Magnetic Nonequivalence in the High-

Resolution NMR Spectra of Diborane

Cite as: J. Chem. Phys. 49, 281 (1968); https://doi.org/10.1063/1.1669822
Submitted: 06 February 1968 . Published Online: 18 September 2003

T. C. Farrar, R. B. Johannesen, and T. D. Coyle

ARTICLES YOU MAY BE INTERESTED IN

Vibrational–Vibrational–Translational Energy Transfer between Two Diatomic

Molecules
The Journal of Chemical Physics 49, 286 (1968); https://doi.org/10.1063/1.1669823

11
B Nuclear Magnetic Resonance Spectrum of the ``Diammoniate of Diborane''
The Journal of Chemical Physics 32, 952 (1960); https://doi.org/10.1063/1.1730834

Nuclear Magnetic Resonance Spectra and Structure of Borohydride Ion and Diborane
The Journal of Chemical Physics 22, 1933 (1954); https://doi.org/10.1063/1.1739944

J. Chem. Phys. 49, 281 (1968); https://doi.org/10.1063/1.1669822 49, 281

© 1968 American Institute of Physics.

 THE JOURNAL OF CHEMICAL PHYSICS VOLUME 49, NUMBER 1 1 JULY 1968

Magnetic N onequivalence in the High-Resolution NMR Spectra of Diborane

T. C.FARRAR, R. B. JOHANNESEN, AND T. D. COYLE
National Bureau oj Standards, Washington, D. C.
(Received 6 February 1968)

Proton and boron-II nuclear magnetic resonance spectra of 11 B-enriched neat diborane have been measured
over the temperature range from _7 0 to -60°C. The terminal-proton resonance and the liB spectrum
exhibit partially resolved fine structure which arises from the magnetic nonequivalence of the terminal
protons and of the boron nuclei due to long-range spin coupling. Spectra were calculated which agree quite
well with the observed spectra and result in a reasonably accurate determination of the magnitudes and
most of the relative signs of the various coupling constants. These values are: JBB= 'F5 Hz, JBHb= +46.2
Hz, JBH,=+133 Hz, J'Bn, = +4 Hz, I In,Hb I =7.2 Hz, In'Hb (trans or cis) =±14Hz,JHtII, (cis or trans)
= ±6 Hz, I JHtII ...... 1 <3 Hz, Ii (Ht-Hb) = -4.50 ppm.

I. INTRODUCTION in which the concentrations of the other isotopic

varieties have been reduced to a negligible level.
The proton nuclear magnetic resonance spectrum of II. EXPERIMENTAL
diborane has been reported previouslyl-4 to consist of a
septet of quintets arising from the bridge protons and a We have examined the proton and boron-11 NMR
quartet of broad lines of nearly equal intensity from the spectra of neat llB 2Hs over the temperature range
terminal protons. The septet pattern of the bridge- from _7° to -60°C, using commercial 60 and 100
proton resonance arises from spin coupling to two llB MHz spectrometers. The isotopically enriched diborane
nuclei. Coupling between bridge and terminal protons was prepared by lithium aluminum hydride reduction
produces further splitting of each member of the septet of llBF3, obtained from Oak Ridge National Laboratory
into the observed quintet. The four lines in the terminal as the methyl ether complex. The isotopic purity of the
proton region are attributed to coupling between these starting material was stated by the supplier to be
terminal protons and a directly bonded llB nucleus. A 98.2% or greater. The diborane was purified by repeated
further splitting of each of these lines into a triplet by vacuum-line distillation through a -160°C trap until
coupling to the bridge protons is expected on the basis no impurities were detectable in the infrared spectrum.
of this first-order analysis but has not been observed. The product was shown to be free from methyl and
It has been proposed3 that this anticipated splitting is ethyl ethers by comparison of the IH spectrum with
obscured by line broadening arising from the strong that of a sample containing traces of those contamin-
boron-proton coupling and asymmetry of the field at ants. A single sharp impurity peak present in the pro-
the terminal positions. ton spectrum is probably due to the presence of about
More recently, it has been suggested4 that the diffuse 0.1 % ethane.
spectrum in the terminal proton region arises from Numerous trial spectra of diborane, both proton and
spin-spin interactions with nonbonded llB and among boron, were calculated for comparison with the ex-
the symmetry-equivalent but magnetically non- perimental spectrum. The program used was UEA
equivalent protons. The observed IH and llB spectra NMR II by Woodman and Harris,6 with the dimension
reported here, along with the calculated spectra, demon- statements enlarged to meet the requirements of the
state that this is in fact the case. B2H6 molecule. The program was run on the Univac
The proton spectrum observed for isotopically normal 1108 computer housed at NBS; it used approximately
diborane is necessarily a superposition of the spectra of 50 000 core storage locations. UEA NMR II is a modifica-
the three isotopically different molecules, IlB 2H 6, tion of LAOCOON II by Castellano and Bothner-By.6
llBlOBH6, and lOB 2Hs, present, respectively, in ap- The modifications which make UEA NMR II suitable for
proximately 64%, 32%, and 4% abundance. Each of the calculation of diborane spectra are a provision for
these molecules has a different spectrum, and while the equivalence factoring to include sets of magnetically
spectrum of llB2H6 predominates, the contributions equivalent nuclei and a provision for nuclei with spin
from the other species obscure the details of structure greater than !.
and line shape. It is therefore desirable to observe the The calculated spectra were plotted by a Cal-Comp
spectrum of llB2H6 in isotopically enriched samples plotter using equal proportions of Gaussian and
Lorentzian line broadening. The line-broadening param-
1 R. A. Ogg, Jr., J. Chem. Phys. 22, 1933 (1954).
2 J. N. Shoolery, Discussions Faraday Soc. 19,
215 (1955). 6 R. K. Harris, University of East Anglia, Norwich England
3 D. F. Gaines, R. Schaeffer, and F. Tebbe, J. Phys. Chem. 67, (private communication). '
1937 (1963). 6 S. Castellano and A. A. Bothner-By, J. Chern. Phys. 41, 3863
4T. C. Farrar, Ann. N.Y. Acad. Sci. 137, 323 (1966). (1964) .
281
282 FARRAR, JOHANNESEN, AND COYLE

11B2H6
lH-100 MHz
t. -50·C

(a)

l H -l00 MHz
t = -50'C
H ---7 I
I FIG. 1. (a) Proton magnetic resonance
20 Hz
spectrum of llB2H6 at 100 MHz. (b)
Detail of low-field multiplets.
I

"-----~d (b)

11B2H6
11B _19250 MHz
t. -40'C
~H

o Hz 50

L -_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

FIG. 2. Boron-ll magnetic resonance spectrum of llB2HG at 19.250 MHz.

 NMR SPECTRA OF DIBORANE 283

20 Hz
>-------<

FIG. 3. Calculated spectra of llB 2H.

using parameters in Table I with upper
sign (a) detail of low-field multiplets in
lH spectrum; (b) llB spectrum.

(a)

50 Hz

Cn)

eters in best agreement with the experimental spectra Similarly, the llB spectrum shown in Fig. 2 is not a
were 2.5 Hz for protons and 4.5 Hz for boron. simple triplet of triplets, but has a considerably more
elaborate structure. The llB spectrum is also temper-
III. RESULTS AND DISCUSSION ature independent.
The 100-MHz proton magnetic resonance spectrum The very sharp line (linewidth 0.3 Hz or less) in the
of llB2lL! is shown in Fig. 1a; Fig. 1b shows the detail bridge proton region due to the ethane impurity
of the first two multiplets in the terminal proton region. demonstrates quite well that the linewidth in the
The 19.3-MHz boron-l1 NMR spectrum is shown in proton spectrum of IlB2lL! is not an instrumental effect.
Fig.2. An instrumentation check was also made for the llB
As Fig. 1 shows, there is no overlap at 100 MHz of the spectra; the llB linewidths in BH4- and BF4- solutions
bridge and terminal-proton resonances. The portion of were about 0.4 Hz.
the spectrum attributable to the bridge protons is as An attempt was made to determine the J BB coupling
described by earlier workers. The terminal-proton constant by llB- f2HI double resonance experiments
resonance, however, is a symmetric set of four complex on samples containing a high proportion (about 50%)
resonances clearly exhibiting considerable fine structure. of llBIOB2H6• When the 2H nuclei were strongly irradi-
The outer members of the "quartet" are distinctly ated, the broad llB resonance narrowed from a line-
different from the inner lines. The appearance of these width of about 75 Hz down to a linewidth of about 18
envelopes, and their widths, indicate that the under- Hz. This places an upper limit of about 3 Hz on JIOBllB
lying structure is more complex than the triplet ex- and hence JllBll B<9Hz.
pected from a first-order analysis. Observed line shapes Although some quadrupolar broadening is present
for the terminal-proton spectrum are identical at 60 (note the linewidth in the bridge proton resonances),
and 100 MHz, except where the lines from bridge and the relative temperature independence of the lH and
terminal protons overlap at the lower frequency. The llB NMR spectra indicates that quadrupole effects are
fine structure is apparent, but less well resolved in small and not primarily responsible for the broad
spectra of isotopically normal diborane. Over the range envelopes observed.
investigated the appearance of the spectrum is sub- In order to explain or to calculate these spectra, the
stantially independent of temperature. following interactions must be considered: (a) Coupling
284 FARRAR, JOHANNESEN, AND COYLE

20Hz
t------i

FrG. 4. Calculated IH spectrum of

llB2H6 with JBB=O: detail of low-field
multiplets.

between terminal protons and the boron nucleus to contributions to the spectrum from couplings between
which they are bonded, J BH,; (b) coupling between symmetry-equivalent nuclei.
bridge protons and boron, JBH.; (c) bridge-proton- The following parameters were taken as known from
terminal-proton coupling, J H • H ,; (d) coupling between inspection of the lOO-MHz IH spectra:
the terminal protons and the remote boron atom,
J' BH,; (e) coupling between terminal protons on op- o(Ht-Hb) = -4.50 ppm, JH,-H, = 7.2 Hz,
posite ends of the molecule, J H,H,ci8 and J H,H,tran8; (JH,-B+J'H,-B) = 137.0 Hz, JHb-B=46.2 Hz.
(f) geminal coupling between terminal protons,
JH,H,uem; (g) boron-boron coupling, J BB ; (h) the The remaining parameters are JBB and the three cou-
bridge-proton-terminal-proton chemical shift, o(H,- pling constants among the terminal protons. The chem-
H b ). ical shift between H t and Hb is so large that it was
There is no reason to assume that any of these inter- considered to be infinite in the calculations. The ob-
actions is necessarily absent. Two possibilities are of served separation of 137.0 Hz which appears promin-
particular interest: ently in both boron and terminal proton spectra was
assigned to the sum of JH,-B+J'H,-B by analogy with
(1) If J'BH,~O, while J H,H,ci8=JH,H,tran8=0, the less-complex centrosymmetric systems. IO ,13 This re-
terminal protons on opposite ends of the molecule will striction was verified by the calculated spectra.
form two magnetically nonequivalent sets, providing, There are too few sharply resolved lines in the llB2H6
of course, JBH,~J'BH,. For the same reason, the boron spectrum for any iterative program to be completely
atoms will be magnetically nonequivalent. The mole- successful. It was only possible to use trial-and-error
cule must then be described as an A 2A'2M2XX' system,1
and the observed spectrum will reflect the influence of TABLE 1. "Best" NMR parameters for llB2H6
the relative magnitudes and signs of J BH , and J'BH,.
(2) If J H,H,ci8 and J H,H,tran8 are not equal to each Value Error"
other (and JBH,~J'BH,), the terminal protons and the
boron atoms will all be magnetically nonequivalent JBB =F5b ±2.0
f BH • +46.2· ±0.5
and the spectra will be of the AA'A"A"'M2XX' type.
+133 ±1.0
In this case, JH,H,u em will also contribute to the spec- fBH,
±1.0
f'BH, +4
trum. I fH,H.! 7.2d ±0.3
This type of magnetic nonequivalence has been dis- f H,H, (trans or cis) ±14e ±2.0
cussed for many simpler systems. Examples include the fH,H, (cis or trans) ±6e ±2.0
difluoroethylenes,8.9 J3C2H6,lO P 2H 4,11 and 29Si28SiF6.12 I fH,H,u",nl <3Hz ±2.0
li(H,-H.) +4.50 ppm ±0.03
In each of these, the magnetic nonequivalence results in
7 The notation used is that of R. E. Richards and T. Schaefer a This is an estimated error. Values taken outside the error limits clearly
[Mol. Phys. 1, 331 (1958) ] in which primes denote magnetic non- give a poorer calculated spectrum.
equivalence of nuclei with the same chemical shift. ',an,.
b The sign of JBB is opposite the sign of J H, H/ i • and J H, H,
8 H. M. McConnell, A. D. McLean, and C. A. Reilly, J. Chern. • This sign must be positive with respect to JBH, in order for the bridge
Phys. 23, 1152 (1955). proton spectrum to be unchanged in going from 60 MHz to 100 MHz. as
9 T. D. Coyle, S. L. Stafford, and F. G. A. Stone, J. Chern. Soc. is observed experimentally.
1961,743. d Reversing the sign of J H, H. has no effect on the calculated spectra.
10 R. M. Lynden-Bell and N. Sheppard, Proc. Roy. Soc. (Lon- e Same comment as (b). It is a consequence of the symmetry of the
don) A269,385 (1962). problem that one is unable to designate which of these constants is the cis
11 R. M. Lynden-Bell, Trans. Faraday Soc. 57,888 (1961). coupling and which is the trans.
12 R. B. Johannesen, T. C. Farrar, F. E. Brinckman, and T. D.
Coyle, J. Chern. Phys. 44, 962 (1966). 1. R. M. Lynden-Bell, Mol. Phys. 6, 601 (1963).
 NMR SPECTRA OF DIBORANE 285

20Hz
>-------4

FIG. 5. Similar to Fig. 3(a) with JaB =

+5 Hz.

--,._-----------------------------'

methods for adjustment of the parameters. Conse- of the spectra for which JBB , JH,H,ci., and JR,H,tram
quently, it is not possible to say that any kind of have the same sign. The spectral parameters in Table I,
"best fit" has been achieved. Nevertheless, some then, represent one assignment which gives calculated
definite results have been reached. First, the spectrum spectra in very good agreement with the observed
of the bridge protons appears to be first order and the spectra. Although it was not possible to demonstrate
bridge proton portion of the calculated spectrum is that this assignment is unique, we were unable to find
affected only by variation of the parameters JH.H, and any other combination of parameters which gave as
J BR.. Second, the model of B2He with the terminal good agreement.
protons equivalent in pairs is incorrect; for every The spectrum of llB10BHe was also calculated using
choice of parameters tried with this model the spectrum parameters appropriately adjusted for the smaller 'Y
consisted of numerous sharp lines rather than the and greater I of lOB. The llB spectrum was not greatly
broad envelope observed. Third, J BB and Jf BH, are different from that in llB2He. However, the proton
small, but not zero. Fourth, J H,H,ci8 and JH,H,Irans are spectrum was much different, being in general more
small, but not zero. The spectrum is not very sensitive diffuse with even fewer sharp lines than in llB2He. It
to the magnitude or sign of JH,H,uem. All that can be has been clearly demonstrated that a superposition of
said is that the magnitude of this last coupling is less llB2H6 spectra and HBlOBH6 spectra, with the spin-
than 3 Hz. coupling constants given in the preceding paragraph,
A total of 32 1H and 32 HB spectra were calculated. satisfactorily accounts for the diborane spectra previ-
The spectra which agreed most closely with the ob- ously published.
served spectra are shown in Fig. 3. The parameters
ACKNOWLEDGMENTS
used to generate these spectra are given in Table 1.
Figure 4 shows the effect on the lH (terminal) spectrum It is a pleasure to acknowledge the assistance received
when J BB = 0.0 and all other parameters are as given in from Dr. James Ferretti and Mrs. M. L. McNeel,
Table I. This spectrum is typical of those for which National Institutes of Health, in obtaining the plots of
J BB is zero. Numerous variations of the proton-proton the calculated spectra, and Dr. E. Lustig and Mrs. M.
couplings were tried for J BB = +5.0 Hz and for JBB= Lake, Food and Drug Administration, for making
-5.0 Hz. If we assume that J BH , is positive, then J BB available to us their 100 MHz spectrometer. We wish
can be positive or negative as long as the signs of also to thank Dr. T. H. Brown and Dr. E. B. Whipple
JH,H,ci' and JH,H,'ran. are both opposite the sign of for their help in the HB - {2H} double resonance ex-
J BB • Figure 5 was generated by the parameters of periments, and Dr. R. K. Harris for making available to
Table I, but with the sign of JBB reversed; it is typical us the program UEA NMR II.