- Email: [email protected]
A positive “through-space” fluorine-fluorine coupling constant in perfluorotoluene
JOURNAL
OF MAGNETIC
RESONANCE
18,241-247
(1975)
A Positive “Through-Space” Fluorine-Fluorine Coupling Constant in Perfluorotoluene C. W. HAICH Chemistry
Department,
University
College,
Swansea
SA2 8PP,
Wales
J. HILTON AND L. H. SUTCLIFFE Donnan
Chemistry
Laboratories,
The University,
Liverpool
L69 3BX,
England
AND
G. J. T. TIDDY U&ever
Research
Laboratory,
Port
Sunlight,
Cheshire
L62 4XN,
England
Received October 2,1974 All the fluorine chemical shifts and the signs and magnitudes of the fluorinefluorine coupling constants of perfluorotoluene have been determined. The most important finding is that the large four-bond fluorine-fluorine coupling constant between an ortho-fluorine nucleus and the perfluoromethyl nuclei is positive. This means, that despite theoretical predictions to the contrary, all known fluorinefluorine coupling constants containing a large “through-space” contribution are positive. INTRODUCTION
Throughout the history of high-resolution nuclear magnetic resonance spectroscopy, fluorine-fluorine coupling constants have been a source of considerable interest. In particular, the large magnitudes of four- and five-bond coupling constants between spatially close fluorine nuclei has been a puzzle (1-6). Recently, the signs have been shown (7) unequivocally to be positive for four- and five-bond coupling constants in the “fixed” conformers of some perfluoroalkyl fluorinated aromatic compounds. One of these [perfluoro(2,4-diisopropylquinoline)], has a very large five-bond fluorinefluorine coupling constant of +196.4 Hz. From both the mechanistic and the diagnostic points of view it is important to establish whether the “through-space” coupling constant contribution is invariably positive; hence we searched the literature for fluorine-fluorine coupling constants of known sign that could be interpreted as containing a large “through-space” contribution. An exception was found for perfluorotoluene, which has the preferred conformation (8) :
Copyright 0 1975 by Academic Press, Inc. All rights ofreproduction in any form reserved. Printed in Great Britain
241
242
HAIGH
ET AL.
Dean and McFarlane (9) made some double resonance experiments on perfluorotoluene and, after assumjng that first-order assignment of transitions is permissible, they concluded that the rotationally averaged coupling constant (JIZ) is negative. However, Cooper (IO), from a study of the solvent effects on some F-F couplings, has suggested that this couphng constant is positive. Because of the small amount of data from such studies, Cooper recognized that this conclusion should be regarded as tentative. Since the fluorine NMR spectrum is not first order we decided to carry out a full spectral analysis in order to establish the sign of (J12) more convincingly. After our investigation had started, Hirao ef al. (II, 12) made some calculations on fluorinefluorine coupling constants and showed that the Fermi contact contribution becomes dominant for fluorine nuclei in close proximity and separated by more than four bonds. Furthermore, some of their calculated coupling constants have negative signs. EXPERIMENTAL
The sample of perfluorotoluene was kindly donated by Professor W. K. R. Musgrave of Durham University. A 75% (v/v) solution in CFCI, was prepared, degassed and sealed into a standard 5-mm o.d. NMR sample tube.
10 iicrtz -
:
iF i
FIG. 1. The fluorine-19 experimental ortho nuclei of perlluorotoluene.
and computer-simulated
NMR spectra at 94.6 MHz of the
COUPLING
243
IN PERFLUOROTOLUENE
Fluorine NMR spectra were recorded with a Varian HA-60 spectrometer operating at 56.4422 MHz, and a Schlumberger FS 3017 frequency synthesizer was used to provide the locking sideband. Fluorine spectra were also recorded at 94.6 MHz with a JEOL PS 100 spectrometer located at the JEOL (U.K.) Ltd. laboratories at Colindale. Spin-tickling experiments were kindly carried out on a Varian XL-100 spectrometer by Mr. M. Pendlebury at Imperial College. Calculations were made for the refinement of spectral parameters using the programs LACX and LAME, firstly with an ICL 1905E computer with a 64k store and then with an ICL 1904s computer having a 96k store. Both computers are at the University College of Swansea.
10 Hertz I
FIG. 2. The fluorine-19 experimental meta nuclei of perfluorotoluene.
and computer-simulated
I
NMR spectra at 94.6 MHz of the
244
HAIGH ET AL. RESULTS
The fluorine spectrum of perfluorotoluene consists of four distinct groups of bands found at -56.49 (CF, group), -140.28 (ortho-F), -161.06 (m&z-F) and -148.37 ppm (para-F) to high fields of CFCl,. The asymmetry present in the ortho (Fig. 1) and meta (Fig. 2) parts of the spectrum makes it possible to extract signs of coupling constants from the spectral analysis. In this analysis, the ortho aromatic coupling constants Jz3 and J,, were taken to be negative (13).
10 Hertz L
FIG. 3. The fluorine-19 experimental and computer-simulated para nucleus of pertluorotoluene.
I
NMR spectra at 56.4 MHz of the
A first-order analysis as an [AM12RX3 spin system showed that Jz3 and Jz5 are of opposite sign. Calculations using LAME gave Jz4 as positive and Jz6 as negative. The asymmetry of the ortho and the metu bands allowed J35 to be given a negative value. Spectra were simulated for all eight combinations of signs for (Jr&, (J13) and (J14) and compared with the experimental spectra: (Jlz) and (J& were found to be positive and (.I& is likely to be positive.
COUPLING
IN PERFLUOROTOLUENE
245
Selective decoupling experiments were carried out in which the outer bands of the CF, group of signals were irradiated in turn. Observation of the meta bands showed that the sign of (J12) is opposite to that of Jz3 +Jz5 (known to be negative from a previous analysis). Observation of the paru bands (Fig. 3) showed that (JIz) has the same sign asJ,,. Thus these experiments established that both (J12) and .Jz4are positive.
Ii I I' !I
FIG. 4. The fluorine-19 experimental and computer-simulated fluoromethyl group nuclei of perfluorotoluene.
NMR spectra at 94.6 MHz of per-
Some spin-tickling experiments were performed in which the para bands were observed while irradiating CF, bands (Fig. 4) : these proved that (.I,,) has the same sign asJ,, and that (J13) has a sign opposite to that ofJ,,, thus (.I& is positive. Observation of meta bands while irradiating the CF, bands gave more complex results and irradiations corresponding to (JIz) cannot be correlated with the coupling constant in the meta band. However, the sign of (J14) was shown to be opposite that of Js4. All the coupling constant values found by iteration and their signs are summarized in Table 1.
246
HAIGH
ET AL.
TABLE FLUORINE-FLUORINE FLUORINE CHEMICAL
I
COUPLING SHIFTS” TOLUENE
CONSTANTS AND FOR PERFLUORO-
Rms <12> = (16) <13> = (15) <14> 23 = 56 24 = 46 25 = 36 26 34 = 45 35
= +22.68 Hz = + 0.62 =+ 1.33b =-20.01 =+ 5.54 =+ 8.34 = - 1.52 =-l&97 = - 0.43
errors
0.011 Hz 0.008 0.006 0.016 0.011 0.016 0.019 0.011 0.019 Rms
CF3
or&-F m&a-F para-F
6 6 6 6
’ Measured upfield b Sign is tentative.
= = = =
- 56.49 ppm -140.28 -161.06 -148.37 from
internal
0.011 0.022 0.009 0.009
errors ppm
CFC13.
DISCUSSION
The fluorine chemical shifts of perfluorotoluene have the expected values (24-16) and the ring fluorine-fluorine coupling constants follow the additive substituent rule devised by Abraham, Macdonald, and Pepper (17). The most important finding in this work is that the sign of the rotationahy averaged coupling constant (5& is positive: this means that no negative four- or five-bond fluorine-fluorine coupling constants have yet been discovered for two fluorine nuclei in close proximity. A possible explanation for the negative sign obtained by Dean and McFarlane (9) is that spectra are reversed in certain modes of spectrometer operation and then a misinterpretation is easily made. An estimation of the value of (.I& can be made by using the results of Jonas, Borowski, and Gutowsky (29). The values they obtained for the individual coupling constants have been modified using the values obtained earlier in this laboratory (7) for perfluoroalkylfluoroaromatic compounds (J12)=[J~a,2+tJlb,2+Jlc,2+Jla,6fJib,6fJlc,61/6
= +86.0 + 7.4 + 7.4 + 0.0 + 19.8 + 19.8 Hz = +23.4 Hz. This value is in good agreement with that of +22.68 Hz obtained experimentally. The verification of the positive sign of (JIz) adds weight to the proposed use (10) of solvent effects for sign determinations of fluorine-fluorine coupling constants. Hirao et al. (II, 12) have not made any calculations on perfluorotoluene or a related molecule but, using their INDO molecular orbital and sum-over-states perturbation
COUPLING
247
IN PERFLUOROTOLUENE
method, they have calculated the four-bond fluorine-fluorine coupling constants in a range of saturated and unsaturated aliphatic compounds. An important feature of their results is that the Fermi contact contribution is much larger than the orbital and spin dipolar contributions. Furthermore, the only large value of the former contribution for a fixed conformer is a negative coupling (Table IX, Ref. 1Z). Using standard bond angles and bond lengths, we have calculated the distance separating the la and 2 fluorine nuclei to be 0.229 nm. Reference to the internuclear distance versus coupling constant diagram published by Hirao et al. (Fig. 8, Ref. 12), gives a calculated value of J,, , 2 of -33 Hz. This poor agreement coupled with criticisms we have made elsewhere (18) of the conclusions of Hirao et al. (If, 22) means that their calculations are by no means satisfactory. ACKNOWLEDGMENTS We thank Dr. K. W. Jolley and Miss S. French (U.K.) Ltd., for allowing us to use their spectrometers for a maintenance grant to one of us (J.H.).
for their assistance. We are also indebted at Colindale and to the Science Research
to JEOL Council
REFERENCES I. 2. 3. 4. 5. h. 7. 8. 9. IO. II. /2. 13. 14. 15. 16. 17. ZY.
19.
L. PETRAKIS AND C. H. SEDERHOLM, J. Chem. P&s. 35,1243 (1961). K. L. SERVIS AND J. D. ROBERTS, J. Amer. Chem. Sot. 87,1339 (1965). N. BODEN, J. FEENEY AND L. H. SUTCLIFFE, J. Chem. Sot. 3482 (1965). J. BURDON, Tetrahedron 21, 1101 (1965). F. J. WEICERT AND J. D. ROBERTS, J. Amer. Chem. Sot. 90,3577 (1968). K. L. SERVIS AND KAI-NAN FANG, J. Amer. Chem. Sot. 90,6712 (1968). R. D. CHAMBERS, L. H. SUTCLIFFE, AND G. J. T. TIDDY, Truns. Faraday Ser. 66, 1025 (1970). R. WASYLISHEN AND T. SCHAEFER, Can. J. Chem. 50,1852 (1972). R. R. DEAN AND W. MCFARLANE, J. Chem. Sot. 509 (1969). M. A. COOPER, Org. Magn. Resonance 1,363 (1969). K. HIRAO, H. NAKATSUJI, H. KATO, AND T. YONEZAWA, J. Amer. Chem. Sot. 94,4078 (1972). K. HIRAO, H. NAKATSUJI, AND H. KATO, J. Amer. Chem. Sot. 95,31 (1973). L. SNYDER AND E. W. ANDERSON, J. Chem. Phys. 42,3336 (1965). N. BODEN, J. W. EMSLEY, J. FEENEY, AND L. H. SUTCLIFFE, Mol. Phys. 8,133 (1964). 1. J. LAWRENSON, J. Chem. Sot., 1117 (1965). E. LUSTI~ AND P. DIEHL, J. Chem. Phys. 44,2974 (1964). R. J. ABRAHAM, D. B. MACDONALD, AND E. S. PEPPER, J. Amer. Chem. Sot. 90,147 (1968). J. HILTON AND L. H. SUTCLIFFE, J. Mugn. Resonance 14,241 (1974). J. JONES, L. BOROWSKI, AND H. S. GUTOWSKY, J. Chem. Phys. 47,244l (1967).
9
OF MAGNETIC
RESONANCE
18,241-247
(1975)
A Positive “Through-Space” Fluorine-Fluorine Coupling Constant in Perfluorotoluene C. W. HAICH Chemistry
Department,
University
College,
Swansea
SA2 8PP,
Wales
J. HILTON AND L. H. SUTCLIFFE Donnan
Chemistry
Laboratories,
The University,
Liverpool
L69 3BX,
England
AND
G. J. T. TIDDY U&ever
Research
Laboratory,
Port
Sunlight,
Cheshire
L62 4XN,
England
Received October 2,1974 All the fluorine chemical shifts and the signs and magnitudes of the fluorinefluorine coupling constants of perfluorotoluene have been determined. The most important finding is that the large four-bond fluorine-fluorine coupling constant between an ortho-fluorine nucleus and the perfluoromethyl nuclei is positive. This means, that despite theoretical predictions to the contrary, all known fluorinefluorine coupling constants containing a large “through-space” contribution are positive. INTRODUCTION
Throughout the history of high-resolution nuclear magnetic resonance spectroscopy, fluorine-fluorine coupling constants have been a source of considerable interest. In particular, the large magnitudes of four- and five-bond coupling constants between spatially close fluorine nuclei has been a puzzle (1-6). Recently, the signs have been shown (7) unequivocally to be positive for four- and five-bond coupling constants in the “fixed” conformers of some perfluoroalkyl fluorinated aromatic compounds. One of these [perfluoro(2,4-diisopropylquinoline)], has a very large five-bond fluorinefluorine coupling constant of +196.4 Hz. From both the mechanistic and the diagnostic points of view it is important to establish whether the “through-space” coupling constant contribution is invariably positive; hence we searched the literature for fluorine-fluorine coupling constants of known sign that could be interpreted as containing a large “through-space” contribution. An exception was found for perfluorotoluene, which has the preferred conformation (8) :
Copyright 0 1975 by Academic Press, Inc. All rights ofreproduction in any form reserved. Printed in Great Britain
241
242
HAIGH
ET AL.
Dean and McFarlane (9) made some double resonance experiments on perfluorotoluene and, after assumjng that first-order assignment of transitions is permissible, they concluded that the rotationally averaged coupling constant (JIZ) is negative. However, Cooper (IO), from a study of the solvent effects on some F-F couplings, has suggested that this couphng constant is positive. Because of the small amount of data from such studies, Cooper recognized that this conclusion should be regarded as tentative. Since the fluorine NMR spectrum is not first order we decided to carry out a full spectral analysis in order to establish the sign of (J12) more convincingly. After our investigation had started, Hirao ef al. (II, 12) made some calculations on fluorinefluorine coupling constants and showed that the Fermi contact contribution becomes dominant for fluorine nuclei in close proximity and separated by more than four bonds. Furthermore, some of their calculated coupling constants have negative signs. EXPERIMENTAL
The sample of perfluorotoluene was kindly donated by Professor W. K. R. Musgrave of Durham University. A 75% (v/v) solution in CFCI, was prepared, degassed and sealed into a standard 5-mm o.d. NMR sample tube.
10 iicrtz -
:
iF i
FIG. 1. The fluorine-19 experimental ortho nuclei of perlluorotoluene.
and computer-simulated
NMR spectra at 94.6 MHz of the
COUPLING
243
IN PERFLUOROTOLUENE
Fluorine NMR spectra were recorded with a Varian HA-60 spectrometer operating at 56.4422 MHz, and a Schlumberger FS 3017 frequency synthesizer was used to provide the locking sideband. Fluorine spectra were also recorded at 94.6 MHz with a JEOL PS 100 spectrometer located at the JEOL (U.K.) Ltd. laboratories at Colindale. Spin-tickling experiments were kindly carried out on a Varian XL-100 spectrometer by Mr. M. Pendlebury at Imperial College. Calculations were made for the refinement of spectral parameters using the programs LACX and LAME, firstly with an ICL 1905E computer with a 64k store and then with an ICL 1904s computer having a 96k store. Both computers are at the University College of Swansea.
10 Hertz I
FIG. 2. The fluorine-19 experimental meta nuclei of perfluorotoluene.
and computer-simulated
I
NMR spectra at 94.6 MHz of the
244
HAIGH ET AL. RESULTS
The fluorine spectrum of perfluorotoluene consists of four distinct groups of bands found at -56.49 (CF, group), -140.28 (ortho-F), -161.06 (m&z-F) and -148.37 ppm (para-F) to high fields of CFCl,. The asymmetry present in the ortho (Fig. 1) and meta (Fig. 2) parts of the spectrum makes it possible to extract signs of coupling constants from the spectral analysis. In this analysis, the ortho aromatic coupling constants Jz3 and J,, were taken to be negative (13).
10 Hertz L
FIG. 3. The fluorine-19 experimental and computer-simulated para nucleus of pertluorotoluene.
I
NMR spectra at 56.4 MHz of the
A first-order analysis as an [AM12RX3 spin system showed that Jz3 and Jz5 are of opposite sign. Calculations using LAME gave Jz4 as positive and Jz6 as negative. The asymmetry of the ortho and the metu bands allowed J35 to be given a negative value. Spectra were simulated for all eight combinations of signs for (Jr&, (J13) and (J14) and compared with the experimental spectra: (Jlz) and (J& were found to be positive and (.I& is likely to be positive.
COUPLING
IN PERFLUOROTOLUENE
245
Selective decoupling experiments were carried out in which the outer bands of the CF, group of signals were irradiated in turn. Observation of the meta bands showed that the sign of (J12) is opposite to that of Jz3 +Jz5 (known to be negative from a previous analysis). Observation of the paru bands (Fig. 3) showed that (JIz) has the same sign asJ,,. Thus these experiments established that both (J12) and .Jz4are positive.
Ii I I' !I
FIG. 4. The fluorine-19 experimental and computer-simulated fluoromethyl group nuclei of perfluorotoluene.
NMR spectra at 94.6 MHz of per-
Some spin-tickling experiments were performed in which the para bands were observed while irradiating CF, bands (Fig. 4) : these proved that (.I,,) has the same sign asJ,, and that (J13) has a sign opposite to that ofJ,,, thus (.I& is positive. Observation of meta bands while irradiating the CF, bands gave more complex results and irradiations corresponding to (JIz) cannot be correlated with the coupling constant in the meta band. However, the sign of (J14) was shown to be opposite that of Js4. All the coupling constant values found by iteration and their signs are summarized in Table 1.
246
HAIGH
ET AL.
TABLE FLUORINE-FLUORINE FLUORINE CHEMICAL
I
COUPLING SHIFTS” TOLUENE
CONSTANTS AND FOR PERFLUORO-
Rms <12> = (16) <13> = (15) <14> 23 = 56 24 = 46 25 = 36 26 34 = 45 35
= +22.68 Hz = + 0.62 =+ 1.33b =-20.01 =+ 5.54 =+ 8.34 = - 1.52 =-l&97 = - 0.43
errors
0.011 Hz 0.008 0.006 0.016 0.011 0.016 0.019 0.011 0.019 Rms
CF3
or&-F m&a-F para-F
6 6 6 6
’ Measured upfield b Sign is tentative.
= = = =
- 56.49 ppm -140.28 -161.06 -148.37 from
internal
0.011 0.022 0.009 0.009
errors ppm
CFC13.
DISCUSSION
The fluorine chemical shifts of perfluorotoluene have the expected values (24-16) and the ring fluorine-fluorine coupling constants follow the additive substituent rule devised by Abraham, Macdonald, and Pepper (17). The most important finding in this work is that the sign of the rotationahy averaged coupling constant (5& is positive: this means that no negative four- or five-bond fluorine-fluorine coupling constants have yet been discovered for two fluorine nuclei in close proximity. A possible explanation for the negative sign obtained by Dean and McFarlane (9) is that spectra are reversed in certain modes of spectrometer operation and then a misinterpretation is easily made. An estimation of the value of (.I& can be made by using the results of Jonas, Borowski, and Gutowsky (29). The values they obtained for the individual coupling constants have been modified using the values obtained earlier in this laboratory (7) for perfluoroalkylfluoroaromatic compounds (J12)=[J~a,2+tJlb,2+Jlc,2+Jla,6fJib,6fJlc,61/6
= +86.0 + 7.4 + 7.4 + 0.0 + 19.8 + 19.8 Hz = +23.4 Hz. This value is in good agreement with that of +22.68 Hz obtained experimentally. The verification of the positive sign of (JIz) adds weight to the proposed use (10) of solvent effects for sign determinations of fluorine-fluorine coupling constants. Hirao et al. (II, 12) have not made any calculations on perfluorotoluene or a related molecule but, using their INDO molecular orbital and sum-over-states perturbation
COUPLING
247
IN PERFLUOROTOLUENE
method, they have calculated the four-bond fluorine-fluorine coupling constants in a range of saturated and unsaturated aliphatic compounds. An important feature of their results is that the Fermi contact contribution is much larger than the orbital and spin dipolar contributions. Furthermore, the only large value of the former contribution for a fixed conformer is a negative coupling (Table IX, Ref. 1Z). Using standard bond angles and bond lengths, we have calculated the distance separating the la and 2 fluorine nuclei to be 0.229 nm. Reference to the internuclear distance versus coupling constant diagram published by Hirao et al. (Fig. 8, Ref. 12), gives a calculated value of J,, , 2 of -33 Hz. This poor agreement coupled with criticisms we have made elsewhere (18) of the conclusions of Hirao et al. (If, 22) means that their calculations are by no means satisfactory. ACKNOWLEDGMENTS We thank Dr. K. W. Jolley and Miss S. French (U.K.) Ltd., for allowing us to use their spectrometers for a maintenance grant to one of us (J.H.).
for their assistance. We are also indebted at Colindale and to the Science Research
to JEOL Council
REFERENCES I. 2. 3. 4. 5. h. 7. 8. 9. IO. II. /2. 13. 14. 15. 16. 17. ZY.
19.
L. PETRAKIS AND C. H. SEDERHOLM, J. Chem. P&s. 35,1243 (1961). K. L. SERVIS AND J. D. ROBERTS, J. Amer. Chem. Sot. 87,1339 (1965). N. BODEN, J. FEENEY AND L. H. SUTCLIFFE, J. Chem. Sot. 3482 (1965). J. BURDON, Tetrahedron 21, 1101 (1965). F. J. WEICERT AND J. D. ROBERTS, J. Amer. Chem. Sot. 90,3577 (1968). K. L. SERVIS AND KAI-NAN FANG, J. Amer. Chem. Sot. 90,6712 (1968). R. D. CHAMBERS, L. H. SUTCLIFFE, AND G. J. T. TIDDY, Truns. Faraday Ser. 66, 1025 (1970). R. WASYLISHEN AND T. SCHAEFER, Can. J. Chem. 50,1852 (1972). R. R. DEAN AND W. MCFARLANE, J. Chem. Sot. 509 (1969). M. A. COOPER, Org. Magn. Resonance 1,363 (1969). K. HIRAO, H. NAKATSUJI, H. KATO, AND T. YONEZAWA, J. Amer. Chem. Sot. 94,4078 (1972). K. HIRAO, H. NAKATSUJI, AND H. KATO, J. Amer. Chem. Sot. 95,31 (1973). L. SNYDER AND E. W. ANDERSON, J. Chem. Phys. 42,3336 (1965). N. BODEN, J. W. EMSLEY, J. FEENEY, AND L. H. SUTCLIFFE, Mol. Phys. 8,133 (1964). 1. J. LAWRENSON, J. Chem. Sot., 1117 (1965). E. LUSTI~ AND P. DIEHL, J. Chem. Phys. 44,2974 (1964). R. J. ABRAHAM, D. B. MACDONALD, AND E. S. PEPPER, J. Amer. Chem. Sot. 90,147 (1968). J. HILTON AND L. H. SUTCLIFFE, J. Mugn. Resonance 14,241 (1974). J. JONES, L. BOROWSKI, AND H. S. GUTOWSKY, J. Chem. Phys. 47,244l (1967).
9