Food and Drug
Administration, 200 C St., S.W., Washington, DC 20204, USA
Contract/grant sponsor: National Science and Engineering Research
Council of Canada (NSERC).
Figure 1. Structure of 1,5-dichloro-2,4-dimethoxyben-
zene (DCDMB).
(
1998 John Wiley & Sons, Ltd.
CCC 0749-1581/98/060403È04 $17.50

4
04
A. D. BAIN, E. P. MAZZOLA AND S. W. PAGE
Strictly, the R values are the initial-slope relaxation
rates following selective inversion, since the full relax-
chemical shift of 7.337 ppm. The sample is appropriate
for NOE measurements for several reasons. First, all of
the signals in the system are singlets, so selective popu-
lation transfer e†ects are not present.12 Second, the
1
ation is bi-exponential. The relaxation matrix in Eqn (1)
includes dipolar interaction plus other relaxation
mechanisms such as random Ðelds and chemical shield-
ing anisotropy. For small molecules in the extreme
narrowing limit, the elements of the relaxation matrix
are given by
relaxation of H should be dominated by dipolar inter-
A
actions with the methoxyl protons. Third, any non-
dipolar relaxation mechanism for H should also a†ect
A
H . Such relaxation can be monitored by observing the
X
relaxation of H . In the sample discussed here, the
relaxation of H is very slow (T [ 200 s), so additional
mechanisms a†ecting H can be largely ignored. On the
R (I) \ Rother(I) ] Rdd
X
1
1
1
X
1
R (IS) \ Rdd/2
(2)
1
1
A
basis of the above three criteria, saturation of the meth-
R (S) \ Rother(S) ] Rdd
1
1
1
oxyl protons in DCDMB was expected to produce a
In the Overhauser experiment, spin S is irradiated, so
that S in Eqn (1) is changed from its equilibrium value.
The steady-state value of I can then be calculated by
setting the derivative to zero. If the matrix elements of
Eqn (1) are given by Eqn (2), then the steady-state value
of I is given by
substantial NOE at H . The observed enhancement
A
was nearly 50%.
DCDMB is not the only spin system to give an essen-
tially full NOE. The half-cage acetate, described by
Anet and Bourn5 in pioneering work, exhibits a ca. 43%
NOE (degassed) for two highly proximate methine
protons.13 Isolated CH groups in which the two
I [ I
S [ S
Rdd/2
Rother(I) ] Rdd
2
=
\ =
1
(3)
protons are diastereotopic also show appreciable
I
I
enhancements. However, the latter protons are also
scalar coupled, which requires that the second-
irradiating Ðeld must saturate the entire multiplet, yet
not a†ect nearby transitions. If all the members of a
multiplet in a coupled system are not equally saturated,
then polarization can be transferred from one spin to
another, independent of the Overhauser e†ect.12
DCDMB provides no such problems.
=
=
1
1
If the irradiated spin is fully saturated, then S \ 0 and
Eqn (3) reduces to the usual equation for the NOE. For
a homonuclear two-spin system with pure dipolar relax-
ation, I \ S and Rother(I) \ 0, and the enhancement
=
=
1
of I is 50%. Other cases can be calculated in a similar
way.
THE SAMPLE
NOE MEASUREMENTS
DCDMB has a proton NMR spectrum consisting of
three apparent singlets (Fig. 2). The methoxyl protons
Normally, NOEs are determined by comparing a refer-
ence spectrum with one in which a spin is irradiated.
Careful integration14 is needed. This yields good data
quickly, but there are no cross-checks in this method
and no really good method of estimating errors or
biases. For the purpose of validation of a standard
sample, a modiÐed spinÈlattice relaxation time experi-
ment, shown schematically in Fig. 3, is proposed as a
check of the steady-state NOE procedure.
(
intensity 6) are at d \ 3.891, the proton (H ) at posi-
A
tion 3, between the two methoxyl groups, appears at
d \ 6.516 and the proton (H ) at position 6 has a
X
In the modiÐed T -NOE experiment, the methoxyl
1
protons are irradiated until a steady state is achieved.
The system is then allowed to relax for some time, q,
and the z-magnetization is sampled with a n/2 pulse.
This experiment is repeated for a series of q values. The
irradiation power must be sufficiently large so that the
Figure 3. Pulse sequence for the combined T and NOE
1
experiment. The methoxyl protons are irradiated until
the system reaches the steady state. Then the decoupler
is shut oþ, relaxation occurs during time q and the
z-magnetizations for all the spins are measured using a
non-selective n/2 pulse.
Figure 2. 300 MHz proton NMR spectrum of DCDMB.
The peak at 7.23 ppm is residual chloroform in the
solvent.
(
1998 John Wiley & Sons, Ltd.
MAGNETIC RESONANCE IN CHEMISTRY, VOL. 36, 403È406 (1998)

PROPOSED STANDARD SAMPLE FOR NOE MEASUREMENTS
405
methoxyl protons are fully saturated but not so strong
(Aldrich) (0.23 g) were placed in a 250 ml round-
bottomed Ñask containing 50 ml of water and 50 ml of
methylene chloride, and the mixture was stirred over-
night at room temperature. The aqueous layer was
extracted with 50 ml of methylene chloride and the
extract added to the methylene chloride layer. The com-
bined methylene chloride solutions were successively
washed with 3 M ammonia solution, 10% NaOH and
salt water. The resulting methylene chloride solution
was dried over anhydrous Na SO and concentrated in
as to perturb H . The residual water signal is useful for
A
monitoring this power level. Since the residual water
and DCDMB are both dilute, there is no real chance of
an intermolecular Overhauser e†ect. The chemical shift
di†erence between the water and the methoxyl reso-
nance is smaller than that between H and the meth-
A
oxyl signal, so o†-resonance irradiation will a†ect the
water more than H . Provided that the water signal (or
A
some other reference line in the same chemical shift
2
4
range) does not vary as a function of q, o†-resonance
e†ects can be assumed to be unimportant. None of the
other parameters in this experiment is critical. The irra-
a rotary evaporator. The resulting product was rec-
rystallized from hexane and a†orded 1.9 g of white
needles.
diation time should be longer than 5T but is otherwise
not important.15 The observation pulse can be of any
Test samples were prepared by Wilmad Glass (Buena,
NJ, USA) for evaluation purposes in the following
manner. Approximately 1 mg of DCDMB was dis-
1
value since there are no Ñip angle e†ects on singlets.16
This experiment illustrates the principle that the
initial conditions in a relaxation experiment are useful
parameters.17 The quantities that govern the course of
relaxation are the initial state, the equilibrium state and
the rate matrix. The NOE is the ratio of the initial
solved in 0.6 ml of CDCl . The sample was placed in an
3
NMR tube (506-PP), subjected to six freezeÈpumpÈ
thaw cycles and then sealed under vacuum. Proton
NMR spectra were obtained (i) at 300 MHz on a
Bruker AC300 spectrometer equipped with a 5 mm,
four-nucleus probe and (ii) at 400 MHz on a Varian
VXR-400S spectrometer equipped with a 5 mm, 1H/13C
switchable probe. The n/2 pulse for protons was 9 ls on
the AC300 and 20 ls on the VXR-400S. The probe tem-
peratures on the AC300 were set to 300 K and con-
trolled to within ^0.2 K using a Bruker BVT2000
variable-temperature regulator; those on the VXR-400S
were set to 298 K and controlled to within ^0.1 K.
(
irradiated) state to the equilibrium state in this experi-
ment, and the rate matrix provides an important cross-
check.
There are many relaxation experiments that could
also be employed, developed for studying chemical
exchange and cross-relaxation.18 The transient NOE of
H is observed after the methoxyl protons are inverted.
A
For a two-spin system, this transient NOE is large if the
steady-state NOE is likewise substantial. However,
when a single spin is relaxed by n equivalent spins
(
n \ 6 in this case), the steady-state NOE is the
same, but the transient NOE is reduced ca. 1/n. For
DCDMB, this transient NOE would be too small to be
measured accurately.
Steady-state NOE experiments at 400 MHz
Preliminary experiments were conducted to determine
the minimum power level at which the six methyl
protons were saturated. The spinÈlattice relaxation time
The formal solution to the di†erential Eqn (1) is
A
I(t) [ I(O)B AI(0) [ I(O)B
\
exp([Rt)
(4)
of H was also measured using the inversionÈrecovery
S(t) [ S(O)
S(0) [ S(O)
A
method20,21 and found to be 4.5 s. Presaturation times
where R is the relaxation matrix in Eqn (1) and “expÏ
means the exponential of a matrix. The program
CIFIT18 was modiÐed to Ðt the present experiment.
Not only the relaxation matrix elements but also the
intensities at q \ 0 and those at q \ O are used as
Ðtting parameters. These intensities provide a “best ÐtÏ
value of the Overhauser enhancement along with realis-
tic error estimates,19 and the relaxation matrix elements
validate the data.
of 30 s ([5T ) were employed to ensure that enhance-
1
ments of H were fully developed. Sixty-four transients
A
each were collected on-resonance (3.891 ppm, OCH )
and o†-resonance (9.15 ppm). The second-irradiating
3
frequency was gated o† during both the n/2 pulse and
signal acquisition. The FIDs were subjected to a 1 Hz
line broadening before Fourier transformation and
spectral baselines were corrected prior to integration of
the H signals.
A
EXPERIMENTAL
Modiüed T -NOE experiments at 300 MHz
1
1
,5-Dichloro-2,4-dimethoxybenzene
SpinÈlattice relaxation rates were estimated using the
saturationÈrecovery method.22 For NOE measure-
ments, the spectrometer “decoupler powerÏ parameter in
BrukerÏs DISNMR program was set to 20L, which cor-
responds to a precession frequency around the decou-
pler Ðeld of 100 Hz. The FIDs were subjected to a 5 Hz
(
Fig. 1, DCDMB)
Dimethyl sulfate (Aldrich) (7.6 g), 4,6-dichlororesorcinol
Aldrich) (1.8 g), NaOH (Fisher) (1.6 g) and the phase-
transfer catalyst benzyltriethylammonium chloride
(
(
1998 John Wiley & Sons, Ltd.
MAGNETIC RESONANCE IN CHEMISTRY, VOL. 36, 403È406 (1998)

4
06
A. D. BAIN, E. P. MAZZOLA AND S. W. PAGE
line broadening prior to Fourier transformation. The
baselines of the resulting spectra were corrected and the
height of each peak above the corrected baseline was
recorded.
vides a valuable reference sample to test and calibrate
the nuclear Overhauser e†ect.
Acknowledgements
RESULTS
The authors thank Paul Cope of Wilmad for his help and A.D.B.
thanks the Natural Sciences and Engineering Research Council of
Canada (NSERC) for Ðnancial support.
Steady-state NOE experiments
Twenty-Ðve experiments were carried out with presatu-
ration times of 45 s, Ðve with 30 s and Ðve with 60 s and
REFERENCES
the observed signal enhancements of H were found to
be identical within the limits of experimental error
A
1. R. A. Bell and J. K. Saunders, Can. J. Chem. 48, 1114 (1970).
2
3
4
. K. F. Kuhlmann, D. M. Grant and R. K. Harris, J. Chem. Phys.
52, 3439 (1970).
(
^1%). The average H MOCH N NOE was 49.1% and
A
3
. J. H. Noggle and R. E. Schirmer, T he Nuclear Overhauser E†ect:
Chemical Applications. Academic Press, New York (1971).
. D. Neuhaus and M. P. Williamson, T he Nuclear Overhauser
E†ect in Structural and Conformational Analysis. Verlag Chemie,
New York (1989).
the median was 49.0%.
Modiüed T -NOE experiments
1
5. F. A. L. Anet and A. J. R. Bourn, J. Am. Chem. Soc. 87, 5250
(
1965).
Because the relaxation of H was very slow, the
saturationÈrecovery method was used to avoid lengthy
equilibration times after FID acquisition. These experi-
6. J. Stonehouse, P. Adell, J. Keeler and A. J. Shaka, J. Am. Chem.
X
Soc. 116, 6037 (1994).
7
. A. Kumar, R. R. Ernst and K. Wu
Commun. 95, 1 (1980).
Ž thrich, Biochem. Biophys. Res.
ments demonstrated that H had a 1/T of ca. 0.004
8. D. Wemmer and N. R. Kallenbach, Biochemistry 22, 1901 (1983).
X
1
9
. D. Neuhaus and M. P. Williamson, T he Nuclear Overhauser
E†ect in Structural and Conformational Analysis, pp. 426È428.
Verlag Chemie, New York (1989).
s~1 (T B 250 s). The NOE experiment yielded the fol-
1
lowing values: for H , I \ 1.54 ^ 0.02, I
A
=
steadyvstate
[
I \ 0.71 ^ 0.02, and for the methoxyl protons,
10. C. W. M. Grant, L. D. Hall and C. M. Preston, J. Am. Chem. Soc.
=
9
5, 7742 (1973).
S \ 10.03 ^ 0.02,
S
[ S \ [9.25 ^ 0.02.
=
steadyvstate
=
11. I. Solomon, Phys. Rev. 99, 559 (1955).
This gives an NOE of 0.49 with an error estimated, as
previously described, of ca. ^0.05 (95% conÐdence
limit, estimated by the proÐling procedure). Repetition
of the experiment produced NOE values in the range
1
2. K. G. R. Pachler and P. L. Wessels, J. Magn. Reson. 12, 337
(1973).
1
1
3. F. A. L. Anet, personal communication.
4. F. A. L. Anet and M. Kopelevich, J. Am. Chem. Soc. 108, 1355
(
1986).
0
.45È0.51.
15. D. Neuhaus and M. P. Williamson, T he Nuclear Overhauser
E†ect in Structural and Conformational Analysis, p. 119. Verlag
Chemie, New York (1989).
These results suggest that large NOEs cannot be
measured as precisely by the modiÐed T method as
with the steady-state experiment. However, the tech-
1
6. S. Scha
(1974).
Ž ublin, A. H oŽ hener and R. R. Ernst, J. Magn. Reson. 13, 196
1
17. A. D. Bain and J. A. Cramer, J. Magn. Reson. A 103, 217 (1993).
18. A. D. Bain and J. A. Cramer, J. Magn. Reson. A 118, 21 (1996).
19. G. A. F. Seber and C. J. Wild, Nonlinear Regression. Wiley, New
York (1989).
nique represents a valuable cross-check, because it is
much less susceptible to systematic errors and biases.
More importantly, two complementary experimental
techniques have been shown to yield essentially the
same full enhancement for proton A upon irradiation of
the methoxyl protons in DCDMB. This molecule pro-
2
0. R. L. Vold, J. S. Waugh, M. P. Klein and D. E. Phelps, J. Chem.
Phys. 48, 3831 (1968).
2
1. R. Freeman and H. D. W. Hill, J. Chem. Phys. 51, 3140 (1969).
22. G. C. McDonald and J. S. Leigh, J. Magn. Reson. 9, 358 (1973).
(
1998 John Wiley & Sons, Ltd.
MAGNETIC RESONANCE IN CHEMISTRY, VOL. 36, 403È406 (1998)