close

Вход

Забыли?

вход по аккаунту

?

Metal Isotope Effect on Metal-Ligand Vibrations.

код для вставкиСкачать
Metal Isotope Effect on Metal-Ligand Vibrations
BY Kazuo Nakamoto[*]
Exploitation of the isotope effect using heavy metal isotopes now makes it possible to
give clear-cut assignments of the metal-ligand vibrations in the IR and Raman
spectra of complex compounds. T h e results previously obtained with this new method
and their current limitations form the subject of this progress report.
1. Introduction
T h e metal-ligand vibrations of coordination compounds
are most important since they provide direct information about the structure of the complex and the nature
ofthemetal-ligand bond. These vibrations appear in the
low-frequency region because of the relatively heavy
mass of the metal atom and the weak nature of the
metal-ligand bond. It is not an easy task t o assign the
metal-ligand vibrations unequivocally since the interpretation of low-frequency spectra is complicated by
the appearance o f ligand vibrations (some o f which
are activated by complex formation) and by the crystal
field effect (lowering of symmetry, appearance of lattice
modes, coupling of lattice with internal modes, ere.) in
the solid state.
Thus far, the metal-ligand vibrations have been assigned
by the use of one o r more of the following methods:
1. A comparison of the spectra between a free ligand
and its metal complex; the m e t a l - h a n d vibration should
be absent in the spectrum of the free ligand. This
method often fails t o give a clear-cut assignment, since
some ligand vibrations activated by complex formation
may appear in the same region a s the metal-ligand
vibrations.
2. T h e metal-ligand vibration should be metal-sensitive
and be shifted by changing the metal o r its oxidation
state. This method is applicable only when a series of
metal complexes have exactly the same structure with
different metals.
3. The metal-ligand stretching band should appear in
the same frequency region if the metal is the same and
the ligands are similar. F o r example, the Zn-N stretching frequencies of Zn"-a-picoIine complexes are similar
to those of Zn"-pyridine complexes. This method is
applicable only when the metal-ligand vibration is
known for one parent compound.
4. The metal-ligand vibration shows a n isotopic shift if
the ligand is isotopically substituted. F o r example, the
Ni-75 stretching band of INi(NH,),lCI, a t 334 c m - '
is shifted t o 31 8 cm - I upon deuteration of the ammonia
ligands"'. The observed shift of 16 c m - ' is in good
agreement with that predicted theoretically for the
Ni-N stretching mode. Similar shifts are observed if
the a-atom of the ligand (atom directly bonded t o the
metal) is isotopically substituted. F o r example, this
[*I Prof. Dr. K . Nakamoto
The Todd Wehr Chemistry Building
Marquette University
Milwaukee, Wisconsin 53233 (USA)
666
method has been used t o assign the metal-ligand vibrations of oxamido(l4N and I5N)l2' and acetylacetonato(I6O and I8O) c o m ~ l e x e s ' ~
It ~should
.
be noted, however, that metal-ligand vibrations as well as ligand
vibrations involving the motion of the a-atom are shifted by this method. Thus it may not provide a clear-cut
assignment for the metal-ligand vibration.
5 . The frequency of a metal-ligand vibration may be
predictable if the metal-ligand stretching and other force
constants are known apriori. At present, this method is
not practical since not many force constants are known
for coordination compounds.
It is obvious that none of these methods is perfect;
furthermore, they encounter more and more difficulties
as the structure ofthecomplex (and hence the spectrum)
becomes more complicated.
In order t o solve this difficulty, we have recently developed a new method which utilizes the heavy metal isotopesL4'.T h e use of isotopes in vibrational spectroscopy
is not new. The pairs of isotopes such as (H, D),
(6Li, 'Li), ("B, "B), (I4N, "N) and (',O and " 0 )
have long been used t o assign the vibrations involving
the motion of these atoms. However, the use of heavy
metal isotopes had not been reported until our first
communication appeared in 1 969'41"1.This is probably
due t o two reasons: (i) Pure metal isotopes were not
commercially available, and (ii) the magnitude of such
a n isotopic shift was thought t o be too small t o be
observed. The former obstacle has been eliminated since
a number of metal isotopes can be purchased from Oak
Ridge National Laboratory[**'.Table 1 lists some stable
metal isotopes, their purity and current prices. Using
these isotopes, it is possible to prepare a pair of metal
complexes in which only the metals are isotopically
substituted. Although metal isotopes are relatively expensive, this factor does not cause serious financial
burden since most of the preparations can be carried
out o n a milligram scale and normally less than twenty
milligramsofsamples are sufficient t o measure the entire
spectrum. Furthermore, some of these metal isotopes
can be easily recovered after the measurement. Table 1
also indicates that several natural abundant elements
may be used as substitutes for pure metal isotopes
such as 52Cr,56Feand 58Nisince the concentrations of
[*I After our first communication ameared in 1969, f . Turte
and J . frrud/zomme independently reported the application of the
metal isotope technique to inorganic salts such as Ni,GeO, and
CaMoO, (C. R. Acad. Sci. Paris, 270. 474 (1970); Spectrochim.
Acta. 26A. 2207 (1970)).
[**I Isotopes Development Center, Oak Ridge National Laboratory. P. 0. Box X. Oak Ridge. Tennessee 37830 (USA).
Angew. Chem. infernut. Edit.
Vol. 11 (1972) No. 8
these isotopes are exceptionally high in natural abundance.
As stated above, the magnitude of a metal isotope
shift is expected to be small since relative mass difference between isotopes (Arnirn) is small and the metalligand bond is generally weak. However, it is beyond
possible experimental errors in most compounds containing the metal isotopes listed in Table 1. For mass
shifts of the bending modes are less than 2.0 c m - ' in
most cases. Thus, the metal isotope method is not so
effective in assigning the metal-ligand bending modes.
It is well known that the Teller-Redlich product rule
holds for the ratio of products of all the fresuencies
between two isotopic molecules. Consider a n octahedral
XY6-type molecule where X is "Ni or 62Ni and Y is
a n NH, ligand (point mass approximation). If the
Table I . Some stable isotopes available at Oak Ridge National Laboratory.
Element
Atomic
Weight
Chromium
51.996
Iron
55.847
Nickel
58.71
Copper
63.54
Zinc
65.37
M ol y hdenum
95.94
Zirconium
91.22
Palladium
106.4
Tin
118.69
Chlorine
Inventory
Form
ZrO,
Pd
Isotope
"OZr
q4Zr
'04Pd
Io6Pd
'OHPPd
34.453
Natural
Abundance (YO)
Purity (YO)
Price ($/mg)
4.31
83.76
9.55
5.82
91.66
2.19
67.88
26.23
3.66
69.09
30.91
48.89
27.81
18.57
15.84
15-72
9.46
23.78
9.63
51.46
17.40
10.97
27.33
26.71
11.81
14.30
24.03
32.85
5.94
75.529
24.471
90-95
> 99.9
> 95
> 90
> 99.8
>90
98-99
86-99.8
95-99
> 98
> 90
> 99
90-99
95-99
90-98
85--99
85-95
90-99
80-99
90-99
90-99
70 85
> 90
> 93
> 95
> 9s
90-99
90-99
80-96
> 98
> 95
2.20
0.90
0.90
IS O
0.10
3.80
0.15
0.20
I .60
0.20
0.35
0.35
0.55
0.80
0.30
0.30
0.45
0.85
0.50
0.25
I .05
4 25
I .95
I .60
3.65
0.30
0.15
differences of 2 units or more, we have so far observed
shifts ranging 10 t o 1 . 5 c m - ' for the metal-ligand
stretching modes. T h e instrumental resolution and reproducibility of typical far-infrared and R a m a n spectrophotometers is 0.5 c m - ' if proper conditions are used"'.
Therefore, a shift of more than 1 c m - ' can be observed
without difficulty if a band is relatively sharp and not
overlapped by other bands. T h u s far, the heaviest
isotopes we have studied a r e '16Sn and '24Sn. As will
be shown later, the Sn-CI and Sn-N stretching modes
of SnC14(py)2are shifted by 5-4 c m - by such isotopic
substitution. This method may be applicable t o elements
heavierthan tinifapropercombinationofmetal isotopes
and a ligand is found.
'
T h e ligand-metal-ligand bending modes are much less
sensitive t o the metal isotope substitution than the
metal-ligand stretching modes. In fact. the observed
[*I It is not necessary to measure the spectrum under high resolution unless two or more bands are closely located. In most cases,
the spectra are run under normal resolution with a slow scanning
speed and on an expanded scale to measure the peak position as
accurately as possible. The peak frequency is calculated by taking
an average value of multiple scans over the desired frequency
range.
Angew. Chem. internat. Edit. / Val. I 1 (1972) No. 8
0 10
0.50
0.50
1.75
product rule is applied t o the infrared active F,,
species which contains one stretching (v,) and one
bending mode (v2), the result is
This gives a general idea about the magnitude of the
metal isotope shift. However, the product rule does not
give information about the individual isotope shift. In
order t o predict individual isotope shifts for v , and v2.
normal coordinate analysis must be carried out on an
XY6-type molecule. Conversely, metal isotope data are
highly useful in refining a set of force constants t o be
used for normal coordinate analysis. In some cases,
vibrational coupling occurs between metal-ligand and
ligand vibrations. If so, even ligand vibrations become
slightly metal isotope sensitive. The degree of vibrational coupling can be estimated from the calculation
of the L-matrix o r the potential energy distribution.
Thus, the metal isotope d a t a a r e useful in confirming
the presence of vibrational coupling predicted by
normal coordinate analysis.
661
2. Phosphane C o r n ~ I e x e s ' ~ ~ ' ~ * ' ~ ~
Table 2 also shows that two bands a t 358.3 and
234.5 c m - ' of '04Pd(PEt,),CI, are sensitive to the
'04Pd-"0Pd substitution. The former can be assigned to
the Pd-CI stretching mode since it is in a frequency
range expected for the Pd-CI
stretching mode
(370-345 cm-')[81.Then, the band a t 234.5 c m - ' must
be assigned to the Pd-P stretching mode. Previous
investigator^'^.^.^] have assigned the Pd-P stretching
bandsofPd(PEt,),X, and Pd,(PEt,),CI, type complexes
at 450-380
cm-'. The Pd-P
stretching bands of
analogous trimethylphosphane complexes have also
Triethylphosphane complexes o f the type, M(PEt3),X2.
where M is Ni" and Pd" and X is a halogen, are
known t o be trans-planar. Thus, one M-X stretching
and one M-P stretching vibration should be active in
the infrared. Furthermore, both bands should be sensitive t o metal isotope substitution. Figure 1 gives the
actual tracing of the spectra of 58Ni(PEt3)2X2(X = C1
and Br) and their 62Ni analogs"'. Table 2 lists the
observed frequencies, isotopic shifts and band assign-
............
.........
................
............
........
P-
N
(D
............
c
w
n
*
- -
0
m
I
62N I B ~ , I P E ~ J ~
........... SBNiBr,lPEt31,
I
I
I
LOO
350
300
-
IAssL1]
I
I
250
200
I
100
150
50
~~crn-'~
Fig. 1. Far infrared spectra of frans-Ni(PEt,),X, (X = C1 and Br).
ments. T h e bands above 420 c m - ' are not listed since
they are not sensitive t o metal isotope substitution. It is
seen that each complex exhibits two bands which give
large isotopic shifts relative t o others. The band a t
4 0 3 . 3 c m - ' ~f~~Ni(PEt,),Cl,isclearlydueto
theNi-CI
stretching mode since this band is strong and not
present in its bromo a n a l ~ g [ ~ .Then,
'~.
the remaining
band at 273.4 c m - ' must be assigned t o the Ni-P
stretching mode. This band persists in '*Ni(PEt,),Br,
at 265.0 c m - ' . Previously, the Ni-P stretching modes
of the Ni(PEt,),X, type complexes were assigned at
426-41 0 cm L6,71. These assignments must be revised
in view of o u r new results.
-'
been assigned at 380-340 c m - ' by many other invest i g a t o r ~ ' ~ . ' ~ .O
' "ur
. results show that these frequencies
are too high for the Pd"-P
stretching modes of
square-planar complexes.
The skeletal bending bands which appear below 200
c m - ' are much less sensitive t o metal isotope substitution and more difficult to assign than the metal-ligand
stretching modes. In Table 2, we have proposed band
assignments which were predicted from a n approximate
normal coordinate analysis described later. The band at
183.8 c m -' of '04Pd(PEt3),CI, gives a n isotopic shift of
1.8 c m -' which is rather large for a pure ligand vibra-
Table 2. Infrared frequencies of triethylphosphane complexes containing metal isotopes (cm - 9.
PEt,
J8Ni(PEt,)2CI,
V
V
408(m)
416.7(m)
403.3 (vs)
372.5(m)
329.0(m)
273.4 (s)
(hidden)
200.2(m)
186.5(shl
161.5(vw)
106.8(m,br)
-
365(m)
330(m)
-
245(m)
AOIal
(0.0)
(6.7)
(-0.1)
(-0.5)
(5.9)
(0.8)
(-0.2)
(-0.5)
(-0.7)
58Ni(PEt,),Br,
A 0 [a1
IMPd(PEt3)2C12
A0 [a1
Assignment Ibl
0
V
413.6(m)
(1.2)
337.8(m) (10.5)Lcl
374.0(m)
(1.1)
327.8(s)
(0.5)Icl
265.0(m)
(4.7)
(hidden)
190.4(s)
(0.7)
155.l(m)
(1.5)
(hidden)
106.0(m) (-0.4)
41 3.0
358.31~~)
375.7bh)
330.8(m)
234.5 (s)
272.0(m,br)
183.8 (s)
168.0(~h)
152.O(vw)
105.5(s.br)
6(CCP)
v,(M-X)
F(CCP)
6(CCP)
v,(M-P)
G(CCP)
6(CPC)
6(MX)
6(MP)
(-0.3)
(3.4)
(-0.3)
(0.0)
(2.5)
(0.0)
(1.8)
(-0.5)
(1.0)
(0.5)
?
[a] AS indicates metal isotope shift, S(J8Ni)-S("Ni) or Wo4Pd)-S("OPd).
Ibl v,, asymmetric stretch, 6 , bending. Ligand vibrations were assigned according to Ref. [121.
Icl In 62Ni(PEt,),Br,, these two bands are overlapped (see Figure 1). Therefore. A0 values are only
approximate.
668
Angew. Chem. infernat. Edit. 1 Vol. I I (1972) 1 No. 8
3 are sufficient t o show that the magnitudes of the
observed shifts are at least reasonable.
tion. This may suggest the presence of vibrational
coupling between this ligand vibration and the Pd-P
stretching mode at 234.5 cm-I.
Similar isotope studies have also been made o n the
Ni(diphos),X, type complexes where diphos is ethylenebis(dipheny1phosphane) and X is a halogen[13'. In this
case, two Ni-X and two Ni-P stretching modes are
infrared active since these compounds have a cis-planar
NiP,X, skeleton. Table 4 lists the four stretching frequencies determined by the Ni isotope method. It also
compares the Ni-X and Ni-P stretching frequencies
of trans-planar, cis-planar and tetrahedral Ni(phosphane),X, type complexes. It is seen that there are
marked differences in Ni-X
stretching frequencies
between cis and trans complexes. The Ni-X stretching
frequencies of the cis-complex are always lower than
those of the corresponding trans-complex, since the
Ni-X bond of the cis complex is trans to the phosphane ligand which exerts a strong trans-effect. Conversely, the Ni-P
stretching frequencies of the cis-
Ni(PPh,),CI, is tetrahedral whereas Pd(PPh,),Cl, is
trans-planar. Thus, the former exhibits two NI-CI and
two Ni-P stretching bands whereas the latter exhibits
one Pd-CI and one Pd-P stretching band. In fact,
58Ni(PPh,),C12 exhibits four bands a t 341.2, 305.0,
189.6 and 164.0 c m - ' which are sensitive t o the
58Ni-62Ni substitution. T h e former two bands are
assigned t o the Ni-CI
stretching modes since they
are relatively strong and absent in the bromo analog.
Then, the latter two bands must be assigned to the
Ni-P stretching modes. In '04Pd(PPh,),CI,, two bands
at 360.3 and 191.2 c m - ' show large isotopic shifts
(4.3 and 4.2 cm-I, respectively) relative t o others.
T h e former band is in the region of terminal Pd-Cl
stretchingfrequencies[*'and assigned safely t o this mode.
Then, the latter must be assigned t o the Pd-P
Table 4. Comparison of Ni-X
and Ni-P
stretching frequencies between various configurations (cm -').
tr-"Nl(PR ,R ') 2X
b
c r\-'XNi(diuhos)X2
b
tet-Ni(PPh,)ZX2
b
v(N1-X)
V(N1-P)
v(N1-X)
v(N1-P)
v(N1-X)
v(N1-P)
X=CI
401 Ial
252 la]
379 Id]
Br
338 [bl
265 Ibl
I
260 Icl
252 Icl
341 [dl
328
290
266
260
212
341 [el
305
265 LO
232
215 Lfl
190 [el
164
193 L f l
184
198 [fl
182
-
365
308
353
278
[a] R = Ph, R'= Et; Ibl R = R ' = Et (Ref. 171); [cl R = Me, R ' = Ph (natural abundance Nil; [dl Form A
(see Ref. 1131); [el 58Nicomplex (see Ref. 151): [fl see Ref. [141.
stretching mode. Again, previous workers'91 assigned
the Pd-P
stretching modes of Pd,(PPh,),CI, at 450
and 428 c m - ' . These bands may be assignable t o the
vibrations of the triphenylphosphane ligand.
In order t o calculate theoretical isotope shifts, we have
carried out a normal coordinate analysis o n a transplanar MY,X, model by assuming that the PEt, ligand
is a single atom having the mass of PEt,. Evidently,
such a n approximation provides only a rough measure
oftheoretical isotope shifts. Even so, the results of Table
Table 3. Comparison of theoretical and observed isotopic shifts
for trans-Ni(PEt,),X, (cm -9.
Ni(PEt ,)2C12
Theoretical [a]
AS
58Ni
Observed
58Ni
AS
Ni-CI Stretch
Ni-P Stretch
Ni-CI Bend
NI-P Bend
402.4
270.8
189.5
168.2
1.7
403.3
273.4
186.5
161.5
Ni(PEt J2BrZ
Ni-Br Stretch
Ni-P Stretch
Ni-Br Bend
Ni-P Bend
340.9
265.9
141.0
147.1
8.6
7.5
0.1
0.5
337.8
265.0
155.1
- [cl
7.2
7.3
0.0
6.7
5.9
-0.2
-0.5
10.5 Ibl
4.7
1.5
[a1 The molecular parameters used for calculations were: NI-CI,
2.301\; Ni-Br, 2.40A; Ni-P, 2.25.k The four angles around
the Ni atoms were assumed to be 90-. The Urev-Bradley force
1.25 for
constants used were: Kfstretchind, 1.35 for Ni-C1,
Ni-Br and 0.75 for Ni-P; H(bending), 0.1 1 and F(repulsive),
0.13. all in units of mdyn/A.
[bl See footnote [cl of Table 2.
[cl Hidden by the neighboring band.
Angew. Chem. internal. Edit.
Vol. I 1 11972) 1 No. 8
complex are always higher than that of the corresponding trans-complex since the Ni-P bond is trans
t o the halogen.
It is well known that complexes of the type,
Ni(PPh,R),Br, (R = alkyl) exist in two isomeric forms:
tetrahedral (green) and trans-planar (brown) forms"".
According to Table 4, the Ni-Br and Ni-P stretching
frequencies of these two forms should be markedly
different. This has also been confirmed by our N i isotope experiments o n both forms of Ni(PPh2Et)2Br2"61.
In general, the tetrahedral form exhibits two Ni-Br
stretching bands near 250 c m - ' whereas the transplanar form exhibits one Ni-Br stretching near 330
c m - ' and one Ni-P stretching band near 260 cm-I.
3. 8-Quinolinolato (Oxinato) Comdexes" 71
Figure 2 shows the far-infrared spectra of potassium
oxinate (abbreviated as KQ) and several anhydrous bis(oxinato) complexes (MQ2). These spectra are too complicated t o allow empirical band assignments. It is possible, however, t o assign the metal-ligand vibrations
by using the metal isotope method and to elucidate
the structure of these complexes based on observed
numbers of metal-ligand vibrations. There are three
probable structures for bis(oxinato) complexes: transplanar (C2,,), cis-planar (C,,,) and tetrahedral (C2")
669
structures. The number of infrared active metal-ligand
stretching bands should be two for C,, symmetry (2BJ
and four for C,, symmetry (2 A , + 2 9, for cis-planar,
and 2 A , + B, + 9, for tetrahedral).
Anhydrous copper oxinate (CuQ,) has been isolated as
two crystal forms, ie., a-form and B-form. As is seen
in Fig. 2 and Table 5 , both forms exhibit two bands
(332.1 and 297.3 c m - ' for the a-form and 324.0 and
289.7 c m - ' for the B-form) which give relatively large
isotopic shifts by the 63Cu-65Cusubstitution. Thus the
structures of both forms must be trans-planar (C,,,).
The present results are in good agreement with those
of previous X-ray studies"8-201.
In general, the metal-oxygen stretching band is stronger
than the metal-nitrogen stretching band in the infrared.
Therefore, the stronger bands (332.1 c m -' (a) and
324.0cm-' (PI) with higher frequencies may be assigned
to the 0 - 0 stretching modes, whereas the weaker
ones (297.3 c m - ' (a) and 289.7 c m -' (B)) with lower
frequencies may be assigned to the Cu-N stretching
modes.
lj3 - Form I
___.
-....
58
NIQ,
6,
........ .._
64
68
4 00
-
........___
I
I
I
300
200
100
c- F
[cm-']
Fig. 2. Far-infrared spectra of KQ, CuQ,, NiQ,, and ZnQ, (Q =
anion o f 8-quinolinol).
Figure 2 and Table 5 show that four bands a t 309.6,
280.8, 242.0 and 231.0 c m -' of NiQ, show large
isotopic shifts relative to other bands by the 58Ni-62Ni
substitution. NiQ, is also known to have a magnetic
moment of 3.24pBL2".Combining these results, it may
be concluded that NiQ, is tetrahedral (C2J.Generally,
the asymmetric stretching mode appears at a higher
frequency and is expected to give a larger isotopic
shift by the 58Ni-62Nisubstitution than the symmetric
one. Considering this trend with their relative intensities
and frequencies, we have assigned two bands a t 309.6
and 280.8 c m - ' to the asymmetric (B,) and symmetric
(A,) Ni-0
stretching modes, respectively, and two
bands a t 242.0 and 231.0 c m - ' to the asymmetric (B,)
and symmetric (A,) Ni-N stretching modes, respectively. It should be noted in Fig. 2 that the Ni-N
stretching bands (242.0 and 231 .O c m -' ) ,are relatively
strong since they are overlapped on the strong ligand
band at around 230 cm-'.
Table 5. Observed freauencies. isotopic shifts, and band assignments for MQ,-type metal oxinates and potassium oxinate (c m-9.
KQ
a
%uQ,
A?[al
B
642 (s)
583 (s)
645.7 (m)
634.3 (s)
0.0
0.5
646.0 (m)
633.0 (s)
553 (m)
488 (m)
481 (m)
585.3 (m)
524.0 (vs)
0.0
0.0
583.0tm)
522.5 (vs)
-
406.0 (vs)
0.0
404.0 (s)
j*NiQ,
AJIal
(I:;
0.0
0.0
-0.5
327 (m)
288 (w)
-
(overlapped) [bl
274.5 (w)
0.0
332.1 i s )
3.0
(overlapped) [bl
277.0 (w)
0.0
324 (vs)
3.5
-
-
224 (sh)
210.5 (s)
0.3
-
297.3 i m )
2.8
-
-
-
-
180 (vs, vbr)
181.5 (m)
168.4
163.2
153.0 (vw)
0.1
0.6
0.4
182.0(m)
168.2 (m)
0.0
0.0
-
-
-
-
141 (vw)
0.0
-0.3
-
110(m)
-
159 (w)
149 (w)
128 (m, br)
97 (m)
~ _ _ _ _
91 (w)
0.0
-
i
21 8.0 (s)
202.4 (w)
289.7 (m)
151 .O (vw)
- 0.4
A? [a1
Assignment
~~
649.8 (s)
627.5 (m)
613.6 (sh)
606.0 (sh)
598.0 (s)
574.0 (m)
519.0 (sh)
1:;:
:?
409.2 (s)
388.5 (sf
0.2
0.5
0.3
0.0
0.0
0.0
0.6
{f~~~d~~~
0.0
:(I
6 (ligand)
6 (ligand)
0.0
0.0
6 (ligand)
6 (C-0) Ibl
409.5 (s)
0.0
403.0 (s)
0.0
389.0 ( m )
0.0
374.0 (w)
0.0
312.5 (w)
0.0
279.5 (w)
0.0
243.0 I v s )
5.0
214.4 /s)
4.2
(overlapped)[bl
ring
deformation
195.5 ( 1 7 7 )
182.4 im)
180.4 (m)
170.0 (w)
158.5 (w)
v,(M-N)
v,(M-N)
ligand
ligand
ligand
ligand
ligand
!Hi {
0.5
330.0 (m)
290.0 (m)
309.6 i s )
280.8 ( S J
246.0 (vs)
-0.6
0.4
6.0
4.8
0.0
242.0 l s s )
231.0 i s )
184.0 (w)
3.8
3.2
0.1
153.8 (vw)
131.(w, br)
0.3
0
561.8 (m)
503.5 (vs)
ligand
hand
v,(M-O)
v,(M-O)
ligand
0.0
3.2
-
2.5
2.0
0.0
0.2
0.0
-
-
138 (w, br)
131 (w , br)
0
_ _ _ _ _ _ ~
[a] A v indicates metal isotope effects; v(~'C U)- v ( ~ ~ C Uv(S8Ni)
).
- v(62Ni)o r v('Zn)[bl Overlapped with v(M-0).
670
MZnQ,
A? Ial
0
v(68Zn).
Angew. Chem. internat. Edit. / Vol. I 1 (1972) / No. 8
In accordance with the expected tetrahedral structure,
ZnQ, exhibits four bands which are sensitive t o the
64Zn-68Zn substitution. Figure 2 also shows that the
general pattern of the far-infrared spectrum of ZnQ, is
similar t o that of NiQ, but is rather different from
that of CuQ, (trans-planar). F r o m the intensity and
freauency considerations, these four bands at 243.0,
214.4, 195.5 and 182.4 c m - ' may be assigned to the
asymmetric Zn-0 (Bl), symmetric Zn-0 (Al), asymmetric Zn-N (B,) and symmetric Zn-N ( A , ) stretching modes, respectively.
4. 2,CPentanedionato (Acetulacetonato)
T h e band assignments o f t h e Cr-0 stretching modes of
tris(acetylacetonato)chromium(iii)have been a subject
of controversy. Originally, Nakamoto et ~ 2 1 . ' ~assigned
~'
the bands a t 459 and 307 c m - ' (calculated) to these
modes based on normal coordinate analysis of the
1 : 1 (metaliligand) model. Gillard er al.'241
extended the
measurement down t o 55 c m - ' and assigned two bands
at 459 and 355 c m- I t o the Cr-0 stretching modes.
On the other hand, Pinchas et
assigned the
592 c m - ' band t o the pure Cr-0 stretching and the
456 c m - ' band t o the C-CH,
bending coupled with
the Cr-0 stretching mode since the former gave the
substitulargest isotopic shift ( 1 9 c m - ' ) by 160-'s0
tion of the ligand. In order to solve this controversy,
we have compared the infrared spectra of 50Cr(acac),
and its 53Cr analog. As is seen in Fig. 3 and Table 6,
the observed isotopic shift of the 592 c m - ' band
(0.7 c m - ' ) is much smaller than those of the 459 and
335 c m - ' bands (3.0 to 3.9 cm-I). Thus, it is most
reasonable to assign the latter two bands to the Cr-0
stretching modes, and the former band at 592 c m - ' to
a ring deformation in which the oxygen a tom is
displaced appreciably.
r r i ea drigorous normal
Recently, Mikami et ~ l . ' ~ ~ ~ c a out
coordinate analysis o n Fe(acac),. In agreement with
our metal isotope study, their results show that the
592 c m - ' band is due to an out-of-plane ring deformation and the bands a t 459 and 355 c m - ' are due to
the Cr-0 stretching modes. As stated before, the band
shifts due to isotopic substitution of the a-atom of the
ligand must be interpreted with caution since the
metal-ligand stretching as well as ligand vibrations
involving the motion of the a-atom are shifted by
such substitution.
5. Complexes of 2,2'-Bipyridyl and
l.lO-Phenanthr~line[~~,~~~
The assignments of the metal-ligand vibrations of triscomplexes of 2,2'-bipyridyl (bipy) and1 .lo-phenanthroline (phen) have been controversial. Originally.
Inskeeo[281
assigned the M-N stretching bands of tris(bipy) complexes of Co", Ni", Cu" and Zn" in the
300-260 c m - ' region by assuming that the highest
freauency band which is absent in the free ligand is
the M-N
stretching vibration. On the other hand,
Clark and Williams[291
assigned these bands to ligand
vibrations and concluded that no bands above 200 c m can be assigned to the M-N stretching modes. In order
to solve this controversy, we have carried out an
extensive study o n the far-infrared spectra of these and
many other bipyridyl complexes by using the metal
Fig. 3. Far-Infrared spectra of "Cr(acac), and its 5 i C ranalog.
Table 6. Observed freauencies, band assignments and isotopic shifts for Cr(acac), (cm -9.
. ..
Nakamoto, er al. I231
Pmchas, et al. I31
A: [a]
:$,'I
n(CH)
667. ring + v(Cr-0)
658.6(CR) v(Cr-0)
609.
5943 z?
459. v ( C r 0 )
416, n?
(373) [dl, ring
(207) [dl. v(Cr-0)
(277) Id], 6(CR)
(202) [dl, ring
+
678. ring
661. x
+ v(Cr0)
592, v(Cr-0)
456, 6 ( C R ) + v(CrO)
416, 6tOCrO)
11
3
19
5
8
Present Work f221
'OCr
AG [bl
791.5
776.0
772.0
68 I .O
659.5
613.2
596.5
463.4
417.8
358.4
250.0 [C]
224.0 [c]
179.0 [c]
Mikami. et al. I251
0.0
0.5
0.0
1 .o
0.7
3.0
0.0
3.9
-
?("O) - c('xo).
[bl J('OCr) - J ( "Cr).
[cl Cr natural abundance. T h e spectra of the comulexes containing pure metal isotopes were not
obtained below 300 c m - ' because the amount o f the sample available was not sufficient to run the
Nujol mull spectrum on our instrument.
[dl Calculated fresuencies.
Abbreviations: v. stretching: z. out-of-Dlane bending: ring, in-plane ring deformation.
[dl
Angew. Chew. internat. Edit. / Vol. Z I (1972) / No. 8
67 1
isotope m e t h ~ d [ * ~ .The
~ ' ~ M-N
.
stretching frequencies
thus determined are listed in Table 7 together with their
electronic configurations and magnetic moments.
Table 7. Electronic configuration, magnetic moment and M-N
stretching freauencies of tris(bipyridy1) complexes.
Metal
Electronic
Configuration
Magnetic
Moment [pJ
M-N Stretching
Freauencies [cm-']
3.78
2.9
2 .o
5.95
385, 349
351. 343
371, 343
224, 191
384.367
382, 308
386, 376
378, 370 [a]
258.227
266.228
235, 184
244, 194
282.258
280,257
291.268
230.184
0
0
0
4.10
4.85
3.71
3.3
3.1
2.23
0
ture of this complex must be trans-octahedral since no
isotopic shifts have been observed in the Raman
spectrum. This is also supported by the fact that the
mutual exclusion rule holds between infrared and
Raman spectra of this compound (Table 8 and Fig. 4).
F o r the trans-octahedral structure, one Sn-CI and one
Sn-N(py) stretching mode should be infrared-active
and both bands are expected t o show relatively large
shifts by 116Sn-124Sn
substitution. In agreement with
this, two bands at 323.0 and 227.5 c m - ' of the [I6Sn
complex are shifted by ca. 4 t o 5 c m - ' while other
bands show almost no shifts by the metal isotope
substitution in the infrared spectra. Since the former
band is strong and insensitive t o the p y - [ D 5 ] p ~substitution, it can be assigned to the Sn-Cl stretching
mode. The latter band is relatively weak and sensitive
to the ligand deuteration. Hence, this band is assigned
t o the Sn-N stretching mode. These and other band
assignments listed in Table 8 have been confirmed by
an approximate normal coordinate analysis o n a transSnX,Y2 model where Y (single atom) represents a
pyridine ligandl3'].
[a1 Values obtained for [Co(phen),l(CIO,),.
As is seen in Table 7, each complex exhibits two M-N
stretching bands and their frequencies are governed by
the electronic structure of the central metal. All the
complexes which have filled o r partially filled t2g(bonding) orbitals and empty e p(anti-bonding) orbitals exhibit
theM-Nstretchingbands in the 390-300 c m - region,
whereas other complexes which have filled o r partly
filled eg orbitals exhibit the M-N stretching bands in
the 290-180 cm-I region. Thus, a dramatic decrease
in the M-N
stretching frequency is seen in going
from Co"' t o Co". However, the M-N stretching frequencies d o not change appreciably in the series Cr"',
Cr", Cr' and Cro although the oxidation state of the
metal changes in a wide range. This result may suggest
that, as the formal oxidation state is lowered, o-bonding
becomes weaker and n-bonding becomes stronger so
that the overall M-N bond strength is approximately
constant. If this is the case, there will be considerable
interaction between metal d n orbitals and the antibonding ligand n* orbitals so that increasing amounts
of electrons of the metal will reside essentially in ligand
orbitals as the oxidation state is lowered. Thus, Cr(bipy)3 may be represented as Cr3'+ (bipy'-)3 as suggested by Koenig and H e r ~ o g " ~who
'
studied the
electronic spectra of these complexes.
..._.......__
124 sn
>
'
c
:= 1
v)
c
m
c
%
U
-
J
T [crn'~
Fig. 4. Low-frequency infrared and Raman spectra of transSnCl,(~y),.
Table 8. Vibrational spectra of trans-SnCI,(pyridine), (cm- '1
~~
S(Il6Sn)
lnfrared
AS [a]
AC [bl
Raman
AS(Sn)
-
5.0
-
-
307.0
244.8
227.5
187.0
4.0
I .O
-
323.0
-
-
170
1 .o
-
-
144.5
-
90
0
-
173.5
-
157.3
-
116.0
Assignment
v(Sn-Cl), E,
v(Sn-C1). A,,
v(Sn-CI), B,,
v(Sn-N), AzL,
NCISnN). E,
G(CISnC1). B2%
G(C1SnCI). E,
v(Sn-N), A , ,
G(ClSnN), A2"
F(C1SnN). E,
lattice
6 . Complexes of P y r i d i r ~ e [ ~ ' ' ~ ~ '
T h u s far, the application of the metal isotope technique
has been limited t o infrared spectroscopy. This technique
is also useful in assigning the metal-ligand vibrations
in Raman spectra. If a molecule contains one metal
atom at the center of symmetry, n o metal isotope
shifts are expected for all Raman-active fundamentals.
Therefore, the lack of metal isotope shifts in the
R a m a n spectrum proves the presence of a center of
symmetry. As an example, Table 8 lists the Raman
frequencies of S n C l , ( p ~ ) (py
~ = pyridine). T h e struc672
In a tetrahedral ZnXz(py)z type complex, two Zn-X
and two Zn-N(py) stretching modes are infrared as
well as Raman active. Furthermore, all these four
modes should be sensitive t o 64Zn-68Zn substitution.
As is seen in Table 9 and Fig. 5, four infrared bands
at 329.2, 296.5, 222.4 and 203.9 cm-I of 64ZnCI,(p~)2
show relatively large shifts when 64Zn is substituted by
68Zn.The former two bands are strong and show little
Angew. Chem. internat. Edit. 1 Vol. I 1 (1972) J No. 8
shifts by the py-[DS]py substitution. Therefore, these
bands can be assigned t o the Zn-C1 stretching modes.
The latter two bands are relatively weak and sensitive
to t h e ligand deuteration. Thus, these two bands can
be assigned t o the Zn-N stretching modes.
to the A , and the latter to the F, species o n the basis
of observed isotopic shifts alone. This assignment has
also been confirmed by the observation that the 427
c m - band of a n approximately 4.0 M ammonia solution
of [Zn(NH,),] I, is strongly polarized.
Similar results are seen in the R aman spectra of these
compounds. In general, the antisymmetric mode is
stronger in the infrared and weaker in the Raman
Other vibrations such as N-H stretching, N H , degenerate and symmetric deformation, and N H , rocking
modes d o not involve the motion of the metal atom
'
c
c
c
350
300
-
250
200
T [cm-'~
150
Fig. 6. Low-frequency Raman spectra of [h4Zn(NH,),11, and
68Zn analog.
Fig. 5 . Low-frequency infrared and Raman spectra of ZnC12(pY)2.
I ~ S
Table 9. Vibrational spectra of Zn(pyr),c12 ( c m - ' )
_ _ _ _ ~
~~
iXaZn)
Infrared
AC la1
AC [bl
C("4Zn)
Rainan
AG bdl
329.2
296.5
222.4
203.9
4.8
2.4
3.6
2.4
0.8
0.2
3.8
4.1
-
-
330.3
294.8
224.1
207.6
200
151.2
3.9
1.3
4.4
3.2
0.0
0.0
1 .o
0.4
5.0
4.2
I .o
6.6
-
-
-
I54
143.3
109.9
~
-
[cl
0.3
0.0
7
4.7
0.7
-
A t Ibl
v(Zn-CI), B,
v(Zn-CI). A ,
v(Zn-N), B ,
v(Zn-N), A ,
G(CIZnN), A,
G(NZnN), A ,
G(C1ZnN). B,,B,
G(CIZnC1). A ,
[a1 C(MZn) - C("Zn).
[bl C(py)-?([D,lny).
[cl Isotopic shifts were not determined due to Door shape of the band.
spectrum than the symmetric mode. Therefore, the
bands at ca. 329 and 222 c m - ' are assigned t o the
antisymmetric whereas the bands at ca. 297 and 204
cm - are assigned t o the symmetric typesL321.
'
and show little shifts by the 64Zn-6sZn substitution.
In order to calculate theoretical isotopic shifts, we have
carried out a normal coordinate analysis on the whole
[Zn(NH,),]" ion (Tdsymmetry). As is seen in Table 10,
the results are in good agreement with those observed.
7. Ammine Complexes'331
8. Other Complexes
Theoretically, two metal-nitrogen stretching vibrations
( A , and F,) should be Raman-active for tetrahedral
(T,) ammine complexes. The totally symmetric A , mode
involves no motion o f t h e central metal at o m and should
thus be independent of the b4Zn-68Zn isotopic substitution. On the other hand, the F, mode involves
the motion of the central metal at o m and should be
sensitive t o the isotopic substitution. Fig. 6 shows the
Raman spectra of crystalline [64Zn(NH3),lI, and its
68Zn analog in the Zn-N stretching region. It is seen
that the 432.0 c m - ' band of the 64Zn complex gives
a shift of 0.5 c m - ' which is within the experimental
error, whereas the 412.0 c m - ' band gives a large shift
of 2.0 c m - I . Thus, it is possible to assign the former
Angew. Chem. internal. Edit.
Vol. I 1 (1972) N o . 8
In addition to those complexes listed above, the metal
isotope method has been applied to mercaptoalkylamine complexes (Ni" and Pd")'341, allyl-Pd com~ l e x e s ' and
~ ~ ] f e r r ~ c e n e ' ~Recently,
~].
this method has
also been applied to metal complexes of octaethylporphin (Ni" and Zn") by Buerger et al.1371.
In principle, the metal isotope method is applicable to
any inorganic and coordination compounds. Practically,
however, its application may be limited by several
factors :
1. Pure and stable metal isotopes are available only for
certain elements.
673
2. Some metal isotopes are too expensive even for a
small scale synthesis.
3. Some compounds are difficult to prepare o n a small
scale. This is particularly true if a compound is gaseous
or liquid.
I1 31 C . Udovich. J . Takemoto, and K . Nakamoto. J. Coord. Chem.
1 , 89 (1971).
1141 P . M. Boorman and A . J. Cartv, Inorg. Nucl. Chem. Lett. 4,
101 (1968).
I151 R . G . Havter and F. S. Humic, Inorg. Chem. 4, 1701 (1965).
[16]J . T . Wang. C. udovich, K . Nakamoto, A . Quattrochi, and
J . R . Ferraro, Inorg. Chem. 9. 2675 (1970).
Table 10 Comparison of observed Raman and calculated frequencies for the [Zn(NH,),lz- ion(cm '1
~
obs
3233
1253
432.0
3275
1596
685
156
-
3275
3150
1596
1239
685
412.0
156
P Z n ( N H ,)412
calc. la1
3213.5
1262.1
432.9
3270.5
1595.3
680.8
162.2
3270.0
1594.8
678.0
3270.3
3214.4
1595.2
1259.5
686.0
410.1
158.8
[6*Zn(NH,)412*
obs.
calc. [a]
3233
1253
431.5
3275
1596
685
155
-
3275
3150
1596
1239
685
410.0
155
~~
Band
assignment
3213.5
1262.1
432.9
3270.5
1595.3
680.8
162.2
3270.0
1594.8
678.0
3270.3
3213.4
1595.2
1259.5
686.0
407.6
157.4
A;
[a1 The molecular parameters used were: Zn-N, 2.01
N-H.
I.OOA,. All the angles were taken to
be tetrahedral. The Urey-Bradley force constants used were: K (stretching), 5.66 (N-H) and 0.54
(Zn-N); H (bending), 0.018(NZnN), 0.048 (ZnNH)* and 0.43 (HNH); F (repulsive), 0.19 ( N . . .S),
0.24( 2 . . H ) and 0.09( H . ..H), all in units of mdyniA.
4. Observed shifts may be too small for complexes
containing heavy elements such as Os, Pt and Pb.
The work described above was mainly supported by the
Petroleum Research Fund administered by the American
Chemical Society (ACS-PRF-3318-C3,5) and the U . S .
Army Research Office (DA-ARO-D-31-124-G1130). The
author wishes to express his thanks to his colleagues,
former students and postdoctoral fellows for their contribution fo this project.
Received. January 4. 1972 [A 894 IEI
German version: Angew. Chern. 84,75511972)
[ 1 I L . Sacconi,
( 1964).
A . Sabatini, and P . Cans, Inorg. Chem. 3, I772
121 P . X . ArmendarezandK. Nakamoto, Inorg.Chem.5,796(1966).
131 S. Pinchas, B. L. Silver, and I . Laulicht. J . Chem. Phrs. 46.
1506 (1967).
I41 K. Nakamoto. K . Shobatake, and B. Hutchinson. Chem.
Commun. 1969, 1451.
[51 K. Shobatake and K. Nakamoto. J. Amer. Chem. SOC. 92,
3332 (1970).
[61 P. L . Goggin and R . J. Goodfellow, J. Chem. SOC. A1966,
1462.
[71 G . E. Coates and C . Parkin, J. Chem. SOC.1963, 421.
[81 D . M . Adams and P. J. Chandler. Chem. Commun. 1966, 69.
191 R . J . Goodfellow, P . L . Gosgin, and L . M . Venanzi, J. Chem.
SOC.A1967, 1897.
[lo] R . J . Goodfellow, P . L. Goggin, and D . A . Duddell, J . Chem.
SOC.A1968, 504.
[ I l l P . J , Park and P . J . Hendra, Spectrochim. Acta 2 5 A , 227,
909 ( 1 969).
I121 J . H . S.Green, Spectrochim. Acta 2 4 A . 137 (1968).
614
1171 N . Ohkaku and K. Nakainoto, Inorg. Chern. 10, 798 (1971).
1181 R . C . f f o y and R . H . Morris, Acta Crystallogr. 22. 476
(1967).
[191 F. Kanakubo, K . Ogawa, and I . Nirta. Bull. Chem. SOC.Japan
36, 422 (1963).
1201 G. J . Palenik, Acta. Crystallogr. 17, 687 (1964).
[211 W . Matoush and F. Basolo, J. Amer. Chem. SOC. 75, 5663
(1953).
[221 X.Nukamoto. C. Udovich, and J . Takemoto, J . Amer. Chem.
SOC.92. 3973 (1970).
1231 K. Nukamoto, P. J . McCarthy, A . Ruby, and A. E. Martell, J .
Amer. Chem. SOC.83, 1066 (1961).
I241 R. D. Gilland, H . G . Silver. and J. L. Wood, Spectrochim.
Acta. 20, 6 3 ( 1 9 6 4 ) .
(251 M . Mikami, i. Nakanawa, and T . Shimanouchi, Spectrochim.
Acta 2 3 A , 1037 (1967).
f261 B. Hutchinson, J . Takemoto, and K . Nakamoto, J. Amer.
Chem. SOC.92, 3335 (1970).
I271 J . Takernoto, B. Hutchinson, and K. Nakamoto, Chem.
Commun. 1971, 1007.
[281 R . G . Inskeep, J. Inorg. Nucl. Chem. 24, 763 (1962).
(291 R. J . f f . Clark and C . S. Williams, Spectrochim. Acta 2 3 A .
1055 (1967).
I301 E. Koenisand S. f f e r z o g ,J. Inorg. Nucl. Chem. 32,585 (1970).
1311 N . Ohkaku and K . Nakamoto, t o be published.
1321 Y. Saito and K. Nakamoto, t o be published.
[331 J. Takemoto and K . Nukamoto, Chem. Commun. 1'970, 1017;
K. Nakamoto. J . Takemoto, and T . L . Chow. APPL Spectry. 25.
352 (1971).
I341 C. W . S c h b f e r and K . Nakamoto. to be published.
1351 K . Shobatake and K . Nakamoto, J. Amer. Chem. SOC. 92,
3339 (1970).
1361 K. Nakamoro, C . Udovich, J . R . Ferraro, and A . Quattrochi,
Appl. Spectry. 24, 606 (1970).
I371 H . Buerger, K . Burctvk. and J . H . Fuhrhop. Tetrahedron 27,
3257 (1971).
Angew. Chem. internat. Edit. / Vol. !I (1972) / N o . 8
Документ
Категория
Без категории
Просмотров
0
Размер файла
784 Кб
Теги
effect, metali, vibrations, ligand, isotopes
1/--страниц
Пожаловаться на содержимое документа