‘ in viiqdulfl‘ulr
, , ‘ 139%! I.
I .‘ h » L ‘
'- .
. ’. '-. _ {I
“35144. 1'” .
I w. .- .
= 1.55;” . ' .
vilsl c .
ill-
1
fi’
‘40....
x: Idfldn
1‘-
. I:
ll . .
'Vi' .BN‘L
f|
..< p
4 s
n ..t
y. . Os
5 0‘
n I r ‘ ‘ D l .l A. .‘A
I {’1‘ . v A
‘. on . - - {.anss.-«nn.9. Z . ... . .
I. IOuQ’C‘. I‘l‘ ptylvvl‘ . ; t. ~ o w : .-
z I vn\ul.ht TI... ‘1 .6 Q. 0 . A :-~«. c. .
. i “u ... 1-14.. . - . .. -, t: . . -. . -
, ...(7 ((10 . . .. . . o .. quuu. 1. A. «‘ n .v 2. .. ....-
. . v. v v o u .. u ..n A .
. ‘ u I ~ ~ tn V. A-IM w I“. IVI Nu..- t! . o .. . I I ~ ufw ‘
. . .. 03...... mt} . . . . Z a -. l?“
‘ . . . n tyil... s v ... 4‘ O I A
. 1%.-..qu alt . v .. _A .1. a v: _ 1.. -. .313. .
.v ooooo . ‘ . ..-a . .n n . n1| |Ill
I". ‘ I l's|. ‘Q‘l. V 0... I401
‘
7.32
V.
~
~
I\
3. .-.
\Jfifinnm.
:0 .v... ‘
.2 wmxflxmfifildwxzfl
.h» ‘u 1 3‘.<3~|§|I9.v‘0\0“oflnfil
J\ 0
c‘f‘l‘é‘l I .
v!
Q
[It QC.1I.|& stir u... i.
.u ....gkh n! .KHbohnlv...hl1. \:
'Jistl :94. J‘obl- lJIIn-h.."Iu‘ . I- cl Q 4- v, v
n 1 ll :00 v .“||I\0|‘W| {« .Jll‘vs} \v 0.. o 0.: . I. 3 .
- ain‘t ulnofbulIa-va .‘..‘1¢v\t|§1.y§ .. A .. n . Jo...
...»!!! \ J- n-v..‘¥l§.. I .laI‘I-zlo! cs; . . . s ‘ .... .m ... .1"
o 1““. O‘.1Il.|1n3 \c... no .. an . t. -1 > . Infill. ml.
I .u -v.‘ .3 \N‘qlal‘d C‘ll . o..vv~ c 1 . . L4 «v‘ v1.o.\ ....) h‘Vl 1 ..l “I“
t |.nOol N1. 3. I ‘.~ z . . k. in? .. v 4-¢AHN>PIHA..I“~|W¢M.
. A v«~.io\1,“r“ 6%.1 o v v ., .‘1 Allyn/\‘filmmn
I I '0. c‘tQ 9.1 c ‘ ..I.f..n.o ;. ll all
. . O. Y5 I utt\lti pl)..|"
. t..lr.l»‘.b:l 1.1"."U
-.lo‘.,00 .d‘f‘?
. 4!!le ‘1: »‘« ‘ 1.
..l 'iii‘o‘clz. I
l.\l..t"‘kf""\'lr ll'4
:0 ,‘a$\1“-I.l‘n I
31.. .17., AI|IJ$IQII 19
zil: lyiflhfi wv -vl t
A.'..0s'|n (. 1! (I‘.
[Itls‘gv ‘l I}.
. -‘vt‘... .9- .!
.gh‘lb, l- 0 O ‘
43:9.va100 x
, oilild ‘ 9
.v ‘ ‘llltt
- 1-
:c D .l‘l'h‘r1 ‘ ‘
‘ 31:33.- (It. 5 .
. - . ...- . ll'...x...q..l..?..k|v-
|||l..l.‘.. v I!
‘lit. 1
«IIII.J..Y { .I
- 1.3.... . ,
F-I
V?
,‘
Li
/""
i a
a l
...‘v
‘ ‘ .1 3:2" ...: l
figs..- ~gra nggt
5 ’ «bull? e
v.3;- . . .' . ,
‘.‘ . '. ’ .gv, “r _
"- "" 1’ .4. ‘ . .
& viii’ifl‘grfiaig‘jl
a w -
I‘-‘ "“x " - ‘ -4
«a. ..«, -, ,. ‘
t;-‘.>«_.%;:.--;:5n———
This is to certify that the
dissertation entitled
Gas Phase Transition
Metal Ion Chemistry
presented by
Sunkwei Huang
has been accepted towards fulfillment
of the requirements for
EA D degreein (l’fmifl/Vy
9%. flax/ire
/ Major professor
MSU is an Affirmative Action/Equal Opportunity Institution 042771
MSU
RETURNING MATERIALS:
Place in book drop to
remove this checkout from
LIBRARIES
1—!!— your record. FINES will
be Eharged if book is
returned after the date
stamped below.
~ :«m “r a” 2'31 am:
h ' a: [9 3;? a "in. “an?“ 4-: In
5a.:
*__
GAS PHASE TRANSITION METAL ION CHEMISTRY
By
Sunkwei Huang
A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Chemistry
1983
ABSTRACT
GAS PHASE TRANSITION METAL ION CHEMISTRY
By
Sunkwei Huang
Catalysis attracts tremendously broad interest and con-
stitutes an important research field for chemists. Although
there are a plethora of papers being published to probe the
catalytic behavior of surfaces and homogeneous organometallic
complexes, the behavior of a single metal center is usually
not well understood due to the complications of ligand
electronic effects (for the former) and solvent effects (for
the latter).
Recently, Muetterties e£_al. have reported studies on
metal clusters in an attempt to build a conceptual bridge
between molecular and solid state chemistry. However, all
of these fields have suffered from a lack of sufficient
thermochemical and structural information and are still
inevitably somewhat speculative in nature. Gas phase transi-
tion metal ion/molecule reactions help to provide answers to
the following questions:
l. How do metal centers react with molecules in the absence
of solvent molecules?
2. What factors control these reactions?
3. How can we apply these answers to obtain a better under-
standing and control of macroscopic synthetic systems?
In this dissertation, the chemistries of ions such as
Fe+, Cr+, and Ni+ (containing multiple numbers of C0
ligands) with multifunctional organic molecules are report-
ed. Mechanistic effects, ligand effects, the differences
in chemistries of these metal ions and their reaction
mechanisms with ethers and polyethers will be discussed.
Also, a unique "double insertion, double 8-H shift" mechan-
ism is proposed to explain the results of these metal ion
reactions with cyclic polyethers.
Since the reaction products we have observed in this
ICR study are mass spectrometric peaks, the assignment of
ion structures sometimes ambiguous, although some techniques
such as double resonance, CID, and the use of labelled com-
pounds can provide some useful information. Ab initio
calculations can provide much information such as energy
levels of various electronic states, orbital occupancies,
etc. The structure of CrCHz+ is given as an example to
elucidate the use of this technique.
ACKNOWLEDGEMENTS
I would like to thank my wife Chimiao Lieu for taking
care of my daughter Catherine and all the hosework that
enabled me to spend all my time doing research. Without her
encouragement, I would not be where I am today.
I also want to thank my parents and parents-in-law and
all the members in our two families for giving me confidence
and keeping the families in good shape so that I didn't need
to worry about my father or father-in-law's sickness.
Thank you, Dr. Allison. For four years, I have learned
very much from you. I really appreciate your sincere direc-
tion of my research work and also for giving me the courage
to be a good scientist and teacher. I will always remember
that you were the preceptor who put me in the right track
for my later career and who helped me in many aspects....
I also want to express my appreciation to everyone in
my group for sharing happiness, discussion and well being.
Acknowledgements must also be made to Dr. Harrison for
directing me to do the ab initio calculations.
ii
Page
LIST OF TABLES.. .................... . ................. vi
LIST OF FIGURES ....................................... viii
LIST OF SCHEMES . ....... . ............................. ix
A. THE TECHNIQUE - ION CYCLOTRON RESONANCE
SPECTROMETRY
1. Introduction ................................... 1
2. The ICR Experiment ............................. l
3. ICR Facilities at Michigan State University.... 8
4. Interpretation of Experimental Data ............ 10
B. THE CHEMISTRY
l. Purpose of Research In Gas Phase
Organometallic Chemistry ....................... l5
2. Historical Review of Transition Metal
Ion Chemistry.. ................ . ............... 20
3. Multifunctional Molecules: The
Macrocyclic Effect ............ . ................ 43
4. The Gas Phase Chemistry of Iron, Nickel
And Chromium Ions And Their CD Containing
Ions With Linear Ethers, Cyclic Ethers,
Cyclic Polyethers And Crown Ethers ............. 43
I. Fe(C0) T Reactions With Ethers And
Polyethers ................................. 45
II. Cr(C0) + Reactions With Ethers And
Polyet ers ................................. 95
III. Ni(CD) + Reactions With Ethers And
Polyet ers ................................. l23
IV. Comparison of Fe(C0)x+, Cr(C0)x+ And
TABLE OF CONTENTS
Ni(c0)x+ in Their Reactions with Ethers ....l46
iii
5. Trends in First Row Transition Metal Ions
In Gas Phase Reactions With Organic
Page
Molecules ...................................... 152
I. Reactions With Propane (C3H8) .............. l54
II. Reactions With Iodomethane (CH3I) .......... l59
III. Reactions With ISOpropylchloride (C3H7Cl).. 165
IV. Reactions Nith cis-2-Pentene (C5H1O) ....... l69
V. Reactions With l-Hexene (C6H12) ............ l75
VI. Reactions With 2-Pentanone (2-C5H100) ...... l80
VII. Reactions With sec-Butylamine
(S-BUNHZ, C4H11N) .......................... l86
VIII. Conclusions ................................ l94
C. AB INITIO CALCULATION
1. Introduction ................................... 196
I. The Importance of Ab Initio Calculations... 196
II. Review of Ab Initio Calculations of
Transition Metal Compounds ................. l98
III. Theory of Ab Initio Calculations ........... 200
2. Use of CrCH + As An Example of An Ab Initio
Calculation ................................... 2]]
3. Discussion ..................................... 22l
APPENDIX A. Schematic Diagram for Voltage Controls.... 228
APPENDIX B. Marginal Oscillator Setup in ICR
Experiment ................................ 238
APPENDIX C. The Relationship Between Magnetic
Field and Mass In The ICR Experiment ..... 239
APPENDIX D. Alternate CID Circuit for Conventional
ICR ....................................... 24o
iv
APPENDIX E. Calculation Of The Collision
Frequency For Co+ and \/0\/' ............ 243
APPENDIX F. Branching Ratios OF Fe+ Reactions
With VOV .............................. 2'45
LIST OF REFERENCES ..................................... 249
Table
10
ll
12
l3
T4
15
16
17
LIST OF TABLES
Page
Collisional Parameters as Functions of
Pressure for Co+ Reactions with Et20 ............ l2
Summary of Metal-Ligand Bond Dissociation
Energies ...... . ....... . ......... ... ..... . ....... 20
Heats of Formation and Avera e Bond Ener ies
of The Positive Ions From Fe%C0)5 and Ni?C0)4... 22
Fe(CO)5 Reactions with P- dioxane and
P- dioxane- "d8 ..... ......... ....... .. ............. 52
Ion/Molecule Reactions of Fe+ with l2-crown-4
and l5-crown-5 ....... . ...... . ....... . ........... 68
The Reactions of Fe(C0)x+ with Ethers ........... 9l
Cr(CO)5 Reactions with P-dioxane and P-dioxane-
d8' .............. . ........... ... ................ TOO
Cr+ Reactions with Cyclic Polyethers....... ..... ll2
Neutrals Lost In The Reactions of Cr(CO):
With Ethers. .......... . ......... . ........... 120
Ni(C0)4 Reactions with P-dioxane and P-dioxane-
d8 ..................................... . ........ 130
Ni+ Reactions With Cyclic Polyethers.. ....... ... l39
Neutrals Lost In The Reactions of Ni(CO)x+
With EtherSOOOOOO......OOCOOOOOOCOOOOCO ......... 145
Number of Reaction Products Observed Metal
Centers In Various States of Coordination ....... l47
Reactions with Propane (C3H8)... ................ l58
Reactions with Iodomethane (CH3I) ............... 164
Reactions With Iodomethane (CH3I) ............... l68
Reactions with cis-Z-Pentene (C5H10) ............ l73
vi
Table
l8
l9
20
21
Page
Reactions With l-Hexene (l-C6H12) .............. . l79
Reactions With 2-Pentanone (2-C5H100) .......... 184
Reactions With sec-Butylamine (S-BuNHz) ........ 192
Results of Ab Initio Calculations of CrCH; ..... 222
Figure
10
11
12
13
14
15
16
17
18
LIST OF FIGURES
Page
Block Diagram of ICR Spectrometer for
Single Resonance Experiment .................... 2
The ICR Cell .................... . .............. 3
Modulat1on of VTrapping (frequency325 Hz) ...... 6
ICR Mass Spectra of Cr(C0)5 + p-Dioxane With
Pressure Ratio l:l at Total Pressure 1.0 x
l0-5 torr... .......... . ............ . ........... 7
Double Resonance Spectrum of m/e 140 in
Figure 4 .................... . .................. 9
Energy Profile: CpNi+ and CH3CH0 .............. 27
Schematic Representation of The Potential Energy
Surface For the Reaction Of A Transition-Metal
Ion With An A1ky] Ha1ide00000000000000000 ...... 28
A Structure of Fe+-12-crown-4 After Fe+ Double
Inserts Into Two C-O Bonds of lZ-crown-4 ....... 79
X-Ray Structure of l5-crown-5 in l5-crown-5
CUBrz.0.0.0..........OOOOOOOOOOOOOOO. .......... 80
6 6
Energy Levels of B1 and A2 States of CrCH2+.. 224
Energy Levels of 4B1 and 6A2 States of CrCH2+.. 225
Trapped Ion Cell Circuitry-Pulsing Section ..... 232
Trapped Ion Circuitry-Drift Section.... ........ 233
Trapped Ion Circuitry-Trapping Section ......... 234
Electron Filament Bias ....... ..... ............. 235
Timings 0f Trapped Ion Cell Circuitry in
Figure 12. .......... .......... ...... . .......... 236
Experimental Setup of ICR ...................... 236
Schematic Circuit Diagram of CID Experiment
For ICR... ......... .. .......................... 242
LIST OF SCHEMES
Page
Scheme I .............................................. 25
Scheme II ............................................. 29
Scheme III ............................................ 33
Scheme IV ............................................. 34
Scheme V .............................................. 35
Scheme VI ............................................. 37
Scheme VII ......................... . .................. 38
Scheme VIII .............................. . ............ 39
Scheme IX ..................... . ....................... 40
Scheme X .............................................. 56
Scheme XI.... ......................................... 64
Scheme XII... ......... ........... ..................... 74
Scheme XIII ........................................ ... 76
Scheme XIV. ............................ . .............. 77
Scheme XV ...................... . ...................... 81
Scheme XVI ............................................ 81
Scheme XVII ........................................... 82
Scheme XVIII .......................................... 82
Scheme XIX. ................................... . ....... 85
Scheme XX ............................................. 85
Scheme XXI...... ...................................... 103
Scheme XXII..... ...................................... 108
ix
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
Scheme
THE TECHNIQUE
A. THE TECHNIQUE - ION CYCLOTRON RESONANCE SPECTROSCOPY
1. Introduction
Ion Cyclotron Resonance (ICR) Spectrometry is a
type of mass spectrometry especially designed for the study
of ion-molecule reactions in the gas phase. ICR experiments
provide a variety of chemical information such as acidity,
basicity, heats of information, bond strengths, proton
affinities and all types of information related to the
chemical reaction].
There are quite a few reviews available on this techni-
que1-9. ICR is based on the dynamics of charged particles
in electric and magnetic fieldsB. The detection system con-
sists of a marginal oscillator (M.0)10’11
12,13
or frequency sweep
detector with phase sensitive detection14. The basic
principle of ICR will be discussed.
2. The ICR Experiment
A block diagram of an ICR spectrometer is shown
in Figure 1. Figure 2 shows a detailed view of the three
section ICR cell which is, in our laboratory, 0.88" x 0.88"
x 6.25". The source is 2.00" long, analyzer region is
3.75" long and the collector is 0.50" long.
The cell is placed between the poles of an electromagnet.
It is housed in a stainless steel vacuum chamber which can
7
be evacuated to a pressure of 1.0 x 10' torr.
DRIFT a
TRAPPING
VOLTAGES
— mmnmu - cum
AMPLIFIER OSCILLATOR
— * .ETECTOR '
MODULATION
[REFERENCE l ammo. $3333
OSCILLATOR ENERGY 5 SUPPLY
— l , , EMISSION
' CURRENT - l I
PHASE- [‘ CONTROL I SWEEP - ‘
SENSITIVE xx RECORDER CONTROL
g DETECTOR _ g] l_2.
Figure 1.
Block Diagram Of ICR Spectrometer For
Single Resonance Experiment
The cell, situated in the magnetic field as shown in
Figure 2, consists of an ion source, analyzer and collector.
The electrons are emitted from a rhenium wire located outside
of the cell by operating the electron energy and emission
control as shown in Figure l. Emitted electrons follow
magnetic field lines, forming a collimated beam which crosses
the cell are collected by a collector located on the Opposite
side of the cell, and are detected as emission current. The
emission controller then regulates the emission current with
a feedback circuit to provide a constant emission current.
The electron current control can be used to adjust emission
current (u A) and electron energy (0 to -100 eV bias).
ISource IAmlyzer 'Conector
r.f from 1
marginal oscillatorfi
l
Drift Plate
Top Analyzer 1
Double Resonance
r. f Oscillator
_-L---- - --- .—
I
Drift Plate 1
Top Source ______1_
1
1
1
, I
I
I
1
1
I
Collector fl~ '
Source
Drift Plate ’ — Drift Plate
._ _ _ Botto A
Bottom Source In nalyzer
a
Figure 2.
The ICR Cell
Once ions are produced by electron impact, they will
'h
move circularly in a uniform magnetic field B with angular
frequency of
_ gfi -
w - mc (rad1ans/sec) (l)
where m is the mass of the ion and c is the speed of light.
Then, by applying a voltage E across the drivt plates
of the ICR, the ions will be "drifted" down to the analyzer
section where the M.0. detector is set to one specific fre-
quency (for instance 153 KHz) with a drift velocity (va) of
Ec
Vd = if (2)
One can then scan the magnetic field to bring ions hav-
ing different angular frequencies due to different masses
according to equation (1) to be at resonance with the mar-
ginal oscillator. The ions can be detected as the power
drop due to the power absorption from the LC resonant
circuit of the M.o.‘°"‘.
Data in this dissertation were obtained under normal
drift-mode conditions using trapping voltage modulation and
phase sensitive detection (See Appendix A for detailed hard-
ware description). The marginal oscillator detector is
based on the design of Warnick, Anders and Sharpl]. In this
technique, ions are produced in the source by an electron
beam with 70 eV electrons. In the presence of a magnetic
field, ions move in a circular orbit in the XY plane. TO
prevent ions from drifting to plates 2 and 4 in the Z direc-
tion, a trapping potential is applied to these plates (V
trap > 0 for positive ions, < O for negative ions), which
creates a potential well near the center of the cell to
trap the ions.
By applying a voltage difference E across the top and
bottom plates which are called drift plates (#l,3,5,6), the
ions experience an E x E force, drifting toward the analyzer
where the top and bottom plates (5, 6) form the capacitive
element of the tank circuit of the marginal oscillator
detector (See Appendix B). When ions move with a cyclotron
frequency equal to the natural frequency of the tank circuit,
they will absorb power from it and are detected. The S/N
(signal to noise ratio) Of the detector output is greatly
enhanced by use of a lock-in amplifier14. The reference
wave for the lock-in amplifier is provided by a function
generator, at a frequency of 25 Hz. To effectively improve
S/N, the signal output of the marginal oscillator must be
modulated at this frequency. This is done by using the same
25 Hz square wave to modulate the signal out of the cell,
which can be easily accomplished by modulating the voltage
on one trapping plate (e.g. plate 4) as shown in Figure 3.
This modulates the presence of ions to be detected in the
cell. If plate 2 is (+), and plate 4 is modulating, when
4 is (+), cations are trapped and when 4 is (-), all ions
are swept out of the cell. As a result, the rf level of the
marginal oscillator varies at the modulating frequency, and
forms a 25 Hz modulated signal output to be processed by the
lock-in amplifier.
Usually, the marginal oscillator is operated at a
frequency of g 153 KHz (wM.0) for convenience (see Appendix
C). The magnetic field is varied (0-18 KG) so the cyclotron
frequencies (we) of all ions present also vary. When w of
c
an ion matches wM 0 , power is absorbed and the ion is
detected.
+"T r
0 Ions Trapped In Cell
Ions Are Not In Cell
- VT .__
Figure 3
Modulation 0f VTrapping (frequency R 25 Hz)
Figure 4 shows typical ICR spectra of Cr(C0)6, p-dio-
xane and a 1:1 mixture (in pressure) of Cr(C0)6 and p-
dioxane.
5 torr, a sufficient
At high pressures, e.g. 1 x 10'
number of ion-neutral collisions occur to produce ion-mole-
cule reaction products. To umambiguously identify reaction
sequences, an ion cyclotron double resonance (CDR) experi-
ment is performed7’15.
Consider the reaction sequence:
+
A + N -————+> 3* + M (3)
B+ + N —————+ c+ + L (4)
where B+ can be formed from many different precursors (A?)
in addition to A+. Also, A+ can be a precursor forming
different product ions (ET) in addition to 3*. If the magne-
tic field is set to monitor B+ (B+ is in resonance with the
M.0.) without scanning the magnetic field, another radio
frequency signal can be introduced into the ICR cell so that
A+ can gain power and be ejected. When this occurs, the
N3 n13 x o...” gmmonm
Horas 3 a . H Scum genome 5;
oedxofivm +385 .8 $83 and. 6H .3 8.3a
8. 09 CV. \
. Ow. . .00. . . .O.m 00 0? ON
‘ I
.
n
+805
52.8% Res. 0.85
o
m U 528QO $2: $96.65 w o
a +0 1.0.1
{1...
1‘11 - b in n h
.. is. 1 2
Eda 3% are
\
m
ocOxofiia + «08.6 Lo 0535 E
hem +6 e no
intensity of 8+ decreases. By scanning the frequency we
applied to the cell, we can eject all possible precursors
A: forming B+ to get this "double resonance" response.
Remember
at constant B, wa = constant, therefore
It is then predictable at which frequency ratio the inten-
sity of B+ will decrease to indicate a unique reactant:
product pair, since
3
8| 8
21>
flea
and wB = 153 KHz
8
Figure 5 shows a typical double resonance spectrum.
In earlier ICDR experimentsg, the second oscillator was
introduced in the analyzer region (on plates #5,6). This can
lead to serious interference problems between this oscillator
and the marginal oscillator which is also coupled to these
plates. If applied in the source region, stronger second
oscillating fields can be applied without affecting the mar-
ginal oscillator. Appendix 0 describes other alternatives.
3. ICR Facilities at Michigan State University
The ICR used in the experiments which will be discussed
was built at MSU.
.: gm 5
03 o}. no 5.5003 momma—omen 33.58 .m E
on: was
m
w m .85
M. be noun.
weaves fl
. SR. w. +8
0 «HR.
1
10
The filament emission controller and plate voltage con-
troller for the ICR cell were designed and constructed by
Dr. M. Raab and Dr. J. Allison in the Department of
Chemistry at MSU. The marginal Oscillator detector is based
on the design of Warnick, Anders and Sharp]]. A Wavetek
model 144 sweep generator is used as the secondary oscilla-
tor in ICDR experiments. The ICR cell is housed in a stain-
less steel vacuum system and is situated between the pole-
caps of a Varian 12" electromagnet (1.5" gap). The electro-
magnet is controlled by a Varian V-7800, 13 KW power supply
and Fiedial Mark I magnetic field regulator.
The instrument is pumped by a 4" diffusion pump with a
liquid nitrogen cold trap, and an Ultek 20 l/s ion pump
controlled by an Ultek 150 mA ion pump controller made by
Perkin-Elmer. The lock-in amplifier used to enhance S/N
is model 128 A, (0.5 Hz - 1000 KHz) from EG&G Princeton
Applied Research.
Samples are admitted from a dual inlet (separately
pumped by a 2" diffusion pump and liquid nitrogen cold trap)
by Varian 951-5106 precision leak valves. Approximate
pressures are measured using a Veeco RG 1000 ionization
guage.
4. Interpretation of Experimental Data
Data were acquired in the following manner. High and
5 6
1<3w pressure (1 x 10' torr is. 1 x 10' torr) spectra of
each compound were taken, and ion-molecule reaction products
11
in either "metal carbonyl" or "organic molecule only"
experiments were determined. The "extra“ peaks observed in
the mixture were taken as ion-molecule reaction products
formed in 1:1 or 1:2 mixtures (in pressure) of metal car-
bonyl to organic, at a total pressure of l x 10.5 torr as
shown in Figure 4. All the product ions were then studied
to unambiguously identify the precursor ions by the ICDR
technique. The following points should be considered in
the interpretation of ICR experiments.
a. Ions are produced with thermal velocities, since
the ions are trapped by small trapping voltages. Fast ions
will not be trapped and thus do not react with neutral
molecules in these experiments.
The application of drift voltages only increases
ion velocities by 10%. This contrasts with conventional
mass spectrometers in which ion extraction from ion sources
requires strong electric fields, accelerating ions to some
keV energies.
b. Ions have relatively long residence times in the
analyzer region. Residence times of milliseconds or more
enable the observation of gas phase ion-molecule reactions.
This contrasts with microsecond intervals in conventional
mass spectrometers.
c. Low pressure experiments yield mass spectra with
fragments experiencing no collisions. Table 1 shows the
reVlationship between collisional parameters and pressure.
12
Appendix E shows how to calculate the Langevin collision
rate, collision frequency and time between collisions.
d. At higher pressure, gas phase ion-molecule reac-
tions are observed. From these, (1) kinetic studies can
be performed; (2) double resonance can be used to identify
reaction sequences; (3) branching ratios can be calculated.
(Appendix F)
e. Various thermochemical quantities can be deduced
from observed ion-molecule reactions. Upper and lower
limits on heats of formation of charged and neutral species,
and limits on bond strengths can be deduced. It is assumed
that processes which are observed must be exothermic or
thermoneutral.
Table 1. Collisional Parameters As Functions
of Pressure For Co+ Reaction With Et90
Time Between Collisions
Pressure Number Density Neutral-Neutrala Ion-Moleculeb
torr I molecules/cm3 sec. sec.
10‘2 3.24 x 1014 1.68 x 10'5 2.67 x 10'6
10'3 3.24 x 1013 1.68 x 10‘4 2.67 x 10'5
10'4 3.24 x 1012 1.68 x 10'3 2.67 x 10‘4
10'5 3.24 x 10H 1.68 x 10'2 2.67 x 10'3
10"6 3.24 x 1010 1.68 x 10“ 2.67 x 10'2
10'7 3.24 x 109 1.68 2.67 x 10'1
——¥
a. Time between collisions]: ZR] /mEt / 4/?62p
b. d(Et 0) = 8.78 x 10"24 cm3, K
culeg, see Appendix E.
L 1.154 x 10"9 cm3/mole-
13
For example,
Fe+ + \/0\/ —————+ Fe(C2H60)+ + C2H4
. _ 18
s1nce C2H50C2H5-—————+ CZHSOH + CZH4 AH - 16.8 kcal/mole
+ .
therefore D[Fe - (o ———-c2H5)] > 16.8 kcal/mole,
H
for the overall reaction to be exothermic.
f. Since most ion-molecule reactions occur with a rate
that is within an order of magnitude of their pre exponential
factor, m l x 10'10 cm3 molecule-1 sec'], it is assumed that
they occur with essentially no activation energy. This
follows from the Arrhenius equation
k(T) = A e(-Ea/RT)
-1 10
where A is on the order of 10101 mol'lsec , namely 10'
cmz‘lmoleculeqsec-1 for bimolecular gas phase reactions.
9. In the absence of solvent, all of the energy of
Observed reactions must be accounted for by the products.
Since we are assuming gas phase reactions are exothermic,
the product ions formed must be more stable with respect to
the reactants. In contrast, the energy can be dissipated
through solvent molecules surrounding the product ions in
solution. Because of this, there is rarely a single product
In an ion molecule reaction (P+ + N -—— I+*). Bonds are
+11:
usually broken and 1+* fragments. At higher pressures, I
can be stabilized by collisions, which is then called an
14
u . . n + +* + *
addltlon product (P + 2N ——> I + N ——91 + N).
h. For some endothermic reactions, the energy barrier
can be overcome by raising the kinetic energy of a
reactant‘g.
THE CHEMISTRY
B. THE CHEMISTRY
1. Purpose of Research In Gas Phase Organometallic
Chemistry.
The work which will be described in this text deals
with the gas phase chemistry of a number Of metal- and
metal-containing ions with various neutral organic sub-
strates. The purposes of studying gas phase organometallic
chemistry include the following:
a. Gas phase results are used in kinetic theorieszo'2].
b. Gas phase reactions can be used to model condensed
phase chemistry.
Catalysis draws tremendously broad interests and consti-
tutes an important research field for chemists. Although
there are a plethora of papers being published to probe the
catalytic behavior observed in surface chemistry and homo-
geneous organometallic complex523, it is not well understood
what occurs at a metal center due to the complication of
ligand electronic effects31 for the former and solvent
effects for the latter24. 0n the other hand, researchers
25 studies claimed that metal
26,27
in heterogeneous metal catalysis
clusters play an important role in catalysts , which
28’29 and can be used as an
interface between molecular and solid state chemistry27.
agrees with ab initio studies
Moreover metal-carbon double and triple bonds are frequently
30
proposed in mechanisms for such reactions However, all
these fields have suffered from a lack Of sufficient
15
16
thermochemical and structural information and are still
inevitably somewhat speculative in nature. Building blocks
of this nature allows the facile correlation of vast amounts
of chemical data. Gas phase metal-ion/molecule reactions
can serve this purpose. Other processes in organometallic
chemistry whose gas phase ionic chemistry parallels can be
readily studied are:
1) Hot atom chemistry32
2) Atmospheric chemistry33
3) Matrix isolation studiesBz’34
4) Metal catalysis chemistry
5) Organometallic chemistry and the solution chemistry
of metal complexes (e.g. ligand substitution
reactions)
c. The study of reactivity trends
Gas phase ion-molecule reactions allow the investigator
to study chemical dynamics and chemical events in the absence
of solvent complications. At Operating pressure (10'5 m 10'6
torr) used in mass spectrometric techniques such as ICR and
35 experiments, single collision events (the basic
ion beam
unit of chemical reactions) can be studied, namely two
species come together to react and separate as products. In
this work, reactants are metal ions (M+),_metal-containing
ions (ML:) and organic neutrals (A). By varying M, L, and
the structure of A, we can study how metal ions react, and
what factors affect their reactivity, and thereby
17
characterize the mechanisms of such reactions.
d. Gas phase chemistry is suggestive of condensed
phase processes
The results from gas phase chemical studies are often
suggestive of condensed phase experiments. Recently,
36 have demonstrated an intriguing
Kametani and Fukumoto
technique called "retro mass spectral synthesis". Fragmen-
tation of a compound on electron impact in a mass spectro-
meter is frequently very similar to chemical degradation
reactions39. For example, cyclohexene decomposes to give
butadiene and ethylene. This compound fragments on electron
impact to give C4H6+ and C2H4. Cyclohexene can also be
synthesized by butadiene and ethylene by a Diels-Alder reac-
tion. Mass spectral fragmentations frequently, therefore,
parallel synthetic pathways as well as degradation pathways.
Kametani and Fukumoto have used this observation to develop
new synthetic pathways for a number of natural product536.
There are many ion-molecule reactions in the gas phase
37’38. Similarly, the nature
having condensed phase analogs
of the bonding of organic molecules on metals can be under-
stood from this work, which can be used to examine what has
been proposed, for example, in methathesisBo.
e. Thermodynamics
Although reaction rates can theoretically provide an
understanding of the effects of structure on reactivity in
organometallic reactions, most observed reactions can be
18
understood in terms of thermodynamics. In other words, bond-
strengths can be reflected in the reactivity as a driving
force to initiate the reaction. Limits on heats of forma-
tion of various complexes, and limits on bondstrengths can
also be obtained. Limits on D(ML:-A) where A is an alkyl
group, hydrogen atom, halogen atom, oxygen atom, alkoxy
group, or various n and n-donor bases can also be obtained.
For example, consider the reactions:
Cr(C0)2+ + \V,O\,r -——4 Cr(Et20)+ + 2C0
Cr(c0)3+ + \v/O\/r —4P4 Cr(Et20)+ + 3C0
we found that D(Cr+-2C0's) = 66.2 kcal/mole and D(Cr+-3C0's)
= 87.7 kcal/mole, therefore, we conclude that 66.2 kcal/mole
< D(Cr+-Et20) < 87.7 kcal/mole.
f. The study of mechanisms of organometallic reactions
Since organic molecules contain alkyl groups, the
interaction of a metal center and an alkyl group is very
important. The following mechanisms are found useful to
explain reactions observed in the gas phase:
1. B-Elimination. This process involves the shift of
a B-H atom from the alkyl group onto the metal4o'42.
+ + CH2
M-CHz-CHR == HM--N
R\\‘JH CHR
43
The relative stabilities of dialkyl compounds ,
CH m pHCH2 m (CH3)3CCH2 >> N-C3H7,
3
l9
N-C4H9 > CZHS > t-C4H9 > i-C3H7, shows the parallel to
carbonium ion stabilities.
2. Reductive Elimination. This is a pathway for metal-
40,44
carbon cleavage with B-atom migration e.g.
D
1 +
+1i/|--C6H5 ————-> MH + CGHSD
H
+ +
(CH3)3AuPPh3-—-——-—-9 CH3AuPPh3 + C2H6'
Other mechanisms proposed in catalytic studiesao’45
have not been used in gas phase organometallic reactions
yet.
The organometallic literature since 1960 has shown that
the thermodynamics of bonding in organometallic compounds
is of interest, however relatively few actual bondstrengths
have been measured. This can be done in gas phase reactions
as described in b.
Table 2 summarizes the metal-ligand dissociation
energies obtained from ion beam studies by J.L. Beauchamp,
which will be of great use in interpretation of results in
this thesis.
9. Li+ use has been demonstrated as a mass chemical
197
reagent . It gives simple spectra, easy to inter-
pret.
20
Table 2. Summary of Metal-Ligand Bond
Dissociation Energiesa (kcal/mol)
_B_ Cr+-R Mn+-R Fe+-R Co+-R Ni+-R
H 35:4 53:3 58:5 52:4 43:2
CH3 37:7 > 48 68:4 61:4 48:5
CH2 65:7 94:7 96:5 85:7 86:6
0 77:5 57:3 68:3 65:3 45:4
a. Data were taken from ion bean studies by J.L. Beau-
champ 94-97
2. Historical Review Of Transition Metal Ion Chemistry
From its birth through the mid 19605, most of the appli-
cations of mass spectrometry have involved organic compounds.
In the past two decades, however, applications have been
extended into inorganic chemistry. Mass spectrometry is now
commonly used to provide molecular weights and formulae of
inorganic and organometallic compounds. Many thermodynamic
data were Obtained from this field. There are quite a few
specific areas of interest in the general area of gas phase
metal ion reactions:
a. Fragmentation Studies. (Unimolecular reactions)
Early work was devoted to investigations of the ionization
52, their fragmentation
patterns, and how these are affected by ligand553.
potentials of organometallic compounds
Other
21
extensive studies of fragmentation patterns in main group
organometallicsso’54 55
50
, metal carbonyls , polynuclear metal
56,57
carbonyls have been re-
and coordination compounds
ported. A typical fragmentation following electron impact
is illustrated below58: +
~———————+ Cr + :C(0CH3)CH3
OCH
Cr —— C -? Cr--CCH
CH
'———————+ Cr---OCH + H2C=CH.
3
Most of this early work was done on volatile organometallic
59 252
compounds, but, field desorption Cf
6O
, and recently
61 62
plasma desorption , laser desorption and
63
, secondary ion
fast atom bombardment techniques have made possible mass
spectrometric studies on nonvolatile compounds and even
ionic substances such as saltssg.
In addition, chemical ionization mass spectrometry (CIMS)
studies have also been performed with organometallic com-
pounds. Examples are CIMS (using CH4) Of sandwich com-
64,65 65
, metal hexcarbonyls , cyclopentadienyl metal
, and arene metal carbony1s64.
pounds
ha1ides65
b. Photochemical Studies. The photodecomposition path-
ways of several metal carbonyl anions66
67
such Ni(CO)3' and
CO(C0)4' and cations have yielded useful thermochemical
data on metal carbonyl fragments, such as heats of formation
and average bond energies as shown in Table III.
Table 3 . Heats Of Formation 699.319.1369 80an Emerges Of The
Positive ions From Fe(CO)5 And Ni(30)4_
photoionization AH; (2v) Average bond
threshold(ev) energies
Ee(00)5 D[Fe-(UO)§]=1.2510.0Bev
Re(oo); 7.08: 0.01 0.37: 0.02 D[Fe-(tio)5]=1.23:0.o3ev
Fe(OO)Z 8.77:0.1 23110.1
Fe(CO); 9.8710.1 4.55:0.1
Fe(00); 10.6810.1 16.51: 0.1
Fe 00* 11.53101 8.51101
Ice“ 14.03101 12.35 10.1
111(00)“ DENi-(m)4]=l.5310.03ev
Ni(00); 83210.01 2.07:0.02 D[N1-(m)4]=1.3610.03ev
111(00); 8.77: 0.02 3.58:0.02
Ni(00): 10.10: 0.1 6.14: 0.1
NICO+ 11.65: 0.1 8.841 0.1
NI+ 13.75: 0.1 12.09: 0.1
4.
a. The assumed processes are 111(00)x + hv——-)(1:(Co)x_n]+noo
22
23
c. Gas Phase Ion-Molecule Reaction Chemistry in
Organometallic Systems.
(i) Metal and Metal Containing Ion-Molecule
Reaction with Organometallic Compounds. The formation of
ions of the type M2(C0): from ion-molecule reactions in
metal carbonyls have been reported for the group VI hexa-
55’68. Fe(C0)567"71 and the negative ion chemistry
72
carbonyls
of Ni, Fe and Cr carbonyls (with accompanying loss of 1
or 2 C0 groups) as follows:
———4 Fe2(00)7+ + co
Fe(c0); + Fe(C0)5 ——+—
+
‘———+ Fe2(CO)6 + CO
N1(CO)§ + N1'(CO)4 --—-—+ N12(CO)g + CO
Fe(C0)3 + Fe(CO)5‘———————+ Fe2(C0)g + 2C0
Cr(CO):1 + Cr((:0)6 ————-> Cr2(CO);3 + 200
Cr(C0)3 + Cr(C0)6 ———————+ Cr(C0)g + 3C0
Further reactions to produce tri-iron carbonyl ions
(Fe3(C0);), and the formation of ions up to Fe4(C0);2 were
also observed. From these studies and ligand substitution
70 with a series of n- and n-donor bases, proton
70,72
studies
of metal carbonyls and bond energies70
affinities (PA)
can also be obtained. For example, P.A. (Fe(C0)5) = 204 i
3 kcal/mole; D [H-Fe(c0);] = 23 : 1o kcal/mole.
24
73
Also, the gas phase ion chemistry of ferrocene and
74
nickelocene have been reported. Predominant features
are charge transfer processes and the formation of a bimetal-
. +
11c complex M2(C5H5)3, e.g.
Fe+ + Fe(C H ) ——> Fe + Fe(C H )"
5 5 2 5 5 2
+ +
+
FeCSH + Fe(C5H
+
5 5)2 "“‘* Fe2(CSH5)3
Also, "triple decker sandwich" complexes were reported
75,76
for these compounds From these studies, the proton
affinity of Ni(C5H5)2 was reported to be 218.9 1 1.0 kcal/
mole74.
MUller also reported the gas phase ion chemistry of
other sandwich compounds such as dibenzene chromium68°77.
He has also studied the chemistry of n- and n-donor bases
with ions formed from electron impact on C5H5CrC6 6’
78
C H6, CSHSVC7H7, and CSHSCY‘C7H7 .
H MnC
5 5 6
In addition, the ion-molecule reactions in
88
C5H5V(C0)47, C5H5Mn(C0)379 and C5H5Cr(C0)2N0 have also
been studied by MUller. The ionic chemistry in mixtures of
C5H5Mn(C0)3 with PF3, AsF3, SbF3 and SF4 was reported by
MUller and Fender179. The ion-molecule chemistry of
CpNiN0(Cp = C5H5) and the chemistry of its ions with n- and
n-donor bases were also reported80’81. From this study, a
4.
series of CpNi-B relative bondstrengths were determined81.
25
(ii) Metal and Metal Containing Ion-Molecule
Reactions with Organic Neutral Compounds. J. MUller has
studied the chemistry of metal-containing ions with hydro-
77,80
carbons , in which H2 loss with simultaneous formation
of a new complex is observed:
CpNiN0+ +[::)H -—————+. CerE:::1 + N0 + H2
-H +
CpCO+ > H2; 2 CpCo[:::]
CH
CpNi+ :[::::>-————9 CpNi---“1 + CSH10
CH
CpNiNO + [::::> —————9 CpNi- 1£::::>¢—————et CpNi ----- m:
H2 loss processes parallel the normal electron impact frag-
mentation mode for similar organometallic compound558, e.g.
wk
1 ‘*
‘ -2co
(CO)3 Fe ———————9 (c0) Fe+ -———————> Fe+C0
3 1
-2e n: -H2 !
J.L. Beauchamp has studied the gas phase chemistry of
82 of
CpNi+ with alkyl ha1ides and found that the reduction
alkyl halides (RX) by CpNi+ to olefins and HX is similar to
that observed for Li+83.
26
H x
CpNi+ + H—e CpNi+---)( + HX
Decarbonylation of aldehydes was also observed82:
___, CpNiC0+ + RH
CpNi+ + RCHO ——1-
,____, CpNiRH+ + C0
The energy profile for this process is shown in Fig. 6.
Deoxygenation of acetone by metal-containing ions have
also been reported78:
+
_——-+ C6H6VO + C3H6
+
C6H6V + 0 =’ C(CH2)2——'
' +
-——3 V0(CH3)2 + C6H6
Recently, J. Allison and D.P. Ridge have studied the
transition metal ions of iron, cobalt, and nickel with polar
organic molecules and found that these transition metal ions
84,85
always insert into a polar bond and is followed by a
B-H atom shift (Scheme I):
Scheme I.
K CD D ‘
Fe+ + CD CH I ———> CD CH -Fe+-I,-—i— ll 2"1J4 —-——> CH CD Fe++ DI
3 2 3 2 CH2 2 2
11
+ CHD 1' + +
Fe(CHDC02) + H1 4— H ---Fe-—-I : CHZDCDz-Fe-I
CD2
( J
27
2on5 Ba 28 u unmade 885 .m 993...
+
m ... N528
:0 U/ a... .9. mo
38 z m/ a / m
I n a 1.0.126
5 1.:8 m 2 o o
a
38 + $26
--~‘_----.o
30(an n: 1 N x4 casinos. b: H a
02095 + Mano
28
This process is proposed to follow the reaction dia-
gram of intermediates shown in Figure 7.
+ MX + R+
M +RX +
Mt u\ + HX
X-M—< x4141- 1
Figure 7. Schematic Representation of The Potential Energy
Surface For The Reaction Of a Transition-Metal
Ion With An Alkyl Halide.
However, this mechanism does not Operate for amines86:
Co+ + >th —--x--» Co+lk + NH3
+
.____, CO(C3H7N ) + H2
.____, CO(C2H5N )++ CH4
+ +
+
—-—+ Co(C2H5N ) + H2
.——-a CO(CH3N )++ CH4
29
Ion-molecule reactions of iron with ketones and ethers87
follow a similar mechanism to that of in Scheme I and is
shown in Scheme II.
Scheme II.
+
0 /Fe
r T + H 11
e ~>
/’ ‘\ z’ \\
CD3 0.020113 01120113
0 o
§>
[003— Fe— C-cnzcaa] [CD/c—Fe— (1:2083]
3
B-anqa
shift [Ifluft
+ -
obj—ne—OO CH ”CD 011%“ -00 °§
C0208 GD/C‘ :6 —- ll2
3 3 C32
13-8
shift 4102;7/ 1413301“)
0 CD
CD '1' ‘1' 2
1 lL—-Fe Fb'—-"
11- Fe? 00 flan—,GHZ‘Fe Loo /c\ 0112
0112:. 0112 0112 0113 II
CD CD CD30D= 082
3 21 EC
:gzoo -li£L—-—9 D— Fe- -CD fldELFe-w
I Shift |
H H ‘/CD
30
Scheme II (cont’d)
.1.
g F 11/“ \1 K (11
C C
+ 0112/ \CD2 CD2/ \CD2 0112/ \Fe:
Fe + 1 I -—-—+ | 1 -——9 ./ CD
CHZN /CH2 CHZ\ /% CH2 /
032 L CH2 J L \Cfiz/ CI"2
13-8
shift
0 ‘ ,
1 1 1 I a
/ \ / \ +/
CD + CD
2 Fe - 2
l I an??? / \>;?CD2
Cflz‘culciiycnz ““2 /CH
K J 1 K CHZ J
1-.....
/F°+\ 1°
CD +
CKCH_CH¢ 2 /Fe
finz
.-cn
. 2
“awrfl/
I“ '3 1 Sift
shift
IE>~E;t-o\€] 1:)F—EJL-j1~ ]
C3H70 -03H6 l -03H8
31
Scheme II (cont’d)
+
{e
F3 + 0 ——) o, 0 +
0 0 “1 0
13.11
() ¥ shift
\Fe+
// -H
.fi-H H
shift shif I
0 1+ ‘ (I T
+
H-—-Fe-f> [:;:Fe
: ‘nflf
1-82. 1
H0 H
I" \ /
38* /Fe;/
.| .
- 0
JHFe-F 1'12
+
F8
Hence, Fe+ inserts into a carbonyl-carbon bond or a C-O
bond in its reaction with ketones and ethers respectively
and form a metalcyclic intermediate in its reaction with
cyclic ketones and ethers. This is followed by a B-H shift,
sfinilar to that in Scheme I.
32
The study of the gas phase chemistry of titanium ions
88 89
with haloalkanes , and alkenes has also been reported,
e.g.
Ti+ + RX 1r————9'Tix+ + R x = Cl, Br, I
_————a R'+ + TiCl R = CC13, CFClz,
MeZCH, CHC12
. + + .
T1Cl3 + RC1 -—————+ R + 1101
. + . . + .
T1C13 + CH3T1C13-—————+ CH3T1C12 + T1Cl4
. + . + .
CH3T1C12 + C2H4 —-———+ C3H5T1C12 + 11014
. + 85% . +
CH3T1Cl2 + C204 -—————9 C3H5T1C12 + H0
CH T'C1+ + c H —————4 C H 1°c1+ + H
3 ‘ 2 3 6 4 7 ‘ 2 2
T'C1+ + C H -—-———e H 1'C1+ + HC1
‘ 3 3 6 C3 5 I 2
Note that titanium has only four valance sites. TiClg
reacts with C3H6 by eliminating HC1 to give C3H5Ti+C12, pro-
bably being a resonance stabilized allyl-TiCl2 cation (I) as
shown below:
33
CH31i+C12 reacts with C204, 85% of the reaction pro-
ceeds by HD elimination. This suggests a mechanism
(Scheme III), in which, after association with the Ti cation
center, the C204 inserts into Ti-C bond of CH3TiC1; as
would occur in the polymerization of ethylene.
Scheme III.
C1 C1 CH
\ .+ In,“ .+ 3
/T1 — CH3 + C204 ——> (T1 002 ————>
c1 c1 /
C02
T H _H
Cl CH3 ‘3‘ \C
'I + \ ”1,".
. ...-.\\
/ \ /C02 ———> 111:, C—0 + H0
Cl CD2 c1] D/C/
”'D
The resulting species undergoes unimolecular decomposi-
tion because it is unable to dispose of its excess internal
energy in this gas phase bimolecular process. Consequently,
1,2-elimination of HD across the B-and y-carbons gives the
allyl-TiClz+ cation (I) and does not react further. To be
effective, mediation of reactivity of CH3TiC12+ by solvent,
the nearby presence of a counterion and another molecule
such as AlMe3 as in Ziegler-Natta catalysts are apparently
necessary.
J. Allison and D.P. Ridge have also reported the chemi-
stry of titanium containing ions with alkenes90 and oxygen-
91
containing organic compounds and reported the following
34
trends:
1. Ti+, TiCl+ eliminate nH2 with olefins (n 3 l).
2. TiClE, TiClg eliminate HC1 with olefins containing
a carbon chain of g 5 carbon atoms.
3. TiC1+, TiCl; eliminate small olefins from olefins
containing a carbon chain of 3 6 carbon atoms.
Scheme IV shows these processes.
Scheme IV.
-2H2
Ti+ + ‘1[:f\-———9 Ti+---:1[:\1 >1111:::>
\* ‘2”
L + 2 +/
TiCl 1- -———> C1Ti --- ClTi
1 , \
NV
\Jv
('5
—l
—_{
u—lo
Q
. +
T1Cl3 +
. +
T1C13 +
//C C H
H @/' O //
Cl ’I L C\\ CH CH
‘71i* H //// H 2 1 2
Cl‘ 1" \\C -———5 Ti+ + c H
H .
H/ \C/ / 4 8
H CH3 C1 C1
35
Scheme IV (cont'd)
H @ @ H
\x //
C———-C -~_
H/i@ @\\\\ //H 4//I:T\
‘“1i+ H C1 | C1
// H
C1 H
H
\\c<::;/’
/
CH3
For oxygen-containing organic compounds, the following
trends were reported:
1. Ti+, TiCl+ deoxygenate aldehydes and ketones.
2. 1101*, TiCl; eliminate HC1 with small aldehydes
and ketones.
3. TiC12+, TiCl; eliminate small olefins from aldehydes
and ketones containing 3 4-carbon chains.
Scheme V shows reactions representative of these processes:
Scheme V.
+ 0 .+
Ti + /fl\¢,\\ —————+ T1 0 + CSH10
0 +
Tic1+ + ,JLV,,\‘-—————a TiClO + C5H10
36
Scheme V (cont'd)
/ \
c1 C1
\+ /
+- " //Ti..Q + 0\
TiCl + RCH CR' -———> C1 ——-> C1 1i——,=C- R + HC1
3 2 5M 2
1 1' 61%
\q/ \R'
RC/ \
H 1
LR
Cl C1 C1
. + ” l.+ 1 + I +
T1C13 + RCHZCHZCR'-—9 CI-Tl -C1-—-+ CI-Tl -C1-—-fi CI-Tl -C1
0"] H orH 06/"+
/H\+ R RI/_‘|_\J\R R'+/\R
H H
(In
————> C1-Ti+-Cl + /\R
0
AK
R' H
Note that HC1 and Olefin eliminations proceed through
a 6-membered ring intermediate with an a-hydrogen needed for
the former process and a B-hydrogen for the latter.
Ti+ reactions with alkanes have also been studied92’93.
—-—->1i+/1L + H2
A _.
e.g.
37
.+~\\
+ -———>C1T1.u>/ + C2H6 + H2
TiCl + /~\/“\,/ ————-
. +
.L__9 T1C1C6H10 + 2H2
+ .t.v>\
T1 +/\/ ————————9 T1; + 2H2
In contrast, transition metal ion chemistries with
alkanesga'96 and alkenes97 by ion beam studies were reported
by J.L. Beauchamp as shown in Scheme VI and VII.
Scheme VI.
H C0$:~$’:rt\H)\+ /CO +)1\——H—-—) Co +—)l\
\\N ‘:§\u\\c W+-/J______9 C°+'—/l1
CH 3-Co+ 1FNCH3
SN
+/ +
Co\\ ————9 Co C2H4 + C2H4
H
CO +
\\\\S H‘\\ +
/
38
Scheme VII.
C3H6 + CoC2H4+‘e—————r— ”.cot_lL
+
C2H4 + C0C3H6 €—————‘
Hence, Co+ can insert into either a C-H bond or a C-C
bond followed by a B-H, B-methyl, allylic H, or allylic
alkyl shift. Note that l-butene can be isomerized to cis-
2-butene and vice versa.
Copper, on the other hand, does not insert into carbon-
polar bonds in alkyl halidesgs, ketones and esters99
(similar to Li+ and Na+85) as is shown in Schemes VIII and
IX.
39
Scheme VIII.
Cu: T
+ C‘ H in“ H 9‘
cu + H ——» ———+ w
k / \J
/ / \ W"
+ i
+ ---, + \__/ \_=_/
CuCl >7< HCl Cu + /_\ HCl +/_\
Halide transfer dehydrochlorination
In order to determine if the two groups in an organic
molecule behave as one "new group", a study of cobalt with
bifunctional organic molecules is being performed in our
100
laboratory Some important results are summarized below:
l. In the case of adjacent functional groups, it can-
not be assumed that products indicative of each group will
be observed in reactions with gaseous metal ions.
2. All of the metal-containing ions derived from
Co(CO)3NO exhibit a rich chemistry with bifunctional mole-
cules. Presumably, in cases where two groups can bond to
the metal, much more energy is released in the intermediate
complex than in the casetrfmonofunctional molecules.
3. In the case of allyl amine, we do see strong evi-
-ence for insertion of Co+ into the C-N bond in contrast
with regular amine586, presumably driven by the strong inter-
action with allyl group.
40
Scheme IX.
acnzccnzcnzn'
91+ + ('m“
‘\ -—-—-% OH H. /'OH
31w )cg-jcnz“—i ll; 53.
RCHZ/ \cncnzav R012 §CH RCH/1\cacnz
:‘Tgcno flui-
Cu, I on
cmz / \
ROI-I 'v CHCHZ
\HJ
0H /
x... 1 ..
...
(hf-- -|| Cum-H2
CHR (:32
Rallzfion'
in“
E\ + ‘——‘—-> u c: R’ c“ E!
__ _ ____9 _
nou’ \oa' g 3 g
H R/ \H l R/ \H
cJ—on'
41
H
Co+ + 7/i\\v/’ +
//z ‘——5 Co -—-NH2
/ \"ii
—Co+——NH2 fi""c°““””2
H
2
—NH2
-NH3
Fe(CH20)(C2H60)
Fe(C H o) ——
2 2 .863
+
--# Fe(CH30)(CzH60)
Fe+ + \\J/O\\,/~—T9§;+ Fe(C2H4)+ + C2H60
.24, +
r Fe(C2H4O) + C2H6
.41 +
‘-—+ Fe(C2H5OH) + C2H4
eiflie Fe(C4H80)+ + H
2
+
Fe + o o -——9 Fe(C H 0)+ + c H o
CH§/’\ A R ?\\CH3 3 6 5 12 3
Basically, all these reaction products can be explained
in terms of Scheme X, i.e. Fe+ inserts into a C—O bond,
56
followed by a B-H shift.
Scheme X.
0
0 + 0 +
+ / \ B-H Fe “—11
Fe + CJ/ \\CHI_—9.CH’/’Fe CH2 shift CH/'1 CH2
3 3 3 // 3H
H8
+
Fe(CHZO) + CH4
S1m11ar1Y9
CH3 CH
\ I’O‘x h/’ 2\\
Fe+ + .~,o\v,. ——+ 6: Fe CH2
H1 H1 H2
2 58%
H 13 o +
shifts H Fe -—C2H5
i” '
+ + H
Fe (CZHGO) e———-CH3—CH2—o—Fe --H CH3
+ i H
C2H4 F +(C H ) + c H OH F +(C H 0) + C H
e 2 4 2 5 e 2 4 2 6
o o o ,/o
Fe+ + CH/ \_/\__/\__f\CH —-9 CH \Ctk
3 3 3 /, CH
H1 2 /,o q\
2 Fet—O-TH-CHZ \—J CH3
H
shifts H1 H2
shifts
_ CH
EET\ fl 3\\0
CH 06 H + CH —o 5 ‘“ | ~. Ho 0 o
3 2 5 3 - a \\
o\ //CH CH/,Fe + \_j CH
CH ll 3
2 CH
57
The successive reaction products, Fe(CH20)(C2H60)+ and
Fe(CH30)(C2H6O)+ in the reaction of Fe+ with dimethyl ether
can be explained as follows:
CH CH'//¢
3 3
CH3 1 CH 1
- Fe
3
\\0 + Fe -———9 \\0--~ + -——————+
/ 1‘7 / I
CH3 CH3
OCHZH o H
\\ ./’ 8
CH2
CH
3
\‘o-~Fe+ + CH
// 2 4
CH3 3
CH2:
CH3 CH3/7 CH
I 3
+ \x +
0 + Fe ——_79 O ------ Fe + CH3.
CH’// 1 //' 1
3 OCHZH CH3 OCH3
That is, the real structure of Fe(C2H60)+ is that in
which Fe+ has inserted into the C-0 bond. When another
molecule approaches, a B-H shifts to eliminate CH4. Alter-
natively, the incoming dimethyl ether can undergo substitu-
tion to replace the CH3 group.
Note that the Fe+ reaction with TDE only produces one
product ion, FeC3H60+. When the reactant neutral becomes
a polyether, multiple metal-ligand interactions are expected,
since these are observed in solution. While the number of
58
atoms in TDE with which the metal ion initially interacts
may be > 1, the actual reactions occur involving only one
site on the polyether. The initial multiple interaction
may be important in directing the metal to a site of attack
(insertion). The mechanism proposed above is typical in
that, when the metal ion rearranges TDE into two smaller
molecules, the smaller one is usually retained as a ligand,
presumably being due to its effective electron donating
ability of the smaller ligands.
b. Reactions of FeCO+
FeCO+ reactions with dimethylether, dimethylether and
TDE are summarized below:
.137 +
4 Fe(C H 0 CH 0
+ 0 + 1 2 6 )( 2 )
FeCO + /\ ———>Fe(CHO) —1
CH CH 2 5 863 +
3 3 '————-> Fe(C2H60)(CH30)
Feco+ +vov "0; Fe(C2H40)+ + C2H6 + co (2)
43-8-4 Fe(C H 0H)+ + c H + co (3)
2 5 2 4
..LUL; F +
Leg—9 Fe(C4H]o)+ + C0 (5)
+
91 +
0 0 0 -L——+ FeC H 0 + C H 0 + C0
5 12 3
CH/3\__/\_j\__/O\CH3 7 3 6 (7)
. .09 +
L——4 FeCOC4H802 + C4Hmo2
(8)
FeCO+ +
59
The processes in (1)-(4) were observed for Fe+ alone
and are presumed to occur by the same mechanism, however
here we have concurrent cleavage of the M-CO bond. Reac-
tion (5) is a commonly observed ligand substitution. Reac-
tion (6) is similar to (2) where an alkoxy group replaces
an alkyl group. FeC3H60+ in reaction (7) was also formed by
Fe+ alone as explained in the last section. In reaction (8),
since (a) C4H802 is not "extracted" from TDE by Fe+ alone,
and (b) the C0 is retained, we interpret this as a MCO+
insertion into the center of the molecule, followed by a
B-H shift.
0
FeCO+ + TDE ——-> I 1 + l—\
H Fe-——CH20 OCH3
.1
I
C gFe + CH30 OCH3
l " \_/
CH30 pCH
Note that, without incorporation of C0 in the inter-
mediate, no B-H atoms would be available to shift. The pro-
duct resembles a metal-ketene complex, which is frequently
observed in such studie591.
c. Reactions of other Fe(C0)x+ ions
As indicated in the reaction product lists, as the num-
ber of CO's present on the metal increases, ligand substitu-
tion becomes the predominant process. Fe(C0)2+ and Fe(C0)3+
60
will react with dimethylether to form FeCO(C2H60)+ with
loss of up to 2 CO's. Diethylether can displace up to 3
CO's and so on, which is typical of other systems studied85-
in fact most compounds which have been studied except alkyl
92’122 have been observed to displace CO from
fluorides
charged transition metal centers.
Successive reactions also become important at higher
pressures, for example:
Fe(CO): +.v,o\,,——+ Fe(C4H100)+ +\V,0\//——9'Fe(C4H10); + xCO
x = 1,2,3
+ +
+
Fe(C4H100)2 + yCO
y = 2,3,4
However, in these studies, no ion of composition
Fe(TDE)+ was observed. The reactions of Fe(CO): with TDE
are summarized below:
Fe(co): q__—4 Fec3H802 +C5H1002 + xCO x=2,3 (9)
___—4 FeCOC3H80 + C5H1003 + (x ‘)C0 x=2,3 (10)
.4 FeC0C3H502 + C5H1o°2 * "2 * (x"’C° x=4 (11)
.___3 Fe(C0)2C 4 H1002+ + C4H8°2 + (x'2)CO x=2’3’1‘12)
or Fe(C5H + C4 H 02 + (x-l)CO
1o 03) 8
61
All of the products for TDE are reactive rearrange-
ments of the TDE; none are C0 displacement. This may
reflect the ability of TDE to complex stepwise with the
metal center i.e.- initial complexation may involve one
oxygen of the ligand. As further oxygens interact, addi-
tional energy is made available for metal-induced decomposi-
tion of the polyether.
Hence reactions (9), (10) and (11) can be explained
H/U%O
as follows:
-CO
Fe(CO)+ + TDE C + C0
2 :\)/00\\_JC
H2 . shifts H3
shifts
i“ Q
/0—
W” +
0 3 \ .fl\,fl\\
// x + Fe+ CH3
C Ee—— CO I
3 C
C=t0 0
/
H -C0
+
62
C0
.,C/0000>_/\__/\__/\H
Fe(C0)2+ + TDE —;£9—+C H// H3 + C0
OL/F elk/H
CO
CH
Ko-—Fe"—Co 4—— CH/0 Fle+ + o
C2H5/ + 3 A/1 UL?\CH
F_9£Q£3fl89_ H
Note that the reactivity (which will be defined here
as the number of products) increases as the neutral changes
from an ether to a polyether. Also note that the CO ligands
can be lost in a stepwise manner.
The formation of Fe(CO)2C4H1002+ is very difficult to
explain, however, the following mechanisms are possible.
0C C0
\..4
I
\
’ \
CH
3\0
Fe(CO)2+ + TDE -———9 {E::U :::;H3 -—————9
\F + 0
/°\/ \/°\CH Coj
CH3 3
63
01"
CH CH
3 3
\. //’
0
+ 1 0
Fe(CO)2 + TDE ———> + E j
o o 0
\L_; 0
u
eC-—4<
H
B-H
shift +
Fe(C5H1003) + C4H8O2
2. Cyclic Ethers
a. Fe+ reactions with THF and THP.
Fe+ reactions with tetrahydrofuran and tetrahydropyran
are summarized below:
Fe+ + THF '6Qe Fe(C3H6)+ + CHZO
"9e Fe(C4H6)+ + H20
__;j!L, Fe(C4H80)+
Fe+ + THP —Hr—Ll§» Fe(C4H6)+ + CH4O
-—4§3—> Fe(C4H8)+ + CHZO
r—-L;§9 Fe(C3H60)+ + C2H4
‘05> Fe(CusO)+
Again, all the products can be explained in terms of a
metal insertion followed by a B-shift. The reaction mech-
anism of Fe+ with THP is shown as an example in Scheme XI.
(The reaction mechanisms for Fe+ with THF is shown in
Scheme II)
Scheme XI.
Fe
Fe + 0 -—-—-> o‘tfl—eo/ > + .2.“
Note that in the THF reactions, the formation of an
intermediate containing C3H5 is a strong driving force. In
THP, there is no favorable mechanism for forming this inter-
mediate. Fe+ does react with THP to form Fe(C3H60)+, which
is reasonable, being geometrically accessible (5-membered
ring) and that Fe+ interacts with both oxygen and carbon
equally well. (Table 2 p.20 ).
b. Fe(CO): reactions with THF and THP
Usually, the more ligands on the metal center, the less
reactive these species will be except in the reaction with
multifunctional organic compounds as explained in the last
section. Hence in the reactions of Fe(CO): with THF and THP,
65
we only see the same products as Fe+ formed (by the same
mechanism presumably) with concurrent cleavage of the M+-C0
bond.
Ligand substitution and successive reactions also
become important such as:
Fe(CO): + THF —————$:Fe(CO):_a(THF) + aCO
for x = 2 a = l
x = 3 a = 2
x = 4 a = 2,1
x = 5 a = 2,1
Fe(CO); + THF —T———>»Fe(1HF)+ + co
-—-—+ FeC0(THF)+ + C0
Fe(THF)+ + THF .5 Fe(THF);
+
-———> Fe(CH20)(THF) + 63116
+
~———9'Fe(C3H6)(THF) + CHZO
3. Cyclic Polyethers
a. Fe+ reactions with 1,3 dioxolan,
l,3 dioxane, p-dioxane-(dglé
lZ-crown-4 and lS-crown-S
Fe+ reactions with 1,3 dioxolan, 1,3 dioxane, p-dio-
xane (d8), lZ-crown-4 and lS-crown-S are summarized below
and in Table 5.
66
+
Fe (C2
Fe+(CH
+
Fe (C2
Fe+(CH
Fe+(C2
Fe+(C2
Fe+(C2
Fe+(CH
Fe+(C2
4.
Fe (C3
.1.
Fe (C2
Fe+(CH
H4)
20)
H40)
202)
H202)
H402)
H4)
20)
H202)
H60)
H402)
2°)
Fe+(C3H60)
Fe+(C2
H402)
Fe+(C2H4)
Fe+(CD
.1.
Fe (C2
Fe+(C3
20)
D4)
060)
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Fe (C20402) +
CHZO
C2H4
CH
CZH4O2
C3H60
c2H6
CH 0
CZH4
C3H60
CHZO
C2H4
C H 0
2 4 2
C D O
3 6
C20402
C020
C2D4
(l3)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
(22)
(23)
(24)
(25)
(26)
67
In its reactions iwth monofunctional ethers, both
acyclic and cyclic ethers, it is energetically favorable
for Fe+ to insert into the C-0 bond, followed by a B-H
shift. If Fe+ interacts with two geometrically accessible
oxygens which are in close proximity, such an interaction
may involve more than twice the energy as the interaction
124
with only one oxygen due to the ligand effect This may
result in a double insertion and double B-shift mechanism.
This process can be stepwise145
, although it is convenient
to use a concerted double insertion, followed by double B-H
shifts (using a stepwise mechanism, it is hard to explain
reactions (17) and (18)). Accordingly, the products in
reactions (13)-(18) can be explained as follows:
,Fef H1 /o
Fe+ + 0A0——> GOO ““19 . J>Fe+ H2
insert1on7
L—j Lj y l\0 H2
+ +
C2H4 + Fe(CHZOZ) H—Fe + Z)
+
Fe,
+
0
x . ./
./ ‘~ double , 4+
Fe + q/Ajp'—_9 (k‘\\,/*’P insertion" <:;:;EE\\J\\
o
CHZO + Fe+(C2H4O) <——1— ...—pet---"
1 CH2
——--0
+
C2H40 + Fe (CHZO) e———I
Table 5
Fe+ +[:
. Ion/Molecule Reactions of Fe+
68
And lS-crown-S.
r—\
o 0:]
QL_P
.22
.13
.41,
.02
+
Fe C2 2
.1.
Fe CZHZO
Fe+C H o
2 4
+
Fe C4H40 + C
2
2
+ C6H14°2
+ C H
6 1202
H o +
4 10 3
+
Fe C4H60 + C4H1003
+
Fe C4H80
+
Fe C4H100+C4
+
Fe C3H60
2 + c4H8°2
2
H603
+ C5H 0
10 2
+
Fe C4H803 + C4H80
+
Fe C6H
1402
_. Fe+(C2H202) + C
+ C2H202
8“160
3
+
+
Fe (C2H6O) + C8H1404
+ .
Fe (C4H802) + C6H1203
+
+ .
Fe (C4H803) + C6H1202
Fe+(C8H]403) + C2H60
3
2
H
H
20
2
+H2
With lZ-crown-4
See
Scheme
XII
XIII
XIII
XIV
XIV
XII
XIV
XIV
XIII
XV
XV
XV, XVI
XVII
XVIII
XVIII
XV
69
+ /\‘ x’ = double ‘
Fe + QL_P 1 OrvL/X’P insertion ’
H1
\—0\ + +
/
1
shift
H
0
’Azx' + +
”,Te———CH2-—é Fe(CZHZOZ) + CH4
\oH
Note that there are only three possibilities for double
insertion, each case gives the corresponding products.
Reactions (23)-(26) were identified unambiguously by
p-dioxane d8 reactions. The product ion in reaction (21)
can be either Fe(C2H202)+ or Fe(C3H60)+. Fe(C3H60)+ is more
reasonable based on the p-dioxane reactions (23)-(26) and
can be easily explained by the proposed mechanism which
follows.
Note that the chair form of p-dioxane is the most
stable conformation based on electron diffraction, sector
125,126 127-129
microphotometer studies , and complexation
studies. It is shown thatZLCOC of p-dioxane is 112.45°
126. There-
which is bigger than that of cyclohexane (108°)
fore we assume that it is the most stable conformer in the
gas phase. From X-ray studies of 2-chlrophenyl-l,3
70
130 131,132
dioxane and conformational analysis and modeling
studiesl3], it is agreed that l,3 dioxane is more stable
in the chair form with theL01C3O3 = 111°, which is puckered
(dihedral angle, T = 63°) and LC C C = 108°, which is
4 5 6
flattened (dihedral angle, 1 = 55°), although the under-
derivatized ligand has not been studied. Since Fe+ has al-
most the same bonding energy to oxygen and to an alkyl
group (Table 2) and because of the geometrical accessibili-
ty, Fe+ is capable of interacting with both oxygens and one
carbon as indicated by dashed lines in the following mechan-
isms. However, it appears that Fe+ is even more likely to
interact with one oxygen and-one carbon instead of two
oxygens due to the geometrical restriction. The products
in reactions (l9)-(26) can then be explained by the "double
insertion, double B-H shift" mechanism, as shown below:
' +
/,Fe-“\ d 1 O\\ fi
u “0 oub e 5 L .../W W
- If, 7 F6 -———9 -----Fe*
insertion
&£::ji::::;z \\__o ‘///EEZ//‘1‘\..O
Fe+(C3H60) Fe+(CH20)
,Fe
. 0
,’ ‘ / 0
I ‘~ double A + --u ,+/’
Fe+(C2H4) Fe+(C2H402)
71
Fei
’x' “‘.0 /0 +
N ——-x—+ Fe’ ——9 Fe (C2H40) (27)
, 0
Similarly,
,oFei. o o
3W“ ...... . \..v H ..... ..4/
ZL/C“;-;7( insertion’ 4\\_:>7__9
CH
+
Fe(CHZO) Fe(C3H60)
Fe+ /g\
M0 30. {01
E Fe“
4 6
5
Ee+
,'1/ 0 0 . 0
FTFTT“/// gaggliion$ £::¢::Fe+ ———:fl:;Fe:::o:>
+ +
(28)
———)(-———> Oe<:> ——>)|;€+<:> (29)
+ +
Fe(CHZOZ) Fe(C3H6)
72
Note that the only reason why reaction (27) doesn't
occur is because Fe? can't interact with two oxygens and one
-CH2- group at the same time. It is difficult for Fe+ to
interact with two separated oxygens by breaking bonds as
shown in (27). Instead, it prefers to interact with one
oxygen and one methylene group. In the case of l,3-dioxane,
Fe+ can either interact with two oxygens or one oxygen and
one -CH2- group equally well. Since the 03-C4-C5 angle is
flattened130
, the -CH2- at C5 can have a weak interaction
with Fe+ unless Fe+ moves closer and weakens its interaction
with another oxygen as in reaction (28). The reason why
reaction (29) doesn't occur is probably because the 01-C2
and C2-03 bonds become weaker than other bonds, due to the
strong interaction of Fe+ with two oxygens, which forms a
strained four membered ring intermediate and results in an
insufficient orbital overlap of 01-C2 and C2-03 bonds, in
contrast with more flexibility of 01C6C5C403 ring on the
other side. Therefore, 01-C2 and 02-03 bonds are easier to
break and C6-01 and 03-C4 are more resistive to cleavage.
Also note that all of the products in the reactions of
Fe+ with 1,3 dioxane and 1,4-dioxane are from the metal ion
"double insertion" only, because B-H's may not be geometric-
ally accessible for shifting after double insertion by the
Fe+, which might have a linear structure for the inter-
mediate, RCH2 -Fe+ -0R. Linear structures for insertion
products have been suggested by D.P. Ridge in the reactions
73
of Fe+ with alkane5194.
In 12-crown-4, the chemistry is even richer. The reac-
tion products together with their reaction mechanisms are
indicated in Table 5.
In some cases, Fe+ appears to break down the polyether
into smaller cyclic ethers, which is also observed on elec-
tron impact of crown ether5133. One product, FeC4H802+,
appears to be the result of cleavage of 12-crown-4 into two
1,4-dioxane molecules which is shown in Scheme XII. The
process is favored since 2 molecules of 1,4-dioxane are
approximately 20 kcal/mole more stable than 1 molecule of
12-crown-4134.
This arrangement of 12-crown-4 is consistent with a
distinct secondary structure for the polyether which is
shown in Scheme XII. This structure was derived from X-ray
crystallographic studie5136.
Thus, we assume that it may
also be a favored configuration for the uncomplexed crown in
the gas phase. (While the molecule is free to assume many
secondary structures, intramolecular hydrogen bonding may
107’111.) Using the numbering
favor this configuration
system of Scheme XII, it can be seen that this chair-type
structure has two oxygen atoms whose lone pairs are
directed above the "plane" of the molecule (1 and 2) and
two below (3 and 4). If Fe+ can complex with this ligand
without extensive changes in the secondary structure, initial
complexation with lone pairs on oxygen atom #1 and #2 will
74
+o~=~o£
«I..—
floxJo ... +0¢...uu..._
07x
“ u
._ ...;
«cameo + woozeu 6.. «......mo «We Imzo
. “fl \\\ d
a O . . O \UNI JUN—...
o s
.. .. .. 1 \on
C . C x
v
«solo d «zolo...
.’ V ’ n N
are“ «Ifilmro; «Ivan «IO-I 10%..
0:30. .0”: o, a .1111. Notwo ifw wax ... on
/ u z N / «as. ..
OlNIU‘ mull. O
+£\ “
SM .mzmxum
75
direct the metal towards C #5. These three atoms are in
close proximity in this model. Scheme XII shows that such
an initial interaction would lead to the FeC4H802+ product
if Fe+ can bring 0#1 and C#5 in close proximity to form two
rings (two p-dioxanes). On the other hand, if this inter-
action brings C#5 and 0#2 in closer proximity, 9-crown-3
can be eliminated, leaving FeC2H4O+ which rapidly loses H2
to form FeC2H20+, which is also observed. Thus, the geometry
of "free" 12-crown-4 suggests that, regardless of which side
of the ring Fe+ attacks, there will be two 0 atoms and one
CH2 unit (which is across the ring) in close proximity for
initial interaction with the metal, similar to the cases of
l,3-dioxane and 1,4-dioxane which have been discussed
earlier. The geometry may facilitate the formation of new
C-O bonds and smaller rings.
Other products can be explained following initial metal
complexation to the two oxygen atoms shown in Scheme XIII
and XIV followed by metal ion double insertion and double
B-H shifts as previously discussed. There are only two
possibilities. Fe+ can interact with either two oxygen
atoms next to each other or two oxygen atoms across the
ring. Scheme XIII shows the former case to form FeC2H402+,
FeCZHZOZ+ (similar to a metal-butadiene complex) and
FeC H
6 14
with oxygen atoms #1 and #2. Following the double insertion,
02+. Note that it also follows initial interaction
we see a situation where both the "left" and "right" groups
76
o. o.
. “on. 1A1 m wan.
Q + . , o. +
I..\ L »
\\ . 1
WI 01/.
on... AAIIIIII fl om...
.... o.\ o l\_._.
«I; aim IaQ 25:00
0)
AU. e9 1.... flow...
3.113.... “W
”EM MEMIUm
77
SCHEME XE
b0
+H
(:fFe 33—»OHQ: :Fe
DOUBLE
SHIFTS
H
(‘9?
c4
limo/—
\_
HHO .
/‘°\ +
H0 -—> O ,Fe+ o
K,o
m/z l60
H603
m/z I30
78
contain B-H atoms which could shift. Apparently, only those
on the "left" do. Molecular models suggest that further
interaction of the oxygens on the "right" group with the
metal ion may move the "right" B-H atoms to positions
spatially less accessible. These additional metal-oxygen
interactions could thus, through an intermediate such as is
shown in Fig. 8, control the availability of 8-H atoms for
rearrangement.
Scheme XIV predicts the remainder of the products
through another double insertion, double B-H shift scheme,
following an initial complexation of Fe+ with two oxygen
atoms across ring. Note that if H's from the "left" shift,
FeC4H100+ is formed, which corresponds to an ion of m/e 130.
An alternative structure (FeC3H602+) for the same ion can be
explained by a double insertion:
flfi —-——)<—O\Fe /“\0
Fe+ +.4E§L—PH 0_—J/Z:/0\3>
Here, Fe+ interacts preferentially with oxygen atom #l
+
-——9 Fe(C3H602) +C5H1002
and carbon atom #5 as indicated in Scheme XII and inserts
into both the C-6 and the C-0 bonds similar to reaction (28).
Nevertheless, CID or high resolution spectra are required to
unambiguously distinguish FeC4H]OO+ from FeC3H602+ .
In the reaction of Fe+ with lS-crown-S, one may expect
to get more products, presumably due to increased interaction
79
+ 4»
Figure 8. A Structtn'e Of Fe-lZ-crown-JJ' After Fe
Double Inserts Into 'IVto C-O Bonds 01‘
1242:011th
80
of Fe+ with five oxygen atoms in lS-crown-S. However, Fe+
apparently can't react effectively with all 5 oxygens.
(lS-crown-S has much bigger cavity than 12-crown-4 does
(1.7 - 2.2 A vs, 1.2 - 1.5 A in diameter)).
Although there are many reports on the complexed struc-
137 and also its free 1igand138, there
112,139,140
tures of 18-crown-6
are only a few reports on complexed lS-crown-S
and the study of the free lS-crown-S molecule has not been
reported yet. However, the structure of lS-crown-S was
taken by X—ray and drawn in reference 139, in which the
copper in CuBr2 interacts with only one oxygen atom of
lS-crown-S, which makes the conformation of 15-crown-5
similar to that of a free ligand as shown in Fig. 9.
Figure 9
X-Ray Structure Of lS-Crown-S
In lS-Crown-S CuBrz.
81
Scheme XV
Hb
a
0 Pet 0 b a (\O */
Q ~ 3 ——+ . x
0d 0 (‘5: :0]
ZA-H
shift
+
Fe(CzHSOH) + [k0 :3)
o
82
Scheme XVI l
3,1) J
b a 0
< fig/j
——>Fe(C482+HO)
0\\/o/ \\/o 605
Scheme XVI I I
(>5 6 [mi
<1 . .403
I :3;
. 3.; ’
+
+
bx“
83
Hence, oxygen atom #4, #7 and #13 are up and #l, and 10
are down. Fe+ then can interact with [0#4 and #7], [0#4 and
0#13], or [0#7 and 0#13]. Since carbon atom #11 is with the
metal up also, the interaction of both [0#13 and C#11] and
[0#4 and C#ll] are also possible. Schemes XV, XVI, XVII,
and XVIII are the proposed mechanisms leading to the observ-
ed reaction products of Fe+ with 15-crown—5.
b. FeCO reactions with 1,3 dioxolang,
l,3-dioxane, p-dioxane, lZ-crown-4 and
lS-Crown-S.
Most products of FeCO+ reactions with l,3-dioxolan,
l,3-dioxane and p-dioxane are seen in the Fe+ case with
concurrent loss of CO except as shown in the following:
/“\
Feco+ + °\__,° ———> Fe+(C3H602) + CO
L————+ Fe(CH20)(C3H602)+
4————--—> Fe(C2H4O)(C3H602)+
‘1’
4.
Fe CO(C3H60) + CHZO
FeC0+ + Q
+ 0 +
FeCO + Ej—a Fe (C4H802) + co
0
. +
, Fe (C4H802) + C0
+
-—————9 Fe (CH20)(C4H802)
(30)
(31)
(32)
(33)
(34)
(36)
(37)
(38)
84
Reactions (30% (34) and (36) are typical of ligand
substitution processes. In the l,3-dioxolane, l,3-dioxane,
p-dioxane systems, the addition complexes are reactive
enough to undergo successive reactions (31) (32) (35) (37)
and (38). Note that the reaction products in (31) and
(35) are also formed from Fe(CHZO)+, which is reactive too.
Presumably, CHZO in reaction (33) carries away most of the
reaction energy, leaving CO retained by Fe+.
FeCO+ forms seven products with 12-crown-4; six of them
were also formed by Fe+ alone. Thus, MCO+ insertion does
not appear to predominate in the reactions with the cyclic
polyether. The new product for FeCO+ with 12-crown-4 is m/e
204; FeC6H1204+, corresponding to loss of CO from the ionic
reactant and C2H4 from the crown, presumably by a mechanism
shown on Scheme XIX.
On the other hand, FeCO+ forms nine products with
lS-crown-S, which are shown below.
FeCO+ + 1505 4411—» Fe+(CZH4O) + C8H1604 + c0 (39)
+
.14 1 Fe (C2H202) + C8H1803 + co
14 +
.20 +
_JEL_. Fe+(C H o ) + C H o + CO
4 8 2 6 12 3
.08 +
_;___s F°+IC6H12°4) + C4H80 + CO (40)
85
Scheme XIX
FeOO' + 12-crown-4 A
°‘1°°
.Fef’ j
(,0
T
(A
E j
o o
[0 o) + 02H4+GO
Scheme XX
86
+ .04 ‘ +
FeCO + 15C5 7 Fe (C8H1403) + CZHGOZ + C0
.06 + +
Of which, six products were also formed by Fe+ alone.
Again, MCO+ insertion does not seem to play a role in this
larger cyclic polyether.
The new product Fe(C2H40)+ in reaction (39) is presum-
ably formed via the mechanism shown in Scheme XV, following
double insertion as indicated by the pathway b. The C0
ligand plays an important role here. Apparently, available
energy can be used to break a M+-CO bond instead of inducing
B—H shifts. The product ion Fe(C6H1204)+ in reaction (40)
is formed through interaction with oxygen atoms #7 and #13
as shown in Scheme XX. Note that we do not see this ion in
the reaction with Fe+. Since oxygen atoms #7 and #13 are
far apart, the CO ligand here may play an important role in
bringing them closer as shown below:
FEW
.“ j/ ’:0
‘Fe’ :)
0 0
\__/
87
Once Fe+ double inserts into C-O bonds, it may further
interact with oxygen atoms #1 and 4 to eliminate CO and
C4H80. The retention of CO on the product ion in reaction
(41) clearly indicates that no further stabilization energy
is produced (product distribution ratio for this ion is 0.06
and only 0.05 in the reaction with Fe+).
c. Fe(CO): reactions with l,3-dioxolan,
l,3-dioxane,_p-dioxane, 12-Crown-4 and
lS-crown-S.
Higher CO-containing iron ion molecule reactions with
1,3-dioxolan, l,3-dioxane, p-dioxane and 12-crown-4 only
undergo substitutions and successive reactions. Note that
in the reactions with 12-crown-4, a new ion, FeC0C2H202+ is
formed at m/e 142, presumably by the same mechanism as is
shown in Scheme XIII.
In the reactions of Fe(CO): with lS-crown-S, there are
few new ions formed:
4 + +
Fe(CO)2 + lS-crown-5-———+ Fe(C2H602) + C8Hl403 + 2C0 (42)
+ +
Fe(CO)X + lS-crown-5-———+ FeCO(C2H602) + c8Hl403 +
(x-l) CO x = 2,3,4 (43)
+
Fe(CO)2 + lS-crown-S ———9 Fe(C6H1204 + C4H80 + 2C0 (44)
+ +
Fe(CO)2 + lS-crown-5 ———+ Fe(c8H1605) + C2H4 + C0 (45)
+
Fe(CO)3+ lS-crown-S ———9 Fe(CBH1605)+ + C2H4 + 200 (46)
88
The product ion, Fe(C6H1204)+ in reaction (44) can be
formed by the mechanism shown in Scheme XIX. The loss of 2
CO's might be the result of strong interactions of Fe+ with
oxygen atoms in lS-crown-5. Fe(C2H602)+ and FeCO(C2H602)+
in reactions (42) and (43) are formed by the mechanism shown
in Scheme XV. Instead of forming the Fe(C8H1403)+, which is
formed in the Fe+ and FeCO+ reactions, these two products
are formed by retaining the smaller ligand (C2H602) with a
concurrent loss of 1,2 or 3 CO's indicating the preference
of iron ion to retain the smaller ligands (as noted by the
product distribution ratio, .2 for the reaction (42) and
only .06 and .05 for the formation of Fe(c8H1403)+ and
FeC0(C8H1403)+ respectively). Finally, a similar mechanism
to that shown in Scheme XIX can be used to explain the forma-
tion of Fe(C8H1605)+ in reactions (45) and (46). Note that
as in the 12-crown-4 case, C0 ligands appear to be important
in forming this big metallocyclic product, since the energy
which can be used for B-H shifts is instead used to break
M+-CO bonds.
4. Thermodynamic Conclusions
Table~ 6 lists all neutrals lost in the Fe(C0)x+
reactions with linear ethers, polyether, cyclic ethers and
polyethers which we have just discussed in last sections and
branching ratios. (Product distributions)
Fe+ induces the rearrangement of dimethylether into
CHZO and CH4, but FeCO+ does not. This can be readily
89
understood in terms of thermodynamics: (Dimethylether will
displace one, but not two CO's from Fe+)
Assumption
Feco+ + CH OCH
+
3 3-—————+ Fe (CZHGO) CO AH < O
Y‘Xl'l
+ +
Fe (C0)2 + CH30CH3 —*—— Fe (C2H60) + 2C0 Aern > O
This implies that the initial Fe+-dimethylether inter-
action energy is 27.17 kcal/mole < D(Fe+-— MeZO) < 73.17
kcal/mole.
Note that there are two ways to interPret these
'ts
results. (All AHf's are taken from ref. 142 and all un1
are in kcal/mole).
A. In terms of bondstrengths
Feco+-—————— Fe+ + co .°. AH = 27 kcal/mole
282 -26.42
+ + .
Fe(CO)2 —————— Fe + CO . . AH = 73 kcal/mole
282 -52.83
.'. 27 kcal/mole < D(Fe+-Me20) < 73 kcal/mole
90
B. In terms of heats of formation
FeCo+ + CH OCH -———9 Fe(C2H60)+ + C0,
3 3
228 -43.99 -26.42
. +
. AHf(Fe -Me20) < 2l0
Fe(CO): + CH OCH3 ———+ Fe(C2H60)+ + 2C0,
3
156 -43.99 -52.83
. +
But,
+ + +
Fe(CZHGO) ‘———9 Fe + C2H60 Aern = D(Fe ~Me20)
D(Fe+-Me 0) = AH (Fe+) + AH (Me 0) -
2 f f 2
AHf(Fe+-Me20) = 282 + (-43.99) -
[l65 < AHf(Fe+Me20) < 210]
°. 28 kcal/mole < D(Fe+-Me20) < 73 kcal/mole
If, however, the structure of Fe(CZHGO)+ is really
CH -Fe+-0CH3, then
3
+ +
D(CHB-Fe -OCH3) = AHf(Fe ) + AHf(CH3) + AHf(0CH3) -
+ _ > 165
109 kcal/mole < D(CH -Fe+-0CH ) < 155 kcal/
3 3 mole
91
8.: 8? 33.3 m
.‘I 8 8“ 8" .2
tn; 8~ ~ fl .
E. V RAHNMMIWM amen 8~ EV ax . ....J 8.. Q 8” a. a; was...
AM. V 8Tr=koo Emu SN 92 Rive—3:00 8” 8A 8‘ 8— V _
. . 8" V 8~
ASH“ 8~ .noflw. ax 8~ 5 .Vlfifiofmflml 1 a 8N E a. BN V at. .luléil
GN. ”NJ Sen. . AS. V 8~ . ....Nu _ V
E.“ 8 .. my} 8~ V. E. V 82 frag 8" 5 V
m? 8102....“ an a. _ 8.. as}: a. a. a. a. _ a. a. V 42V...
53-8 :23:- Iil 3 o TAB; Bataan—Mn FN V 83 aznu _+
$3 nomenn A8; 8 . Norm. . h
$~.V 8 .53.an V 29V 8N . owzau _
:~.V 8 .nooaznu V Go V 8~. SNOMIMUS ARV 8‘
. .95.? 8 a V... u 3: R; V V
V 3: 8~ V .
3.: Eu . saw so; Nam—.... Run 8" V 3.. V Bracing E: 89...]. 8V 8. 8. _ 8V _
23 2.2.3300 _ 31V 8.68.4". 87:3. 8 V :V 8~.no..£mu 2: Sauna 8~ 8 an H 8~ V B WEB...
. V 33 No.2”... V V. V
_ V 25V 8 . ~%=~u V _ V _
V8; 8 . ......u V V 33 8....amxau , , V.
. ~ . u _ . SUV 8 . omzau V V
Go V 8 . oéd . _ V
E: 8 . ox} V .3 V amass} EV 8 _ V V
1.1V 8 . “05:... So; 8 8~ V 83°..er 3.88. :6 _ V V
REV 8.5046 3: 8 . m: 29V 8 7: V 8.3.3.} 3982.0 :13 8 , ARV 8 3: 8 V V
EV 8 .62.} ARV 8.93.} 2.9V 8.3.6 ”8.83“: g V Examrmu A$.V8...=~u “2.8.415 , 2~.VR..Vm~u A8.V8.a=..u V .
. o :5; woo—nap 81V 8.0.} Ao~.V8.onVo ASV 83°}? ARVBSHE 2.18.3} 2.98 238.38 V 8 8 .82.
AS V ~00sz V V
2~ V N022$ is. V :5 V. V
_ EV nan”? 2.... V _
Va; .4. EV a mu ......V ....m. 2.3 ....m. .3 Agnes...
. d no. u
3.3 “:0 SNV WV... . GNV ”flux". AR. V are a V o=~u 3o. V cwau :3 axm ”at” ....uw
. 3V 3 E V o _ a: mu REV 0:1... 3: wean”. Noawe 2 . we r.
. n . . u . n . a V 39V ~ V o «3 V cmzu u<
:3 ... .31.} 3 5 amino E V 0...». .8 V ed as w: at} a V ~39 E. V :1: V cacao V :3 ox? :3 and EV «a ....
- 2) 0 . G ..O V O
44:53 m... E ‘ E.— O _ O /o\
El: 5:. “ABE. co 1.3%.! I: 5 an: 55.... Jljlc . an.
92
Also, the energies required to rearrange/fragment the
organic molecule are 0.9 kcal/mole and 82.6 kcal/mole for
the following two processes respectively
CH30CH3 —77——<> CHZO + CH4 AH 0.9 kcal/mole
> CH3 + OCH3 AH
82.6 kcal/mole
Obviously, the first process (involves Fe+ insertion
into C-O bond, followed by a B-shift) is a low energy pro-
cess, which has to pass through an intermediate (insertion
product) related to second process, in which the insertion
product is formed. Although B-H shifts might not require
energy (i.e. naturally occur), the extra energy after form-
ing the insertion product will be removed by the Fe+-CO
bond breaking (2 27.8 kcal/mole)142
to get no further reac-
tion and stop at Fe(CH20)+.
Since the initial complexation energy is smaller than
the energy required to dissociate dimethylether into CH3 and
0CH3, the real mechanism might look like:
[ Fe+ r Fe+ /0
+ 5 x’ g ,+
Fe + /0\—) 5 —-> /0“\: ——-) /Fe CH3
./°\J L CH3 CH3
(II) (III) (1v)
0
+ +
CH4 + Fe(CHZO) é——-CH4...Fe ...//JL\\
93
Fe+ reacts with dimethylether to form (IV) with much
energy released which is available for further reactions.
However this energy will be taken away by a C0 ligand in the
FeCO+ reaction. This explains why precursors of (IV) are
composed of only 8.6% of Fe+ and 9l.4% of FeCO+ (see
Appendix F).
Analogously, diemthylether will displace up to three
CO's from Fe+ implying that the initial Fe+-diethylether
interaction energy is 96.6 kcal/mole < D(Fe+-Et20) <
l33.3 kcal/mole.
The energies required for Fe+ insertion into a C-O bond
in diethylether, and for rearrangement procuts are:
C2H50C2H5 —— > C2H5 + OCZH5 AH = 82.3 kcal/mole
_———> C2H50H + C2H4 AH = l6.8 kcal/mole
The second process is the lowest energy process among
these three processes. A possible reason why the formation
of Fe(CzHSOH)+ can displace up two CO's from Fe+ and only
one CO for forming Fe(C2H4O)+ is that Fe+ may form a stronger
bond with CZHSOH than that of C2H40. However, it is not
understandable what makes so much difference in complexation
energy of Fe+ with dimethylether from that of diethylether.
Could it be from the polarizability difference?
The rest of the reactions can be explained in a similar
way. The data to date can be summarized as follows:
94
l. Whenever B-H atoms are available, they will shift
to form stable products, following insertion into a C-O
bond. If CO ligands are present, the breaking of M+-CO
bonds may compete with this process.
2. A stable addition complex can only be formed with
the concurrent loss of one or more CO ligand on the metal
ion (to take away the energy).
143" is seen. In
3. A "mechanistic macrocyclic effect
the case of lZ-crown-4, the Fe+ actually induces reactions
to product 9 products, but only one product is observed in
its linear analog (TDE).
4. In the case of l2-crown-4, we assume that the Fe+
actually induces reactions from "inside" the crown cavity.
This is implied by the fact that Fe+ reacts with lZ-crown-4
to form 9 different products, FeC0+ gives 7 products,
Fe(CO)2+ gives only 2 products, and as more CO's are added
to the metal, no reactions are observed. Thus, ligands can
prevent the metal from entering the crown cavity and induc-
ing reactions‘43. However, this may not be true in the
l5-crown-5 case, which has a larger crown cavity. Hence,
Fe+ reacts with l5-crown-5 to form 7 products, FeCO+ gives
9 products, Fe(C0)2+ gives 6 products and both Fe(CO)3+ and
Fe(C0)4+ give 2 products respectively.
5. Many products of reactions involving cyclic poly-
ethers, lZ-crown-4 and l5-crown-5 can be explained using a
double metal insertion, double B-H shifts process. In these
95
reactions, CO ligands act predominately as spectators.
6. When CO's are present on the metal center, they
can act either as spectator ligands or as active groups
(vis MCO+ insertion). An alternate interpretation is that,
after M+-C bonds are formed, a CO ligand on the metal may
insert into the M+-C bond.
7. The strength of a metal-ligand bond alone does not
guarantee complexation in the species studied here. In the
case of bulky ligands, the presence of 3C0's on a metal
center may prohibit sufficiently close approach for inser-
tion.
+ .
II. Cr(C0)x Reactions With Ethers And Polyethers
A. Results
1. Linear Ethers and Polyethers
a. CrflCO)x reaction with dimethyletherLCzflfigl
Ions formed as products of ion-molecue reactions in a
mixture of Cr(CO)6 and dimethylether are listed below, with
their precursors as identified by double resonance.
m[e stoichiometry precursor(s)
98 Cr(C2H60)+ Cr+C0, Cr(C0)2+
+ + +
126 CrCO(C2H60) Cr(C0)2 , Cr(C0)3
+ +
144 Cr(C2H60)2+ CrC0+, Cr(C0)2 , Cr(C0)3 ,
Cr(C0)4+
96
b. erC0)x+ reaction with diethylether
L24fll 09.).
Ions formed as products of ion-molecule reactions in a
mixture of Cr(C0)6 and diethylether are listed below, with
their precursors as identified by double resonance.
mle stoichiometry precursor(s)
+
96 Cr(C2H4O) CrT
11o Cr(C3H60)+ CrCO+
+ + + +
126 Cr(C4H100) Cr , CrCO , Cr(CO)2
149 (C H O) H+ OCH T OC H T C H 0+
4 1O 2 3 ’ 2 5 ’ 4 11
The ion at m/e 149 is (C4H100)2H+, the protonated
diethylether's dimer is formed by C4HHO+ reaction with a
neutral molecule of diethylether, which in turn is formed
+
by OCH + and OCZHS .
3
c. Cr(CO): reaction with triethylene glycol
dimethylether (TDE, C851894l
Ions formed as products of ion-molecule reactions in a
mixture Of Cr(CO)6 and TDE are listed below, with their pre-
cursors as identified by double resonance.
mle stoichiometry precursor(s)
84 Cr(CH4O)+ CrC0+, Cr(CO)2+
+ +
91 C4HHO2 C2H30
+ +
96 Cr(C2H40) Cr
.1.-
98 Cr(C2H60)+ CrCO+, Cr(CO)2+, Cr(CO)3
+ + +
101 C5H902 C3H60 or C2H202
97
gig stoichiometry precursor(s)
11o Cr(C3H60)+ CrC0+, Cr(CO)2+
112 CrCO(CH40)+ CrC0+, Cr(C0)2+
l26 Cr(C3H602)+ Cr+, CrC0+, Cr(CO)2+
128 Cr(C3H802)+ Cr+, CrCO+, Cr(CO)2+
l38 CrC0C3H60+ Cr(CO)2+
14o Cr(C0)ZCH4O+ Cr(CO)2+, Cr(CO)3+
142 Cr(C4H]002)+ CrCO+, Cr(CO)2+
156 CrCOC3H802+ CrCO+, Cr(CO)2+
l66 Cr(C0)2C3H60+ Cr(C0)2+
168 CrCOC4H802+ CrCO+, Cr(CO)2+, CrCOC3H60+
17O CrC0C4H1002+ ’ CrCO+, Cr(C0)2+, CrC0C3H60+
2. Cyclic Ethers and Polyethers
a. CrLCO)x+ reactions with tetrahydrofuran
(THF, CQHBQ)
Ions formed as products of ion-molecule reactions in a
mixture of Cr(CO)6 and THF are listed below, with their
precursors as identified by double resonance.
ELE stoichiometry precursor(s)
124 Cr(C4H80)+ Cr+, CrC0+, Cr(CO)2+
+ + + +
152 CrC0(C4H80) Cr(CO)2 , Cr(C0)3 , Cr(C04
+ + + +
l96 Cr(C4H80)2 CrCO , Cr(C0)2 , Cr(C0)3 ,
Cr(CO)4+, Cr(C4H80)+,
CrC0(C4H80)+
98
b. Cr(CO)x+ reactions with tetrahydropyran
QTHFz Csfllog)
Ions formed as products of ion-molecule reactions in
mixture of Cr(C0)6 and THP are listed below, with their
precursors as identified by double resonance.
gig stoichiometry precursor(s)
11O Cr(c3H60)+ Cr+, CrCO+
138 Cr(C5H100)+ Cr+, CrCO+, Cr(CO)2+
166 CrCO(C5HwO)+ Cr(C0)2+, Cr(CO)3+
224 Cr(CsH100)2+ Cr+, CrCO+, Cr(CO)2+,
Cr(C5H100)+
c. Cr(CO)X+ reactions with 1,3 dioxolan
1&3fl5921
Ions formed as products of ion-molecule reactions in a
mixture of Cr(C0)6 and l,3-dioxolan are listed below, with
their precursors as identified by double resonance.
m/e stoichiometry precursor(s)
82 Cr(CH20)+ Cr+
+ + +
126 Cr(C3H602) Cr+, CrCO , Cr(C0)2
+ + +
154 CrCO(C3H602) Cr(CO)2 , Cr(CO)3
+ + + +
156 Cr(C3H602)(CH20) Cr , CrCO , Cr(C0)2 ,
....
Cr(C3H602)
99
d. Cr(CO)x+ reactions with l,3-dioxane
Ions formed as products of ion-molecule reactions in a
mixture of Cr(CO)6 and l,3-dioxane are listed below, with
their precursors as identified by double resonance.
gig
82
110
112
140
144
168
170
196
197
228
01"
stoichiometry
Cr(CH20)+
+
Cr(C3H60)
+
Cr(CzH402)
H O +
C7 12 3
CrCO(C4H802)+
+
Cr(C4H802)(CH20)
+
Cr(CzH402)(C3H60)
Cr(CO)2(C4H802)
Cr(CO)2C4H902)+
+
Cr(C4H802)2
precursorLs)
Cr+, CrCO+, Cr(CO)2+
+ +
C3H5 , 50 , C4H7O
Cr(CO)2+, Cr(CO)3+
+ + +
Cr , CrCO , Cr(CzH402)
+
C2H 2
Cr(C0)4+
+ +
C3H70 or C2H302
Cr+, CrC0+, Cr(CO)2+,
+ +
Cr(C4H802) . Cr(CO)3 ,
+ +
CrCO(C4H802) , Cr(CO)4
Cr(CQ)x+ reactions withgp-dioxane (C4H892)
and its d8 isotope
Ions formed as products of ion-molecule reactions in a
mixture of Cr(CO)6 and l,4-dioxane and p-dioxane-d8 are
listed in Table 7, with their precursors as identified by
double resonance.
100
+Amomachnu mmomxaoVBuu
. NMBVHV v.8 mAuomnaoVno EN . «mommaora .
+o~8 £83088 :8 . +A8V8 p.88 mAmowzaoVB 9.8
mA8V8 .WABVHV “moan—8388 92 mABVB .wBVhV mmooxaoVBHV 8H
w 8V8 w 8V8
noon. “.8 Mmomnao V8 mi ...88 n8 MmoszVanu 9:
+8 +3~8V8 3m .8 ..AomfiVuo we
QVuowusooHn 36.83303 “Va WVHomHVoam Faafidsn ME
magi 23x31
$6188.68... 82 8888... 5.2V 8368.. m~83 . V. 638
101
f. Cr(CO)X+ reactions with l2-crown-4
55—8-51 694)—
Ions formed as products of ion-molecule reactions in a
mixture of Cr(C0)6 and lZ-crown-4 are listed below, with
their precursors as identified by double resonance.
gig stoichiometry precursor(s)
96 CrC2H4O+ Cr+
112 CrC2H402+ Cr+
128 CrC2H4O3+ Cr+
l56 CrC4H803+ Cr+, CrCO+, Cr(CO)2+
zoo CrCGH1204+ Cr+, CrCO+, Cr(CO)2+
228 CrCBH1604+ Cr+, CrCO+, Cr(CO)2+
g. Cr(C0)x+ reactions with l5-crown-5
LC4052095)-
Ions formed as products of ion-molecule reactions in a
mixture of Cr(CO)6 and lS-crown-S are listed below, with
their precursors as identified by double resonance.
mle stoichiometry precursorjs)
+ +
110 Cr(CzH202) CrCO
+ p + +
ll3 Cr(C2H502) CrCO , Cr(CO)2
+ +
126 Cr(C4H]OO) Cr
+ + +
156 Cr(C4H803) Cr , CrCO
+ + + +
200 Cr(66H]204) Cr , CrCO , Cr(C0)2
102
B. Discussion
l. Linear Ethers and Polyether
6. Reactions of Cr+
Cr+ does not react with dimethylether. Products of
Cr+ reactions with diethylether and TDE are listed in the
following.
+
cr + ‘~/0\v/_fi~
.64
Légé'Cr+(C4H100)
+ .66
Cr + ‘//O O O 0\\ Cr (CZH40) + C6H 140 3 (48)
3 15 Cr (CH 80 ) + C H O (49)
2 5 1O 2
.18
'fl Cr+ (CH 602) + C5H120 2(50)
Reactions (47)-(50) can be explained again based on
what was discussed in the case of Fe+, namely, metal ion
insertion, followed by a B-hydrogen shift. This is shown
below and in Scheme XXI.
Cr + O ——-> Cr'+/- ---> ”1-7K
\V/ \V/- \\/, H \v/ I H
l
+
Cr(C2H4O ) + c2H6
Note that Cr+(C3H802) and Cr+(C3H602) in reactions
(49) (50) are formed following insertion into the bond
between skeletal atoms 5 and 6, which is in contrast to Fe+
CY+ (C2H4 0) + C2H6 (47)
103
«on . «an
+0 z 65 . ..o 1 us
_ as.
V
. \o
..o tot "/16
up. U ...tuox
. I
a 0
...V \J . .20/Lena 1» us.
«09:66 tw no ofu I o\
V _
A . .20 (Loo...~ once «.6
\l/ I +.. _
220 0:0 _ v ‘ .10
@1161? :«uaooz _. .. V
a .+ . . o
x
.V
«:6 n w -
:3 01:0\ +5 _
/I\ _ .... 0 /\.......ol\/
6: o a: \J \J \J
/ x u Al All 36 o 0 022 + c
51 5101.: o/
6.20 as.
a MEMIum
104
which inserts into the bond between skeletal atoms 4 and 5
(p. 56). Attack of a more centrally located bond in the
polyether may be evidence for a symmetric intermediate in-
volving multiple metal-ligand interactions. This possibi-
lity is shown as the first step in Scheme XXI.
Also, the strong Cr+-0 interaction (D(Cr+-O) = 77 i 5
kcal/mole) may favor the formation of small ligands with
more oxygen's than does Fe+ - explaining the difference in
site of attack for Fe+ and Cr+. In Scheme VXI, both possible
B-H atoms can shift to get corresponding products. How-
ever, there is only one B-H shift which occurs in the case
of Fe+. The probability of such a H-shift depneds on the
stability of the final product. Both pathways in Scheme
XXI form ligands which can donate 4 electrons to the elec-
tron deficient metal. Thus, the option of which B-H's will
shift is related in part to the ability of the final pro-
duct tO be a good ligand (strong bond to Fe+).
Cr+ also reacts with TDE to form CrC2H40+ (reaction
(48)). A variety of mechanisms have been considered, how-
ever none are consistent with other mechanisms and observa-
tions. We can speculate that, to form a rearrangement
product consisting of 2 carbons and one oxygen, an insertion
into the bond between skeletal atoms 3 and 4 by Cr+ would
be necessary. However, this intermediate, has no H atoms
which are 8 - to the metal. The intermediate, written as
follows, shows that a six-membered ring intermediate gould
105
form a 4-membered ring product CrC2H40+ with subsequent loss
of CH3(OCZH4)ZOCH3.
2 / \ +
0 -——> 0 Cr + CH 0 0 OCH
CH ——0 0 OCH 3 3
\CHz—CrV 2 \ /\ / 3 \CH2/
1
cr+o o .03 <-—+ Cri"fiH
CH3
Note that the explanation above is based on Cr+ inser-
tion into a C-C bond instead of a C-O bond, which is not
usual, because Cr+-alkyl bonds are weak.
TDE, then, appears to react via the same mechanism as
diethylether. Cr+ and Fe+ apparently choose to attack dif-
ferent C-O bonds of this polyether. This choice may be in
part forced by the structure of the initial metal-ligand
complex, and further favored by the stabilities of the
final products.
b. Reactions of CrCO+
The CrCO+ reactions with dimethylether, diethylether
and TDE are summarized on the next page:
106
+ +
L +
Cr(C2H60)2
;§14.Cr(C4H]0)+ + C0
+ .005 +
CrC0+CH00(‘OCHT——->Cr(CHO)+CH0+CO(53)
31.]L/ 3 007 2 6 5 12 3
' x + .;
7 CFC3H602 + C5H1202 + C0 (54)
.013 +
.006 +
.069 +
_———+ CFCH30H + C7H1403 + C0 (57)
.170 +
.104 +
.011 +
.011 +
.003 +
CrCO+ reactions with dimethylether and diethylether only
result in ligand substitution and successive reactions. The
product in (52) is not formed by Cr+ alone. If the metal
inserted into a C-C bond, there would be fl2_B-H shift and
eliminate CH4. Thus, we conclude that the CO of CrCO+ is
actively involved in the process and inserts in toto into
107
the C-0 ether bond. Based on past observations, the neutral
product of (52) could be C2H40 or [CO and CH4].
0
0 H
+ \\\ +
CrCO + \\/1r\/,-———a \\\T”’ Cr__—.C
H “"
l
H o
H—t +Cr
H l
c
H
"—c e—a c:~(c3+150)+
I
This mechanism leads to acetaldehyde as the neutral lost
in (52).
The products in (54)-(56) were formed by Cr+ alone, and
thus have been discussed in Scheme XXI. Apparently CrCO+
extracts methanol from TDE. Cr+ alone does not do so, how-
ever the CO is not actively involved in the mechanism, since
both CrCH30H+ and CrCOCH3OH+ are products. Thus, we inter-
pret this via M+ insertion process, in which the C0 acts as
a spectator ligand as shown in Scheme XXII. Once the metal
center has induced the elimination of metanol from TDE the
Cr(CH30H)(CO)+ can use the remaining available energy to
break the weaker metal ligand bond, Cr+-C0.
The products in (60)-(62) can also be explained by
CrCO+ insertion into the central skeletal bond of TDE as
shown in Scheme XXIII. Note that both Crcoc4H802+ and
O + have a small contribution from m/e l38,
4”10 2
CrCOC3H60+, as their precursor, however, it would not be
CrCOC
108
Scheme XXII.
+
Croo+ mo 0 0 me
\_/\_/\_I
l
mac—cg— on
”\2\
H‘-CH--(KL_/;) 9MB
00
CrCHBOfi+1F—-——-CHj--O--Cf<; 'f' Vt—-O p; 9MB
. \
+
CO
(CHBOH)($'((D)
Scheme XXIII
CrGO i‘ PBQ\—J/jt_/;\L‘/DMB
mMLJLZJOJmfimMKCIJ
l K *0
"e o
+ '-. +
cL‘HBo2 + orcocl‘ 0°26— C: Cr c
He H 1'0
+ +
c53803+ arcanmo2 “(3511303)
....
C43100 2
109
expected to be reactive enough to experience successive
reactions with large molecule like TDE.*
Also note that the formation of CrCOC4H1002+ illus-
trated in Scheme XXIII occurs when CrCO+ inserts into a C-C
bond, followed by a B-H shift, and the C0 is retained on the
metal center.
Products in (59) are formed by the same mechanism as
in Fe+ case, where the metal ion inserts into a C—O bond
(skeletal atoms 4 and 5), followed by a B-H shift. Since
Cr+ does not do so, but both CrC3H602+ and CrC003H602+ are
products, the CO on Cr+ might be a spectator as described
above in the formation of CrCH30H+. The formation of
Cr(CZH60)+ in reaction (53) is again explained by CrCO+ in-
sertion into skeletal atoms 3 and 4, so that B-H is availa-
ble for shifting.
* This is based on the assumption that a large molecule such
as TDE has a steric effect which could prevent CrCOC3H50+
from approaching TDE to do any insertion, B-H shift process
1.e.
CH CH CH
3‘0.. 3\o‘ 0/ 3
' “'-c;-oo+ ms: 'ciftm + oo+c o
E; Fe... :1; .... 53123
CH2
“#3168
0113\0‘
L—x——-> (SING;- o/Cfl3
110
c. Reactions of CrLCOlX:
Higher CO-containing chromium ion reactions with
diemthylether, and diethylether only give substitution
reactions. However, it is more striking to realize that all
of the products for TDE are reactive rearrangement of TDE -
none are simple C0 displacements. This may reflect the
ability of TDE to complex stepwise with the metal center
i.e. - initial complexation may involve one oxgyen of the
ligands. As further oxygens interact, additional energy
is made available for metal-induced decomposition of the
polyether e.g.
Cr 3
CH
I l
a a
”‘31.ij ”wag—””31?
+
061-11202 + 61.638602
111
CO here also acts as an energy mediator to stabilize
the system, so that Cr+ can interact with another oxygen
without decomposing.
2. Cyclic Ethers
a. Reactions of Cr+
Cr+ doesn't react with THF except by forming an addi-
tion complex. It reacts with THP to form only one product,
Cr(C3H6O)t which can be explained analogously to Fe+
4.
.Cf
\
x’ x +
0; 7 ‘———A Cr(C3H60) + 02H4
3. Cyclic Polyethers
Cr+ + THP
a. Reactions of Cr+
Cr+ reactions iwth l,3-dioxolan, l,3-dioxane, p-dioxane
(d8), lZ-crown-4 and l5-crown-5 are summarized in Table 8.
Presumably, the products in (63), (64), (65), (66),
(67) can be explained in terms of the mechanisms which have
been proposed for Fe+ (p. 64) for the reaction of Cr+ with
l,3-dioxolan, l,3-dioxane and p-dioxane. Note that Cr+ is
even more "selective" compared with Fe+. Cr+ reactions with
l,3-dioxolan and l,4-dioxane give only one product, Cr(CH20)t
0n the hand, the reaction of Cr+ with l,3-dioxane produces
the same products as Fe+ except for Cr(C2H4)+. Two impor-
tant factors may control these product distributions:
(1) Cr+ forms strong bonds with oxygen (77 i 5 kcal/mole,
Table 8. Cr+
/\
+ o 0
Cr + \_J
0
Cr+ + [:;F
0
0
:1
0
112
Reactions With Cyclic Polyethers
O}
01
w
01
4..
Cr (CHZO) + C2H40.
4.
Cr (C3H602)
l———> Cr(CH20)(C3H602)+
Cr+(CH20) + C3H60
Cr+(C3H60) + CHZO
Cr+(C2H402) + C2H4
Cr+(C4H802)
Cr+(CH20) + C3H60
Cr+(C4H802)
Cr+C2H40 + C6H1203
Cr+C2H402 + C6H1202
Cr+CZH4O3 + C6H120
...
Cr C4H803 + C4H80
+
.1.
+
Cr (C4H100) + C6H1004
4.
Cr (C4H803) + 06H1202
4.
Cr (C6H1204) + C4H80
(63)
(64)
(65)
(66)
(67)
(68)
(69)
(70)
(71)
(72)
(73)
(74)
(75)
113
O
/
w + b
0
shift
it
L)
T
O———Cr
L.)
+ Cr(CHZO)+(——
C2H4O
Ne 2w
l
0
.T
« ‘l
h
C O
2
H
C
r
C
+
O
O
6
H
3
a c
O O
/[I\
+
+
r
C
,G-H
-—————9
ShlfT
+’0
H
Cr
+
Cr + ___>
0 0
o
cmcwof' <—— \
242 ‘—
‘ +
Cr(L2H4)
114
see (Table 2) compared with Cr+-C(H2)(R) (37 i 7 kcal/
mole), that is, it may prefer to produce products having
a high O/C ratio and bonds to oxygen atom directly. (2)
Stepwise metal ion double insertion, followed by B-H's
shift is occuring so that chormium can insert into a bond
which is in geometrical proximity, due to a weaker initial
complexation energy as a driving force (see thermodynamic
conclusion section) as shown on page 113.
The chemistry of Cr+ with l2-crown-4 differs from that
of Fe+, apparently due to the fact that, relative to Fe+,
Cr+ interacts more strongly with oxygen and less strongly
with carbon.
The first product, CrC2H40+ in (68) could be explained
using a mechanism similar to that in Scheme XII. Note that
products in (69)-(72) have ligands with a higher O:C ratio
than those formed in Table 5 with Fe+. Consider the pro-
duct CrC2H402+. This could be considered as either Cr+
complexed to Eigflflg or as a metallocyclic 5-membered ring.
The latter will be assumed because of the strong Cr+-0 bond.
These products can then be explained if the Cr+ complexes
in the crown cavity, and, because of its size, can only
interact with two oxygens at a time. Apparently no H-shifts
occur. The pathways leading to products (69), (7l), (72)
are shown in Scheme XXIV. Note that all pairwise combina-
tions of oxygens are sampled.
The product in (70) has a very high O:C ratio in the
115
SCHEME m
0 O
C J
O O /—'\
\_/ \ o o
C 3
0,. “O
\__/
fl 0 O O
O O
C Cr: 1 C 96' j V C” 1
‘~ / ’ “ ‘0 o
o. ,0 9,4,0 , l
F" ’ ‘ ’ ‘
\ + o o o o
0 Cr E j < >
L/ 0
O \ +/
+ Cr +
0 + /C'+\
( 7 C2H4 O O
116
ionic product, and may be the result of the metal ion
"puckering" up the molecule in an attempt to interact with
3 oxygens of the crown.
0 —— CH2
CH2/ \“ k \CH/0 0
I \Cr+ 2 I ”20,, \\\ +_ 0
CH2 ,I‘V¢\\ /,CH2 ‘———? l Cr-—-0 + 02H4 +
\ I, “ /CH2 H C /
0 0 2 \0
\ /
It is striking to notice that Cr+ forms an adduct ion
(addition complex) with lZ-crown-4. Again, this could be
the result of the small metal ion puckering up the molecule
in an attempt to interact with 3 or 4 oxygens of the crown
117
as some alkali metal ions do However it is hard to
interact all 1 oxygens because Cr+has a larger size than
the cavity of Iz-crown-4, (or+ has a radius of 0.8l A146’147,
0
compared with the cavity of lZ-crown-4 which is l.2 A -
1.5 A in diameter), Cr+ must be to some extent "away" from
the cavity center, which would result in a longer range
interactions (ionic rather than covalent)* with all four
oxygens and no bond breaking will occur. Fe+ has the same
tendency for interacting with both oxygen atoms and carbon.
* + 6 ‘ . . 5
Cr has a 5 ground state corresponding to an [Ar] 3d con-
figuration157. The next exaited state is 6D at l.47 V,
corresponding to [Ar] 45 3d . Despite the energy gap of
1.47 eV, it was reported that a long-lived metastable excited
state of Cr+ exists in ICR on electron impact 02
Cr(C0) 14‘ 143. It is not surprising that Cr+( S) will inter-
3
act wigh oxygen atoms electrostatically through both intrin-
sic and induced dipoles. To form covalent bonds, it has to
use an sd hybrid and Cr(5D) will form covalent bonds with
oxygens.
117
If they interact, energy is released and much chemistry
occurs, since it can only form bonds covalently. Thus Cr+
exhibits quite a different chemistry with lZ-crown-4 than
does Fe+. This can be accounted for by the greater pre-
ference of Cr+ for bonding to oxygen, which appears to be
the predominant driving force in the Cr+ reactions.
The attempt to interact with g and 4 oxygens by Cr+
is also observed in its reaction with l5-crown-5. Products
in reactions (74) and (75), namely Cr(C4H803)+
Cr(C6H]204)+ were formed by the mechanism shown in Scheme
XXV, in which Cr+ need only make an effort to pucker oxygen
atom #1, since oxygen atoms #7, 4, T3 are already on the
same plane.
The product in (73) can be explained by a mechanism
similar to that of in Scheme XIV - after forming an inter-
mediate (V) in Scheme XXV, Cr+ inserts into two C-O bonds,
followed by double B-H shifts to form Cr(C4H100)+.
Scheme XXV.
:3 ..d;.oj bund;,d + o
C. a. ..7 —-> «W2 Li
J: LJ (V)
C17) (‘o/W
{—on :1 ——> (0‘02?) 03—) “<%’*3°3) + 8
\-—’° °\__Jo
118
b. Reactions of_Cr(C0)x:
Cr(CO)x+ (x = 1-6) reactions with l,3-dioxolan, l,3-
dioxane, p-dioxane, lZ-crown-4 only result in the addition
complexes, substitution reactions and successive reactions.
Cr(C0)x+ reactions With lS-crown-S yield two new
products:
CrCO+ + lS-crown-S ——> Cr(CZH202)+ + caumo3 + co (76)
Cr(C0)x+ + lS-crown-S ——-a-Cr(C2H502)+ + C8H1503 + xCO (77)
x = 1,2
Products in (76) and (77) can be explained by a mechan-
ism similar to that shownin Scheme XIII. Formation of
Cr(C2H502)+ may be an indication that Cr+ is sequentially
inserting into C-O bonds, followed by B-H shifts to avoid
an intermediate with an unusually high formal oxidation
state for the metal. Further B-H shifts might not occur if
they are geometrically inaccessible after one B-H shifts.
Referring to Scheme XIII, no products corresponding to
double or single B-H shifts from the larger ring side are
observed. Since Cr+ has a strong bond energy with oxygen
atoms, the formation of products in reactions (76) and (77)
is accompanied by the loss of CO to carry away the extra
energy.
119
4. Thermodynamic Conclusions
Table 9 lists all neutral lost in the Cr(C0)x+ reac-
tions with all the ethers discussed above and their branch-
ing ratios (product distributions).
Dimethylether can displace up to two CO's from Cr+, as
can diethylether, implying that the initial Cr+-ether inter-
action, 66.2 kcal/mole < D(Cr+-Me20), D(Cr+-Et20) < 87.7
kcal/mole, is less than that of Fe+, 96.6 kcal/mole <
D(Fe+-Et20) < l33.3 kcal/mole, but larger than that of the
dimethylether case, 27.l7 kcal/mole < D(Fe+-Me20) < 73.l7
kcal/mole.
Cr+ can only react with diethylether by H-shift
pathway (p. , Scheme X) to get Cr(C2H4O)+. The difference
from that of Fe+ may be in part thermodynamics.
Recall,
L—-——e C2H50H + CZH4 AH l6.8 kcal/mole
Cr+ can induce the rearrangement indicated in the first
process, which requires very little energy, AH = 0.3 kcal/
mole. Fe+ induces both rearrangements. More energy is
required for the second process, AH = l6.8 kcal/mole. Thus
Fe+ forms "higher energy" products than does Cr+
However, one must be careful when thermodynamic conclu-
sions are drawn for Cr+, since electron impact on Cr(CO)6
+l4l,l48
produces excited states of Cr Nevertheless,
were
SnQB floor}
3w. v8n+no~a=ou
8N
n
.335
120
2.“...V8N . omzau
E . V8~ «of?
AHOQ 8Nw
3H; 8~ 2.3”
SN; 8~+om=aom
soda-8 532%.- 938. o< c
8.
SN.
338 + N lac
E348 +~oon=nu
So; non—2mg
938 +~o2=mu
Amc.vao~ oudmzau
a: near}
ASJB oao~fl=no
A8. v8~ .~oon=uu
2048.233}
3~¢8 «near?
A$¢8~+no~fl=mu
Go. v8~ .nomnzou
AB.V8~+no£=~.u
8N
-—--.-_-...———.__
were
$38 . op}
$.18 ~05.on
Aow. V8 .nonxmu
ANH. V8 4nomH=mu
v
-— —-———-—4
2o; 8
SA; 8 ...=~u
I‘vil
so; on;
63 Noafio
EN; ..oofoo
GN; 3
:3 as}
8
2.;
an; 8+J:Nu_
8
28; Nomzao
38; «our?
28¢8+~om=au
A§.v8.no«£uu
82¢ no}?
$38.93“?
38; «com—mu
88. V8 .Noormu
A§.V8.~o~n=no
Anoo¢8¢no§=ou
2.0; 8
an. Voéf
68.5
mm
w<
w...
8n; coxno
Ann; 9.
“no; oarnu
23;
A8; a=~u
U<
U<
8a; ~o~n=ng
$3 wooden".
So; noise;
33 .2”
cm; ozwm
“IcaDHOIflH
ox
r
u
O.
O
O
E
£23 5; 495.5 no 23308.. 9:. 5 Illa-3 3.53.. .dllloSa.
+
_
\]o\/_
\
xo
121
although the dissociation energy of dimethylether into
CHZO and CH4 is only 0.9 kcal/mole, we neither see
Cr(CH20)+, nor addition complex from Cr+. The reason for
this is not clear.
It is of interest to note that no ion of composition
Cr(TDE)+ or Fe(TDE)+ was observed. In the case of
lZ-crown-4, no ion of composition Fe(lZ-crown-4)+ was
observed. Cr(12-crown-4)+ is formed, with 3 precursors:
Cr+ + lZ-crown-4 -———9 Cr(l2-crown-4)+
CrC01 + l2-crown-4 -———9 Cr(12-crown-4)+ + C0
Cr(c0)2+ + lZ-crown-4 -———> Cr(12-crown-4)+ + zco (78)
Since Cr(CO)3+ is not a precursor to Cr(l2-crown-4)+,
one may interpret this as an indication of the Cr+-crown
bondstrengths:
D(Cr+-CO) + D(CrCO+-CO) < D(Cr+-12-crown-4) <
D(Cr+-C0) + D(CrC0+-CO) + D(Cr(C0)2+-CO)
however, this may be an incorrect interpretation. The metal-
polyether interactions should be very strong; complexation
appears to release enough energy such that subsequent frag-
mentation of the polyether always occurs, except for Cr+,
which appears to form weaker bonds to ethers than Fe+.
Rather, the reaction
122
/"‘\
O 0
Cr(C0)3+ + E: :j ———+ Cr(12-crown-4)+ + 3C0
may not occur because the carbonyl ligands prevent this
large ligand from getting close enough to the metal for
significant orbital overlap. This is verygraphically seen
as we progress from diemthylether to diethylether to TDE to
l2-crown-4. The reactions due to M(C0)n+ decrease as n
increases. Thus, reactions ofM+ alone are presumably due
to the metal in or very close to the crown-cavity. As CO's
are added to the metal, reactions occur with the metal out
of the cavity to the point where CO's prohibit metal-ligand
interactions with this bulky ligand. Thus, when 3 or more
60's are present on the metal, sufficiently close approach
of the bulky l2-crown-4 ligand is prevented and no reactions,
not even simple idsplacement is observed.
Thus, it is difficult to interpret the data here in
terms of estimating M+-polyether bondstrengths. The Cr+-
lZ-crown-4 bondstrength, based on reaction (78) must be >
66 kcal/mole.
In conclusion, Cr+ has a weaker complexation energy
with ethers than Fe+ does. Cr+ doesn't form an adduct ion
with lS-crown-S, possibly due to the large cavity of
lS-crown-S. As CO's are added to the metal, it only under-
goes substitution and successive reactions in smaller ethers.
123
In the reactions with lZ-crown-4 and l5-crown-5, however,
the interaction with these bulky ligands is prevented or
confined to interact with only a few oxygens in the crowns.
III. Ni(C0)x+ Reactions With Ethers and Polyethers
A. Results
1. Linear Ethers and Polyethers
a. Ni(C0)X+ reactions with dimethylether
1.92m).
Ions formed as products of ion-molecule reactions in a
mixture of Ni(C0)4 and dimethylether are listed below, with
their precursors as identified by double resonance.
ELE stoichiometry precursor(s)
88 Ni(CH20)+ Ni+, Nico+
104 NI(C2H60)+ Ni+, Nico+
132 Nic0(c2H60)+ Ni(C0)2+, Ni(c0)3+
150 Ni(c2H60)2+ Ni(C0)2+, Ni(c0)3+
l60 Ni(CO)2(C2H60)+ Ni(CO)3+, Ni(c0)4+
178 NiCO(C2H60)2+ Ni(C0)3+, Ni(c0)4+
190 Ni2c0(c2H60)+ Ni+, NiCO+, Ni2(c0)2+
218 Ni2(c0)2(c2H60)+ Ni+, Nico+
b. Ni(CO)X+ reactions with diethylether
L£4fl1091
Ions formed as products of ion-molecule reactions in a
mixture of Ni(CO)4 and diethylether are listed below, with
124
their precursors as identified by double resonance.
gig stoichiometry precursor(s)
86 Ni(C2H4)+ Ni+
102 Ni(c2H40)+ Ni+, NiC0+
104 Ni(C2H50H)+ Ni+, Nico+
. + .+ . + .
132 N1(C4H100) N1 , NICO , N1(C0)2+
. + +
143 N1(CO)3H czns
145 Ni(00)2(ocn3)+ 01130+
. + + +
155 N1(C0)2C3H5) “ C3H5 , C3H7O
160 NiC0(C4H100)+ Ni(C0)2+, Ni(c0)3+
188 Ni(CO)2(C4H]0)+ Ni(CO)3+, Ni(C0)4+
+ +
191 (C4H100)2(C3H7) CH30
206 Ni(C4H]00)2+ Ni(C0)2+, Ni(C0)3+,
. +
N1C0(C4H100)
218 Ni2C0(C4H100)+ Ni+, NiC0+
, + . + . +
234 N1C0(C4H]00)2 Ni(c0)4 , Nl(C0)2(C4H]00)
The ion of m/e l43 is a proton transfer product. The
ions of m/e l45, 155, l91 are products from organic ion
reactions with neutral Ni(CO)4 formed from 0CH3+ and C3H5+.
c. Ni(C0)x+ reactions with triethylene glycol
dimethylether (TDE, C8fl1894)
Ions formed as products of ion-molecule reactions in a
mixture of Ni(C0)4 and TDE are listed below, with their
precursors as identified by double resonance.
102
103
116
118
130
133
134
144
148
160
162
163
188
204
221
236
125
stoichiometry
Ni(C2H40)+
Niczhso+
Ni03H60+
Ni(03H80)+
Nic0(02H40)+
Ni(c3H702)+
Ni(C3H802)+
NiC0(C3H60)
Ni(C4H]002)+
. +
N1CO(C3H602)
)+
+
Ni(C5H]202
Ni(c4H903)+
Ni(00)2(c3H602)+
Ni200(02H40)+
Ni(c7HMo3)+
. +
N1(C7H]504)
Ni(10E)+
precursor(s)
.+
N1
+
02H30+, C3H70
Nico+
Ni+, Nico+
Nico+
Ni+, Nico+
Ni+, Nico+
Ni(CO)2+, Ni(00)3+
Ni+, Nico+
NiCO+, Ni(C0)2+, Ni(00)3+
Ni+, Nic0*, Ni(00)2+
Ni+, Nico+
Ni(C2H40)+, Ni(C0)3+,
Ni(00)4+
NiCO+, Ni(c0)2*,
Ni(CO)3+, Ni(C0)4+
NiCO+, Ni(00)2+
Ni(C0)2+, Ni(C0)3+
Ni(CO)4+
chlic Ethers and Polyethers
a. Ni(CO)x+ reactions with tetrahydrofuran
THF
Ions formed as products of ion—molecule reactions in a
mixture of Ni(C0)4 and THF are listed below, with their pre-
cursors as identified by double resonance.
ELLE
100
102
112
127
130
155
158
186
202
216
230
244
Note that c H +
stoichiometry
Ni(c3H6)+
Ni(CzH4O)+
Ni(c4H6)+
Ni00(c3H5)+
Ni(THF)+
Ni(CO)2(C3H5)+
NiCO(THF)+
Ni(00)2(THF)+
Ni(THF)2+
NiZCO(THF)+
. +
N12(THF)2
Ni2(00)2(THF)+
3 5
precursor(s)
Ni+, Nico+
Ni+, Nico+
Ni+, Nico+
+
C3”5
NiC0+, Ni(00)2+
+
C3”5
NiC0+, Ni(c0)2+
. + . +
Nl(C0)3 , Ni(00)4
NiCO+, Ni(CO)2+,
Ni(C0)3+, Nic0(TH1-')+
Ni+, NiCO+. Ni2(00)2+
. + . +
N1(C0)2 , N1(C0)3 ,
Ni(CO)4+, Ni(CO)2(THF)+
Nico
can displace up to 3 CO's from Ni(CO)4
to produce m/e 127, Ni00(c3H5)+ and m/e 155, Ni(C0)2(C3H5)+.
127
b. Ni(CO)x+ reactions with tetrahydropyran
(THP, Csfllog)
Ions formed as products of ion-molecule reactions in a
mixture of Ni(CO)4 and THP are listed below, with their pre-
cursors as identified by double resonance.
mze stoichiometry precursor(s)
. + .+
112 N1(C4H6) N1
. + +
127 N1C0(C3H5) c3115
144 Ni(THP)+ NiCO+, Ni(00)2+
. + +
155 N1(C0)2(C3H5) c3115
172 Nicomip)+ Ni(C0)2+, Ni(00)3+
205 Ni(C0)2(THP)+ NiC0+, Ni(C0)3+, Ni(00)4+
230 Ni(THP)2+ NiC0+, Ni(C0)2+. Ni(00)3+
Nicomip)+
Note that c3H5+ reacts with Ni(00)4 to form m/e 127 and
m/e l55 as was observed in the mixture containing THF.
c. Ni(C0)x+ reactions with l,3-dioxolan
I£3fl5921
Ions formed as products of ion-molecule reactions in a
mixture of Ni(CO)4 and l,3-dioxolan are listed below, with
their precursors as identified by double resonance.
m[e stoichiometry precursor(s)
86 Ni(02H4)+ Ni+
88 Ni(CH20)+ Ni+, Nico+
128
gig stoichiometry precursor(s)
102 Ni(C2H4O)+ Ni+, NiCO+
130 NiCO(C2H4O)+ c2H40+
132 Ni(C3H602)+ NiCO+, Ni(c0)2+
160 Ni00(03H602)+ Ni(CO)2+, Ni(c0)3+
188 Ni(CO)2(C3H602)+ Ni(CO)3+, Ni(c0)4+
206 Ni(C3H602)2+ Ni(CO)2+, Ni(CO)3+,
Ni00(c3H602)+
216 Ni(00)3(c3H602)+ Ni(00)3+
234 NiCO(C3H602)2+ Ni(C0)3+, Ni(00)4+
. +
N1(C0)2(C3H602)
Here, the reactive organic species is CZH40+, which
reacts with Ni(CO)4 to displace 3 CO's.
d. Ni(CO)x+ reactions with l,3-dioxane
L94fl8921
Ions formed as products of ion-molecule reactions in a
mixture of Ni(C0)4 and l,3-dioxane are listed below, with
their precursors as identified by double resonance.
m[e stoichiometry . precursors(s)
. + .+
86 N1(C2H4) N1
88 Ni(CH20)+ Ni+, Nico+
. + +
115 N1(C3H50) C3H50
. + . +
118 N1(C2H402) N1CO
. + . +
132 N1(C3H602) N1C0
129
mlg_ stoichiometry precursor(s)
145 Ni(C0)2(0CH3)+ . c2H +, OCH3+
146 Ni(C4H802)+ Ni+, NiC0+, Ni(c0)2+
160 Ni00(c3H602)+ Ni(C0)2+, Ni(c0)3+
174 Nic0(c4H802)+ Ni(C0)2+, Ni(c0)3+
188 Ni(CO)2(C3H602)+ Ni(CO)3+, Ni(c0)4+
202 Ni(c0)2(c4H802)+ Ni(c0)4+
220 Ni(C4H802)(C3H602)+ NiCO+, Ni(C0)2+,
Ni(C4H802)+
234 Ni(C4H802)+ NiCO+, Ni(C0)2+, Ni(00)3+
. + . +
N1(C4H802 , N160(C4H802)
Where m/e llS and l45 are products arising from the
H 0+ and OCH + ion-molecule reactions with netural Ni(C0)4.
c3 5 3
e. Ni(CO)x+ reactions with l,4-dioxane
12,889.22
Ions formed as products of ion-molecule reactions in a
mixture of Ni(CO)4 and l,4-dioxane and p-dioxane-d8 are
listed in Table 10, with their precursors as identified by
double resonance. Note that m/e 87 was formed by H atom
abstraction by Ni+.
f. Ni(C0)x+ reactions with 12-crown-4
fl8fll 5941
Ions formed as products of ion-molecule reactions in a
mixture of Ni(C0)4 and lZ-crown-4 are listed below, with
their precursors as identified by double resonances.
l
130
“Naming v82 .momzao
.mABXz .wBXz wwomzsoxz +3
Amomooo V82 .momnso momma
.mABXz @832 mfimomasozz RN £83: .mABVE ANomzsovaBXz 8m
“#8:: .mABXz $093383: SN mamas". .
+
momnau .mABXz .wBVE .uwomESBE 8S
.mABXz .w8zz «$933338: 8a m8: 8: n2 +A~om=suzz 9:
+82 .+2 “$28“..qu NNH +82 ma ...Aosxwuxz «3
+82 ...“: Mosnmoxz 82 +82 “2 +845sz 8
3
+A 8N8: +82 w: +32 8
mo: pg +3883. co m: +3332 cm
awquomnsoowm NM» 652230: fl. auogomu Egowngopw fl
mm ocdxotflTA ocaommih
mu 2338A 3 3:688 5:. 838.98: 88:... .3 33..
gig
100
102
103
118
146
157
174
178
202
204
206
214
216
232
234
131
stoichiometry
Ni(C2H20)+
Ni(c2H40)+
Ni(002H5)+
Ni(c2H4oz)+
Ni(c4H802)+
Nic0(c2H30)+
Ni00(c4H802)+
. +
N1(C5H]203)
. +
N12CO(C2H202)
. +
N1ZCO(C2H402)
NiCO(CSH1203)+
. +
N12(CO)2(C2H20)
. +
N12(C0)2(C2H40)
Ni(08HMo4)+
, +
N1(C8H1604)
precursor(s)
Ni+, NiCO+, Ni(c0)2+
Ni+, Nico+
+
CZHSO
Ni+, NiCO+, Ni(00)2+
NiCO+, Ni(CO)2+, Ni(00)3+
c2H30+
Ni(C0)2+, Ni(CO)3+,
Ni(c0)4+
Ni(C0)2+, Ni(00)3+
Ni+, NiC0+, Ni2(c0)2+
Ni+, NiC0+, N12(C0)2+
Ni(C0)2+, Ni(C0)3+,
Ni(00)4+
Ni+, NiCO+, Ni(C2H20)+
Ni+, NiC0+, Ni(C2H4O)+
Ni, Nico+
NiCO+, Ni(C0)2+, Ni(C0)3++
Ni(CO)4+
Note that 02H50+ is the most prominant ion in E1
spectra of all crowns
from C2
Ions formed as products of ion-molecule reactions in a
5
H 0+
149
The ions m/e l03,
and C2H3O+ reactions with netural Ni(C0)4.
g. Ni(C0)x+ reactions with lS-crown-S
LC40520951
mixture of Ni(C0)4 and lS-crown-S are listed below, with
157 are formed
132
their precursors as identified by double resonance technique.
gig stoichiometry precursor(s)
118 Ni(02H402)+ Ni+, Nico+
152 Ni(c4H803)+ Ni+, Nico+
190 Ni(CGH]203)+ Ni+, Nico+
192 Ni(C6H1403)+ Nico+
207 Ni(CGH]3O4)+ Ni+, Nico+
278 Ni(C1OH2005)+ Ni+, NiC0+, Ni(CO)2+,
Ni(CO)3+, Ni(c0)4+
B. Discussion
l. Linear Ethers and Polyethers
a. Reactions of Ni+
Ni+ reactions with dimethylether, diethylether and TDE
are summarized below:
Ni+ + /// \\ —794914 Ni(CH20)+ + CH4 (79)
LgeiieiNi(c2H60)+
Ni+ + o —~9e5-1-)»Ni(czu4)+ + C2H60 (80)
9413—>Ni(c2H40)+ + C2H6 (81)
0.20 . +
-————+ N1(C2H60) + C2H4 (82)
0.21 - +
_—--9-N1(C4H10)
133
.+ 0,6] . +
N1 Me%_JR_Ja_JPMe —v-——91N1(CZH4) + C6Hl403 (83)
0.10 - +
0.07
———+ Ni(c3H702)+ + 05111102 (85)
0.10 . +
«————> N1(C3H802) + 05111002 (86)
0.03 - +
0.02 . +
-————7N1(CSH1202) + C3H602 (88)
"9:925 Ni(c4H903)+ + 541190 (89)
Products in reaction (79)-(82) are similar to those
reported for Pet Note that Ni(CH20)+ doesn't undergo succes-
sive reaction. Also, Ni+ can directly form an addition com-
plex with diethylether. This may, perhaps, indicate that Ni+
has a weaker complexation energy with oxygen in ethers than
Fe+ or Cr+.
Products in (83), (85), (86) and (88) can be explained
similarly to that of Cr+in (48)-(50) except that no B-H
shifts in (85). It may shift but be retained on the metal
center. The product in (84) can be explained in a similar
way:
0 CH
Ni+ + TDE-———a MeQ¥_/N1//J\V/O pMe-———9 Ni+---0<:i 3 +
H CH3
0 0 OMe
\\_/\_/
The product in (87) might be formed similarly to that
134
in (83):
.... 2::
. +
0 3e/SH ;:}————9 N1(C4H]002) 1+ [::j
LE4
The product in (89) is hard to explain. One possibility
is that Ni+ inserts in a stepwide manner into c-o bonds
between skeletal atoms (5,6) and (10,11):
Me
0
Ni++Me0 o 0 0Me———>Me0 0 +
\_J\_/\_/ ‘
\\ 0
Nt\v//
0 )+ + c
9 0\]———>Ni(C4H9 3 4H90
5. Reactions of Nico+
NiCO+ reactions with diemthylether and diethylether
only result in substitution and successive reactions. NiCO+
reacts with TDE yielding some new products:
NiCO+ + 105 -w———+ NiC3H60+ + c5H1203 + co (90)
i———» NiCO(C2H4O)+ + €5H14°3 (91)
-———9NiC0(03H602)+ + 05111202 (92)
_———a Ni(c7HMo3)+ + CH40 + c0 (93)
135
)+ + CH + c0 (94)
-———9 Ni(C7H]504 3
Products in (90 and (92) follow the same mechanism as
that in (84) and (86) except that the B-H shift occurs
from the "other" side after Ni+inserts into C-0 bonds. In
both cases, CO acts as a spectator with and without con-
current 1055 of C0 in (90) and (92) respectively.
The product in (91) is 28 mass units above Ni(CzH40)+,
however, the incorporation of C0 might be an indication of
a different mechanism, because it makes a B-H shift availa-
ble for shifting after NiCO+ inserts between skeletal atoms
3 and 4. Similarly, the product in (93) can be explained
by Ni+ insertion into the c-o (skeletal atoms 10 and 11),
followed by a B-H shift. The formation of Ni(C7H]504)+ is
a high energy process, i.e. once N? inserts into the CH3-0
bond, it eliminates the «CH3 radical.
Ni(C0)x+ reactions with diemthylether, diethylether and
TDE only are substitution reactions. However, one important
result has to be mentioned here:
Ni(00)x+ + TDE-———? Ni(c0)x_2(105)+ + (x-2)CO
x = 2,3,4
Hence, the formation of the Ni(TDE)+ adduct ion is
different from what we have seen in the Fe+and Cr+ cases
where the interactions between the metal center and oxygen
136
atoms (or -CH2-) are so strong that both Fe+ and Cr+
easily induce fragmentations.
From Table 2, it is known that the bond energies of Ni+
to oxygen, a methyl group and hydrogen are small (D(Ni-O) 2
45 kcal/mole, D(Ni+-CH3) z 49 kcal/mole and D(Ni+-H) e 43
kcal/mole) so that it can form an addition complex and
randomly insert into any C-O or C-C bond in TDE as shown
in Scheme XXVI.
Also, the odd mass products corresponding to either no
B-H shift or H atom retention by the metal were observed in
the reactions of Ni+ and TDE.
The Ni+ and NiCO+ reactions with TDE can be summarized
in Scheme XXVI (the numbers on the arrow bar are the skele-
tal atoms of TDE and are used to indicate the bond into which
Ni+ inserts to yield reaction products).
Scheme XXVI.
bonds
bang:
inserted reactions
111+ + CH3OE4jq5: (0,8 28% 3’“ 4(83). (91)
3 “'5 9(8‘1).(90)
5'6 Has). <86). <92)
6'7 e (87)
748 7‘ (88)
10,11 g) (93)
11,12 >(94)
137
2. chlic Ethers
a. Ni+ reactions with THF and THP
The Ni+ reactions with THF and THP are summarized below:
Ni+ + THF '30; Ni(C4H6)+ + H20 (95)
'27: Ni(c3H6)+ + CH20 (95)
——43§-> Ni(c2H40)+ + c2H4 (97)
Ni+ + THP ; Ni(C4H6)+ + CH40 (98)
Products in (95), (96) and (98) are similar to those
observed for Fe+. However, the product in (97),
Ni(C2H40)+, is a new product, and can be explained as
follows:
+ 0 H 0\Ni+
Ni + ———+ ————_9
1 7 H
H
0
0"‘Ni+"'|( ——-9Ni+--) + CZH4
:][: CH3
Note that the enol ligand might rearrange to keto from
which is more stable.
138
b. Ni(CO)x+ reactions with THF and THP
As the number of CO's present on the metal increases,
ligand substitution becomes the predominant process.
3. chlic Polyethers
.+ .
a. N1 react1ons
Ni+ reactions with l,3-dioxolane, 1,3-dioxane, p-
dioxane (d8)’ lZ-crown-4 and 15-crown-5 are listed in Table
11.
Again, most reaction products have been observed in the
reactions of Fe+ with l,3-dioxolan, 1,3-dioxane, p-dioxane
and lZ-crown-4 except that Ni(C2H40)+ is present in the
reactions (106) and (109). Presumably, the formation of
Ni(C2H4O)+ in reactions (101), (106) and (109) follows the
same mechanism as was explained in Scheme XII. Note that
Ni+ a1so reacts with lZ-crown-4 to give Ni(C8H]4O4)+ with
an elimination of one molecule of H2 as shown in reaction
(111). This result implies that Ni+ actually interacts with
one or two oxygen atoms only, and then inserts into a C-O
bond, followed by the B-H shift.
Products in (112) and (113) are similar to that ob-
served for Fe+ as shown in Scheme XV and XVII. Products in
(114) and (115) can be explained in terms of Scheme XXVII
and XXVIII.
Table 11. N
Ni+ + 6Ao
\_J
+
N1 +6?)
.+
'l
.29
~37
k:
.20
.10
.11
.21
L———>
139
, +
N1(CH20) 1021140
+
Ni(C2H,+) + 011202
111(c2114o)+ + 01120
+
“(914) +C211402
11210311120)+ +C3H60
Ni( 0411802 )4’
——->+Ni(C2H4) +028402
._2§_, 111(01120) +c31150
L———+\+N1(021140) +c21140
N1“+ (cf-‘2) —=&-> 111(021120 )++ (161-11403
.o .
—5—)N1.(C2H,+0)++ c611
12° 3
lanficzntpz) +c611120 2
L92——9Ni(08H140“) + H2
.20
-'——91~11(0211,+02)+ *°8"16°3
.12
——-+N1(c,+H803)* + 0611120 2
0?
~N1(C6H1203): + 0411802
ieui(clouz:o5)*
Reactions With Cyclic Polyethers
(99)
(100)
(101)
(102)
(103)
(104)
(105)
(106)
(107)
(108)
(109)
(110)
(111)
(IR)
(113)
(114)
(115)
(116)
140
Scheme XXVII
H
+ 0
\N23 ——)Ni(C6l-11203) + (0:)
H
1_1 " +
2 B -H CONN? + C4H602
L)"
C2115
Scheme XXVIII
The formation of the adduct ion, Ni(C10H2005)+ might
indicate that Ni+ interacts with one or two oxygens in the
crown and since their interactions are not so strong, frag-
mentation does not occur.
b. NiCO+ reactions
Nico+ reactions with l,3-dioxolan, l,3-dioxane, p-
dioxane, 12-crown-4 and 15-crown-5 are summarized below:
/\
+ 0 0
NiCO + \_j
Nico+ + (:3) ——1
141
.7___.-14 N1(CH20)+ + c2H40 + c0
.67 . +
_____, N1(C3H602) + co
413—» Ni(CzH40)+ + c1120 + co
499—) Ni(CH20)+ + C3H60 + co
.23 . +
.20 . +
___—’N1(C3H602) + CH2 + CO (118)
0
NiCO+ + [j ——
0
24.81., Ni(C4H802)+ + co
r31.. Ni(CH20)+ + C3H60 + co
Jae Ni(C2H40)+ + C2H40 + co
499—» Ni(C2H402)+ + c211 + co
4
[—1
NiCO++ ESL—PO) n
425—) Ni (C4H802)+ + co
fig-e Ni(c21120)+ + C5H14°3 + co
.05 - +
‘_—_9 N1(C2H40) + C6H1203 + C0
.36 - 4'
-—_—>N1(CZH402) + C6H1202 + C0
014) ' +
/ N1(C4H802) + C4H802 + C0 (119)
ll
. . +
~———a N1(CBH1404) + H2 + C0
.02 . +
—9N1(08H1604) + C0 (120)
142
0 + co
o/W
- + Ni(C5H]203)+ + (x-l) Co +
C H 02 x = 2,3
4 6
Ni(C0)x+ + 12-crown-4 ———> NiCO(C5H]203)+ + (x-2)co +
C H 0 x - 2,3,4
4 6 2
These two products can be understood in terms of the
structure as shown in Scheme XII (p. 74), in which the Ni+
will interact with oxygen atom #1 and carbon atom #5 as
shown below:
-co 0\\’u+ \ ZB-H \
° O/NL JO shift7
4/ c
H
0
1‘
9
< :.+
N1“\ + C4H602
0 ‘0
\ / \CH3
It is of interest to note that both 12-crown-4 and
lS-crown-S can displace up to 4C0's from Ni+. If the reac-
tions are induced from the cavity center, C0 ligands on Ni+
will prevent it from getting close to all of the crown's
oxygen atoms.
144
4. Thermodynamic Conclusions
Table 12 lists all neutrals lost in the Ni(CO)x+ reac-
tions with all ethers discussed above and their branching
ratios.
Dimethylether can displace one CD from Ni+, and diethyl-
ether can displace two CO's from Ni+, suggesting that 6.58
kcal/mole < D(Ni+-Me20) < 71.17 kcal/mole. Similarly, for
diethylether, 71.17 kcal/mole < D(Ni+-Et20) < 101.74 kcal/
mole.
In comparison with the complexation energy of Fe+, Ni+
has a smaller complexation energy. Thus, we did not observe
Ni(C4H80)+ as a product in the reaction of Ni+ with diethyl-
ether, since less energy is available. Such an analysis in
the reactions with cyclic ethers and polyethers may not be
useful, since orbital comparability, orientation in space and
many other factors are involved. Both THF and THP can dis-
place up two CO's from Fe+ and Ni+, but only one product is
observed in the reaction of Ni+ with THP, (which produced
three products in the case of Fe+). In contrast, Ni+ forms
three products in its reaction with THF, (which also forms
three with Fe+). The difference is that although both com-
pounds can displace up to 2C0's from both Fe+ and Ni+, the
bare Ni+ does not contribute to the formation of the addi-
tion complex. This in turn relates to the bonding of
ligands to the metal center and the orbitals used by the
metal center, etc.
145
21:5: 30¢ 8...
AR¢8n+o£nu :3 8~+~5 28.33 + 0:6
8.: A$.v§+~OD—_30 8N A0.v 8N 8N 8N 8N as. v8N¢NONH=h0 8N ”RSV“:
“magmatic “WW.” . ...s
3M¢Snmmwunaw A84 830 9. A8; 8+~o§¢no
Ao~.v8~.~om:..u 8” A8; SN 8H 8a 8a Ro¢8~mofiunu 8”
Bo fin We”? E Q¢8~ NB 8~ 8~ 8~ €1v8~Mo~fno 8~ 8 mA8v=
3.38 + osuru Go; 8~
GH¢8~ + as} G”; 8~ + mo
:38 . use...” :3 SN . of
69x3 2%.? 2K; 8 “no; 8~+~Oo=no
A-¢8~+~ofi=ou 8 A8; 8 n5 8 8 8 28¢ 8+~ofiuno 8
8~ AS.V8~Mo£=ou 8“ AS; 8~ 8N 8N 8~ A8; 8+noflnnu 8N 8 mgr:
8a; 8 + ...u
“so; 8 + can.
33 8 + oozau
A8; 8 Keogh".
3o; ~o~fl=no
8.38 A8; 8 A8; 82%.?
:58 + Raid; 8 w: RH; 8+~oon=m°
3:8 +~oo=¢2adv8 +~om=ao Tn“; 8 8.3 8 AR; 8 3o; Shear}
SJB .~%=JAR.V8+~oN£ou “8; 8.3} 8: 836 Co; 8 $3 83$ 3”; noise... an; 8
A8.V8+~o~a=w%no.v8.no~n=ou “8.863%29; 8.33 A2; 8+o~6 A8; 83mg 88; 8+noormo 2a; 8+a=~u So; 8
A8.v8.now§3n.v8.nosroo f~¢8soonn2$v86¢=nu $fi86§~u 8 2b; 8:30 Go; 8+nofi=uo 8.3 8+oz~u an; 8.35 p82
33 lawman
A8; Nowano
in; 2 3.0.; 2 Re; «amuse
2.0; or} A8; m: 3H; fun 8“; ~oon=mu A3; 3
Co; ~om=cu 8.0.; «03.00 83 oazmu RN; 2 3...; ONE 33...} Cb; ~03} 8w; 3%...
3.3 No}? 20.10300 83 one: 33 as} 2.3 as? 33 i 33 “02.x". :3 new :3 3
SN; no}? 3“; no}? A8; 9.3.9.. AR; was} 58¢ Nora 0:5 2.83an So; no}? 2...; cons... AS; 30 +2.
a n
n ncxaonnn 41.59.84." 0 O Q Q G Bu. )0 \/ o \5
.933.
285» 5:. .88: 8 28:32. .5. 5 :3 $.53.
+
146
In the case of polyethers, the thermodynamics conclu-
sions are more difficult to make. The rough estimation of
complexation energy for Ni+ with TDE, 12-crown-4 and 15-
crown-S will be greater than 126.32 kcal/mole. It is unclear
how many M+-O interactions this figure reflects.
Obviously, many more experiments have to be conducted
for an understanding of metal ion-polyether interactions.
Metal ion reactions with multifunctional molecules are some
what an interface between molecular and bulk interactions.
Although the explanation of product ions is somewhat
speculative, this is the first attemp to carry out this kind
of study. Other attempts still have to be tried on smaller
molecules to determine other factors controlling the forma-
tion and distribution of products. Section 5 is an attempt
to do this.
+ + . +
x , Cr(CO)x And N1(C(fix In
Their Reactions With Ethers.
IV. Comparison Of Fe(CO)
Table 13 summarizes the ether reactions for Cr+, Fe+
and Ni+, (excluding successive reactions and complexation
reactions). By consulting Table 6, Table 9 and Table 12, it
is readily seen that, as the number of 00's on the metal
center increases, the "reactivity" decreases and only sub-
stitution reactions are observed.
In the reaction of Fe+, a clear macrocyclic effect is
observed, namely the number of products changes dramatically
147
305603 cogmxoamsoo and 30:80am :33de gamuooofim 3305 won moon *
erJix we no .2 mm .8 2 .....H .8 2 we no 2 mm no n+2
§§ 0 \V\ o § o \Q 0 “Q o WAS?
& o o R o o I& H o & o o I\\A o o mA8V=
o o o o o o n a o o o o o o 0 £8?
. o o o o o o n n N o o o o o 0 was:
0 o o o o o w w 9 o H o o o o wsvz
o o H m m o HH m 0H m n H H o 0 +82
H m H m m o H. H m m nL-...H1.iH 2H rm; +2
0 g a \/o \/ “no [ammo . :oH pagodmm
*éowpmcflynoou mo monopm «:03; 5 «H350
Hot: no... ugsflo 388$ 833% mo nBssz .mH «Ea.
Table 13. ( cont'd)
148
15-crown-5
c) §§§§S§:§.
Mm
COO
12 -cmwn-li
NE§§S§\\V§
3:\
CrF/e///N/iCrF/e//N/1/
s \\\\\\\
~.\\\\\\\\)‘
.\\\\\\\\\
o s
MT-Cr Fe N/iCrFe NICrF/e Ni
Reactant Ion
AA
VV
AAA
VVV
149
from linear polyethers to cyclic polyethers. THF and THP do
not produce as many products as diethylether, due to their
geometrical constrants, possibly leading to insufficient
orbital overlap. Hence, Fe(CO)2+ is still reactive in the
reaction with diethylether and unreactive in its reaction
with THF and THP. However, in its reactions with polycyclic
ethers, Fe+ starts to interact with multifunctional atoms
to induce more products from the cavity center, exhibiting
macrocyclic effect with lZ-crown-4. When the ring size
increases, this effect is decreased slightly and we expect
that the reactions may be more like those observed linear
polyethers in the reactions with lB-crown-6 and 21-crown-7
with Fe+ inducing products more randomly, presumably due to
its incapability of interacting with all functional atoms
at a time. .It is of interest to note that in the reactions
with small polycyclic ethers, the Fe(CO)x+ ions (x 3,2) are
unreactive, possibly implying that geometrical restrictions
are present, although the initial interaction could be
strong. Moreover, in the 12-crown-4 and lS-crown-S systems,
the addition complexes were not observed. Also, no substitu-
tion reactions occurred, suggesting that Fe+ interacts with
all oxygen atoms in the former and 3 or 4 oxygen atoms in
the latter and the interaction must be very strong to under-
go prompt fragmentations. It also implies that Fe+ induces
reactions from the cavity center or very close to it. Con-
sequently, as more CO's are present on Fe+, it can no longer
150
get close to the cavity center of lZ-crown-4. Note that we
also see some ligand effects, in which CO can act as either
a spectator or a participator to give new products which
are not seen in the ligand-free metal ion's case, eSpecial-
ly in the TDE reaction.
The above model can also be applied to Cr+ reactions
with the knowledge that Cr+ prefers to retain bonds to
oxygen. Note that the formation of Cr(12-crown-4)+ may not
be the result of the interaction of Cr+ with 4 oxygen atoms
from the cavity center, because it will result in fragmenta-
tion, due to its strong interaction with oxygen atoms.
Instead, it could be the result of interaction with two
oxygen atoms as in p-dioxane. Again, we have no evidence
that Cr+ could interact with all 5 oxygen atoms in lS-crown-
5. The failure of Cr+ to form an addition complex with
lS-crown-S could be due to the insufficient oribtal over-
laps. Also note that CO ligands play an important role in
the reaction of this metal with TDE (see Table 9).
Since Ni+ has relatively weak bond energies to oxygen
and alkyl groups, the interaction between it and ether
oxygen atoms is weak. Thus, we don't see the macrocyclic
effect in this case. From Table 12, it is seen that NiCO+
produces more products than Ni+. Ni+ "randomly" inter-
acts with the oxygen atoms in TDE to give many products.
However, when the reactant cyclic polyether is lZ-crown-4,
it only produces 3 small molecular products, with another
151
product having one H2 elimination. It appears that Ni+
only interacts with orbitals in close proximity, possibly
implying that the smaller size d-orbitals are used. Forma-
tion of addition complex of Ni+ with TDE, 12-crown-4 and
lS-crown-S may indicate that Ni+ is unable to interact
strongly with many oxygen atoms because of its low bonding
energy and that Ni+ can't efficiently overlap with orbitals
wich are far from it.
Another way to look at the product distributions is
from Staley's bond dissociation energy studies for two
ligands in the gas phase158'162. Staley found that metal
ions are softer (based on HSAB theory) across the Periodic
Table in the following order:
H+, Li+, A1+ > Mn+ > Fear+ > Co+ = CpNi+ > N0+ > Ni+ > Cu+
They also found that the bonding distance of interact-
ing center of ligands to the metal ions is increasing in
the following order:
+ + +
+ < NO+ < A1 < Ni < Mn < L1+ < CpNi+
H
Hence, we expect that Cr+ prefers to retain harder
acids such as oxygen acids with higher O/C ratios than Fe+
does. Ni+ prefers to retain softer acids with lower 0/C
ratio or simply alkenes. On the other hand, the shorter
bonding distance will reflect a larger tendency to retain
larger ligands and show a greater substituent effect.
152
Unfortunately, the experimental data is not sufficient yet
to be used here.
5. Trends in First Row Transition Metal Ions In Gas
Phase Reactions With Organic Molecules
In this work, Co(C)3NO, Cr(CO)6, Ni(CO)4) Mo(C0)6 and
W(CO)6 were obtained from Alfa products. Fe(CO)5 and cis-
2-pentene were obtained from Aldrich Chemical Company.
l-hexene, 2-pentanone and sec-butylamine were obtained from
Chem Service. Iso-propyl chloride and propane gases were
obtained from Matheson Gas Products Inc. Methyl iodide was
obtained from MCB manufacturing Chemical Co., Inc. All
compounds were used without further purification except
cis-Z-pentene which was distilled for 8 hours before use.
All compounds were subject to standard freeze-pump-thaw
cycles before use.
A number of papers have appeared in the literature on
the chemistry of metal ions with organic molecules. Most
of these papers have discussed one metal. No attempt has
been made to discuss the reasons why different first row
transition metals behave so differently with simple organic
molecules. The decision was made to "target" some organic
molecules whose reactivity may provide insights into the
differences of metal ions. A number of factors which
contribute to the chemistry observed in gas phase organomet-
tallic reactions should include:
153
l. The number of available low lying empty orbitals
on the metal ions available for reaction. (For example,
Hg+ (51d10) can have two sp hydrid orbitals but Li(sz)
cannot.)
2. Orientation and size of available orbitals of metal
ions.
3. "Compatibility" of orbitals. For an insertion pro-
cess to occur, bond lengths and bond angles in the initial
metal-ligand complex must be compatible.
4. Thermodynamics (e.g., heats of formation, bond
strength, promotion energy of metal ions (redistribution of
electronic configuration)).
5. Orbital symmetry of the intermediates.
6. Electronic states of metal ions participating
the reactions.
7. Kinetic factor (to be detected in ICR, the rate
9 '11 cm3/molecule/s).
constant has to be in 10' - 10
The initial strategy in this work was as follows:
a. Test if a metal ion inserts into the CH3-I bond.
b. If a metal ion inserts, then test if there is
B-H shift across the metal center by studying
reactions with i-C3H7Cl.
c. Test if 5 or 6 membered ring intermediates are
preferred by investigating metal ion/molecule
reactions with l-hexene, cis-Z—pentene and
2-pentanone.
154
d. Test if the metal ions react with nonpolar
compounds such as propane, to estimate
D(M+-alkyl).
e. Test if the metal ions react with amines to
determine the bond strength of the M+-NR2 bond.
These experiments were performed with a number of 1;;
row transition metal ions.
1. Reactions With Propane (C3fl8)
A. Results
1. Fe(CO)x+ reactions With propane
Ions formed as products of ion-molecule reactions in a
mixture of Fe(CO)5 and C3H8 are listed below, with their
precursors as identified by double resonance.
gig stoichiometry preCursor(s)
+ +
84 Fe(C2H4) Fe
98 Fe(C3H5)+ Fe+, Feco+
+ +
100 Fe(C3H8) FeCO
128 Fec0(c3H8)+ Fe(CO)2+
153 Fe(c0)2c3H5+ c3115+
+ + + +
169 Fe(CO)4H c2114 , c2115 , C3H8
and Fe(c0)3c2H5+
+ + + +
181 Fe(CO)3C3H5 c2115 , C3H5 , C3H7
+ + +
183 Fe(CO)3C3H7 c2115 , c3117
+ + +
197 Fe(CO)5H CZHS , c3117
and Fe(CO)4C2H5+
155
Note that CZHS+ , C3H5+ and C3H7 are reactive organic
+168
species, especially C3H5 , which reacts with netural
Fe(CO)5, displacing two or three CD'S.
2. Co(CO)x(NO)y+ reactions with propane.
Ions formed as products of ion-molecule reactions in a
mixture of Co(CO)3NO and C3H8 are listed below, with their
precursors as identified by double resonance.
mi; stoichiometry .precursor(s)
87 Co(C2H4)+ Co+
101 Co(c3H6)+ 00+, Coco+
103 Co(C3H8)+ Coco+
118 CoNOC2H5 CZHS:
13o CoNOC3H5 c3115
131 Coc0(c3H8)+ Co(c0)2+
133 CoN0(c3H8)+ Coho+
145 Co(CO)2NOH+ C2H5: , Coco+
158 CoCONOC3H5+ C3H5:
16O CoCONOC3H7+ C3H7:
174 Co(CO)3NOH+ c2”5: , Coco+
188 Co(c012Noc3H7+ C3H7
3. Ni(CO)x+ reactionsfiwith propane
Ions formed as products of ion-molecule reactions in a
mixture of Ni(CO)4 and C3H8 are listed below, with their
precursors as identified by double resonance.
ng
86
100
102
115
127
130
143
145
155
157
159
186
156
stoichiometry
. +
N1(C2H4)
. +
N1(C3H6)
. +
N1C3H8)
Ni(CO)2H+
+
5
N1c0(c3H8)+
Ni(c3H8)(c3H5)+
Ni(CO)3H+
Ni(c0)2c3H5+
N1(c0)2c3H7+
N1c0(c3H8(c2H5)+
N1(c0)3c3H8+
NiC0C3H
_precursor(s)
Ni+
Ni+, Nico+
Nico+
c2”5+
3”s+
Ni(CO)2+, c3118+
c H * Ni+, Nico+
C
3 8 ’
CZH 1, N11, Nico+
+
c3”5
+ + . +
c3115 , C3H , N1CO(C3H8)
+ . +
C3H7 , N1CO(C3H8)
N1+
The ion at m/e 100 has a very small peak intensity com-
pared with that of the comparable product for Co+, but it is
similar to that observed for Fe+.
Cr(CO)x+ reactions with propane
Ions formed as products of ion-molecule reactions in
a mixture of Cr(C0)6 and C3H8 are listed below, with their
precursors as identified by double resonance.
mZe
80
96
124
stoichiometry
Cr(c2H4)+
Cr(C3H8)+
Crc0(c3H8)+
precursor(s)
Cr+, orco+
orco+
CrCO+
157
5. Mo(CO)X+ reactions with propane.
Ions formed as products of ion-molecule reactions in a
mixture of Mo(CO)6 and C3H8 are listed below, with their
precursors as identified by double resonance.
gig stoichiometry (precursor(s)
162 Moc0(c3H6)+ MoCO+, Mo(00)2+
19o Mo(CO)2C 3 H6+ Mo(CO)2+, Ho(c0)3+
218 No(CO)3C 3 H6+ Mo(CO)3+
8. Discussion
Table 14 summarizes the reactions with propane for all
five metal ions together with product distributions.
All the product ions seen can be explained in terms of
the mechanism pr0posed by Beauchamp94'96.
M+ +,/”\\-—r———>11—1+—<<5-—————+ “\‘Mi—l ———%>M-—
shift H//M U\
'————>H——M+
jiji///zH M = Fe, Co, Ni
5 shift
- e
shift
"\Mt—H—ent-H
:_//’
""9 “‘3‘“ W /
CH3
M Fe, Co, Ni,
Cr
158
No 9“. a.”
n. Hm. 3. Ha.
0H a." H m. oH “we 0;” “5.
n. N H m H a w H ,N H "W
.8 Hz 8 ea
8. 3. HN.
o.H om. Hm. mu.
.5 Hz 8 9m
Ammmuvozou Tll m=mo + .38
+
H: + wamovaoowzil
H: + 8 + HmzanHiBwsfl
8w + AmznovmthSwzj
8+ Amenszflswfl
38:
N: + momma:
$8 + kamHuEUI mama + +2
Hanna 2888 5:. «5308: . 8H 838.
159
Thus, Fe+, Co+ and Ni+ exhibit very similar reactivity
164. It is of interest to note that Fe+, Ni+
with alkanes
and Cr+ have stronger preference to inserting into C-C bonds
to eliminate smaller alkanes. Mo+, never induced any reac-
tions from propane unless it has ligands attached. CO
ligands on Mo+ might affect the bonding abilities of other
bonding orbitals. It is not surprising then since Mo+ is
a second row transition metal, which has larger d orbitals
(60% size of 55 orbital) to make the bonds using the 4 dz2
orbital, which in turn makes the second bond possible by use
of other d orbitals (the promotion energy form 4dn to 55]-
4dn"l to make a 55 orbital available for bonding123 is
about 72 kcal/mole). This contrasts with first row transi-
tion metals which bond almost exclusively using the 4s
orbital. The 3d orbitals are only 30% the size of the 4s
orbital and are tightly bound to the nucleus. (It is pos-
sible to mix 45 with 4p to make sp hybrid orbitals).
II. Reactions With Iodomethane (CH3I)
A. Results
1. Fe(CO)x+ reactions with CH3;
Ions formed as products of ion-molecule reactions in a
mixture of Fe(CO)5 and CH3I are listed below, with their
precursors as identified by double resonance.
160
m/e stoichiometry precursor(s)
183 FeI+ Fe+
198 FeCH3I+ Feco+
226 FeCOCH3I+ Fe+, FeCO+, Fe(00)2+
+ + + +
254 Fe(CO)2CH3I Fe(CO)2 , I , Fe(CO)3 ,
+ +
CH3I , Fe(c0)4
282 Fe(CO)3CH3I+ Fe+, Fe(CO)3+, CH3I+,
Fe(CO)4+
2. Co(CO)x(NO)y+ reactions with CH3;
Ions formed as products of ion-molecule reactions in a
mixture of Co(CO)3NO and CH3Iareelisted below, with their
precursors as identified by double resonance.
ELE stoichiometry precursor(s)
186 CoI+ co+
201 CoCH3I+ Coco+
229 CoCOCH3I+ CoCO+, Co(c0)2+
231 CoNOCH3I+ CoCO+, CoCONO+, Co(CO)2NO+
259 CoCONOCH31+ CH31+, Co(CO)2NO+,
Co(c0)3Ho+
287 Co(CO)2NOCH31+ Co(CO)3NO+
3. Ni(CO)x+ reactions with CH3;
Ions formed as products of ion-molecule reactions in a
mixture of Ni(CO)4 and CH3I are listed below, with their
precursors as identified by double resonance.
161
gig stoichiometry precursor(s)
185 Nil+ Ni+, Nico+
200 Ni5H3I+ Ni+, NiCO+, Ni(c0)2+
213 Nic01+ NiCO+, 1+
228 NiCOCH3I+ NiCO+, Ni(CO)2+, Ni(50)3+
255 Ni(CO)ZCH3I+ Ni(CO)2+, 1+, Ni(CO)3+,
Ni(50)4+
284 Hi(50)35H3I+ Ni(CO)3+, Ni(CO)4+
4. Cr(CO)x+ reactions with CH3;
Ions formed as products of ion-molecule reactions in a
mixture of Cr(CO)6 and CH3I are listed below, with their pre-
cursors as identified by double resonance.
m/e stoichiometry, precursor(s)
179 CrI+ Cr+
194 CrCH31+ CrCO+, Cr(CO)2+, 5H31+
5. Mo(50)x+ reactions with CH3;
Ions formed as products of ion-molecule reactions in a
mixture of Mo(CO)6 and CH3I are listed below, with their
precursors as identified by double resonance.
162
mle stoichiometry precursor(s)
219 MoI+ Ho+
234 MoICH3+ Mo+, MoCO+, 1+, Mo(50)3+
5. A1(5H3)_.,+ reactions with CH3;
Ions formed as products of ion-molecule reactions in a
mixture of A1(CH3)3 and CH3I are listed below, with their
precursors as identified by double resonance.
m/e stoichiometry precursor(s)
+ + +
AlMe3 , CHBI
*
199 AlIMe3
* Me denotes a methyl group
163
B. Discussion
Table 15 summarizes all the reactions with iodomethane
for all metal ions in this owrk, together with their product
distributions.
It appears that all metal ions insert into the polar
bond (C-I in this case), followed by fragmentation to retain
the iodine atom. Among them, Ni+ seems to have the strong-
est interaction energy with CH3I to induce more reactions.
For example, Ni+ is the only metal ion that produced MCOI+
and MI+ from NiCO+. Mo+ follows almost the same reaction
pattern as that of Cr+.
There are two possible mechanisms leading to the forma-
tion of Mi+:
(1)197,198
M+ + RI ——> R-M+-I ——>MI+
(2)85,98,169'170
X H
M+ + IM'I.
\ + e
x H ———-—>MX + d
J. Allison and D.P. Ridge197’198
reported a number of
first row transition metal ions reactions with alkyl halides
and found the following results:
164
$6 + MH£H 56.5 kcal/mole)
than Mo+.
Therefore, it seems that thermodynamics only can explain
all the reactions presented to this point. The mechanism
can be summarized below:
H
I
M+ + >_c1 __, >_Mt_ c1 ——>J--M+——Cl HJH-H” + HCl
l \\\\3 M+=Mo M+=Fe+,Co+,Ni+
+ +
M(C3H6) MCl Cr+,MO+
M+=Fe,Co,Ni,Cr M+=Cr ' '
The formation of MC3H7+ for Fe+, 00+, Ni+ and Cr+
implies that 0(M-53H7)+ > 82.03 kcal/mole.
168
o. H... .Hou :8. .. +2... ...»...HozHu...
\HHS...
o .. N “HquHuvnTHHquHSouxT way...
0.. 0.. 8.. ...N Essay M98...
0 .. H+HoH .. HN.:chH 8 .....Ho... 1.8...
u UCO‘UUHUK U’dQOUUU-um
HH .... +HH HHuHsHBHesHSczou
...... ..HH .HHHcHzHuzcsH $9.8 HHHquHo +HkuH=Huczou
H... 8 + H8 HxHu.ozHux.8.o.
HH. 8.. + 8 + HmzHSczHuchSoo mszcSou
HH. ....H.. + MuszcuvzeJ
H... e.H NH. 8N .. MH8H..HH.NL.H8.T
..H. .3. ..H. ..H. NH. NH. 8 + «HuHxHuvHuxH8.:1l
.... HH. ..H. .... .... 8.. .+ 3.6.5821
HH. 8. NH. HN. HN. H... H..H H... ..H. 8 + 8.. .. «HzHuvHuwHouvkaHSw:
N H .H H N H N H N H N H H:
NH. monHurTl
HH. H... HH. HH. .8 + H..H ......I
..N. H... HaHu + 86.11
HH. H... H... 8. H... 8.. +HH=Huwzil
o: mum. aHz a a .II. .H :5 + 55118:. H": ++x
AHqunu. ucHuoH:uHHmcumoo~ :NHa acoHuuuom
.Ofl 0.23.
169
IV. Reactions w1th Cis-Z-Pentene (C5H10)
A. Results
+ . . .
l. Fe(CO)x react1ons w1th c1s-2-Csfl1O
Ions formed as products of ion-molecule reactions in a
mixture of Fe(CO)5 and (25H10 are listed below, with their
precursors as identified by double resonance.
ml£_ stoichiometry precursor(s)
95 Fe(C3H4)+ Fe+, Feco+
98 Fe(c3H6)+ Fe+, FeCO+
110 Fe(C4H6)+ Fe+, Feco+
124 Fe(csHB)+ Fe+, Feco+
125 Fe(c5Hm)+ I FeCO+, Fe(50)2+
154 Fe50(55Hm)+ FeCO+, Fe(CO)2+,
. Fe(CO)3+
182 Fe(CO)2(C5H10)+ C3H6+, Fe, Fe(5013+
and Fe(50)353H6+ Fe(CO)4+
210 Fe(CO)3(C5H10)+ Fe(CO)4+, Fe(CO)5+
238 Fe(CO)2(C5H]0)+ Fe+, FeC0+, Fe2(CO)4+
Note that m/e 182 is a mixture of Fe(CO)2(CSH10)+ and
+
Fe(CO)3(C3H6) .
+ . . .
2. C0(CO)X(NO)y react1ons w1th c1s-2-Csfl10
Ions formed as products of ion-molecule reactions in a
mixture of Co(C0)3NO and 05H10 are listed below, with their
precursors as identified by double resonance.
170
m/e stoichiometry precursor(s)
99 Co(C3H4)+ 5o1, CSHIJ , Coco1
101 5o(c3H6)1 5o1, Coco1
113 C0(C4H6)+ 5o1, Coco1
127 C0(C5H8)+ 5o1, Coco1
129 Co(C5H]0)+ Co1, Coco1, 5o(50)21
130 CoN053H51 C3H5: , Coc01, Co(50)21
155 5o50N053H51 c3H31, c3HJ
157 CoCOCSHlJ Coco1, 5o(50)21
158 CoCON0C3H J c3HJ , Co(CO)2NO+
159 C°N0C5H10+ 5o50N01, 5o(50)2N01
187 C0C0N0C5H10+ c3HJ , 05H101, 5o(50)2N01,
Co(50)3No1
195 Co(C5H8)2+ Coco1, Co(50)21
199 Co(C5H]0)2+ Coco1, Co(50)21
215 Co(CO)2NOCSH10+ CSHIJ
215 C02CO(C5H10)+ Co1, Co250N01
218 5o250N0(c3H6)1 0o1, 5o(c3H6)1
229 CoNO(C5H]0)2+ Co1, C5H10+, Coco1,
Co(50)2N01, Co(50)3H01
230 Co(c4H6)(5o50N0)1 5o1, Coco1, 5o(c4H5)1
244 Co250N0(55H8)1 Co1, Coco1, Co2(50)2N01
245 C02C0N0(C5H]0)+ 5o1, Coco1
Note that 03H5+, C3H3+ and C3H 6+ from organic fragmen-
tations are reactive with neutral Co(CO)3NO to form m/e 130,
l7l
l56, 158 and m/e l87.
3. Ni(C0)x+ reactions with cis-2-pentene
Ions formed as products of ion-molecule reactions in a
mixture of Ni(C0)4 and C5H10 are listed below, with their
precursors as identified by double resonance.
m1; stoichiometry precursorLs)
98 Ni(c3H4)1 N11
100 N1(c3H6)1 N11, N1c01
112 N1(c4H6)1 N11, Nico1
l26 Ni(C5H8)+ N11, Nico1
128 Ni(C5H]0)+ N11, Nico1, Ni(c0)21
155 NiCO(C5H]0)+ Ni(C0)2+, N1(c0)31
17o Ni(c0)41 05H10+
184 Ni(CO)2(C5H]0)+ C5H10+, Ni(C0)3, Ni(c0)41
198 Ni(C5H]0)2+ Ni(80)21, Ni(c0)31
212 N12C0(C5H8)+ N11, Nico1, N12(c0)21
Note that molecular ion of cis-Z-pentene reacts by
charge exchange with neutral Ni(C0)4 to give m/e l70,
N1(c0)41.
4. Cr(C0)x+ reactions with cis-2-pentene
Ions formed as products of ion-molecule reactions in a
mixture of Cr(CO)6 and CSH10 are listed below, with their
precursors as identified by double resonance.
172
gig stoichiometry precursorfsl
+ +
120 Cr(C5H8) orco1, Cr(C0)2
+ + +
l22 Cr(C5H10) CrCO , Cr(C0)2
+ + +
150 CrC0(C5H]O) Cr(C0)2 , Cr(CO)3
5. Mo(Cle+ reactions with cis-Z-pentene
Ions formed as products of ion-moleCule reactions in a
mixture of Mo(CO)6 and C5H10 are listed below, with their
precursors as identified by double resonance.
gig stoichiometry precursor(s)
160 Mo(csH8)1 M01, Moco1
l62 Mo(C5H]0)+ M01, Hoco1, Ho(c0)21
188 MoCO(c5H8)1 Mo(CO)2+
19o MoC0(C5H10)+ Mo(c0)21
215 Mo(C0)2(C5H8)+ Ho(c0)31
B. Discussion
Table 17 summarizes all the reactions of cis-Z-pentene
with metal and metal-containing ions in this work.
It is interesting to note that Cr+ does not react with
cis-Z-pentene; Mo+ only eliminates H2. When CO ligand(s)
are attached, an elimination of H2 is seen for both metal
ions. It is also interesting to note that Cr+ reacts with
C3H8 to eliminate CH4. There thus appears to be a chain
l7l
length effect or the double bond in alkenes prevents Cr+
from inserting into C-C bonds. From Table l4 and l8, it can
173
0..
+Ao£nuv8~8 Tll+28~8¢l| .8
8. NziymovBBNoo .88.
8 + AoannuvBBNoo
0..
+B~A8v~8 1.8
mgr.
8 .wonxnorTI.AomMoV§TlmA8r
~=++Aw=m38~z+ll MABVFHR! .
tnowMSNABFiI “:85 +2
. 33303.» 253825
ma.
oz" 3.
wa.
an.
8. no.
no.
MN.
ea.
c..." R.
MN.
0...“
5.
ma .
ma.
ow.
no.
a.
flu.
ow.
ON.
3.
fi.
no.
8m Janene VB?» 8!
8&3? V8148: U. 248 v.
82858.» 8:1 .
8 ++Aouznu Va...» 8! ...
norms: 8:1
8n. +N: +M£MSN1HA8¥1
8+ n: ésmsfiiart
8 + :5 J33 V?» 8!...
8 Emu". Keane VTHAB
Shim. ...AiuvauxABHU
llwer
aw.
3N.
ma.
agency 1
Nu ++Am=m8x+l
so ++Aw=ao¥11
:MNU .o 50:95:14
can“. + .AaxnurtlorMoéé +.r.
a:
“3....an .=8=&.~130 5:5 98332“
:0de
174
be seen that Mo+ does not insert into a C-C bond in either
alkanes or alkenes.
The mechanism used to interpret the formation of all
product ions can be summarized as follows (similar to those
proposed by Beauchamp97):
+
+ H— + all 110 H\
M —"M" _]———I—_; Hshift 3+7": sT_’1rtH /”> __’”(C 5118)
Janync L141: Fe, oo.N1,Mo|
fishiflt
H \
_.._s -m \+/ +
M? H ”275% m. /“ ———’“(C4“6) EM-Fe'“-"il
5741/»
m3 /’
x +
”a H\+ _[——>M( can“)
— -———>M( C5116)“
hE=Feu3n§§
H
‘—-) «03411,? lLr-F'BJOJTiI
+ |
\M/l
( “H
Note that these mechanisms also show a proposed pathway
leading to the formation of M(C3H6)+ and M(C3H4)+, in which
M+ induces the isomerization of cis-Z-pentene to l-pentene
as a first step. Similar isomerization steps have been pre-
viously reported91’97.
175
V. Reactions with l-Hexene (C5fl121
A. Results
+ .. .
l. Fe(CO)x react1ons w1th l-Csfl12
Ions formed as products of ion-molecule reactions in a
mixture of Fe(CO)5 and C6H12 are listed below, with their
precursors as identifiedby double resonance.
ELE stoichiometry precursor(sl
+ +
84 FeC2H4 C2H4
98 Fe(c3H6)1 Fe1, Feco1
11o Fe(c4H6)1 Fe1
+ + +
T40 Fe(C6H12) FeCO , Fe(CO)2
+ +
153 (C3H5)Fe(CO)2 C3H5
+ +
l68 FeCO(C6H]2) Fe(CO)3
+ +
181 Fe(CO)3(C3H5) C3H5
195 Fe(CO)2(C6H]2)+ Fe1, Feco1, Fe(c0)41
224 Fe(C0)3(C6H]2)+ Fe1, Feco1, Fe(c0)21,
Fe(CO)5+
250 Fe2(CO)2(C6H]0)+ Fe1, Feco1
252 Fe2(CO)2(CGH]2)+ Fe1, Feco1, Fe(c0)21
266 Fe2(CO)3(C5H]O)+ Fe1, Feco1
278 Fe2(CO)3(C6H]0)+ Feco1
+ + + +
280 Fe2(C0)3(C6H12) FeCO , Fe(c0)2 , Fe(CO)3
Note that C3H5+ reacts with Fe(CO)5 neutral to displace
two or three CO's to form m/e 18l and 153 respectively. The
formation of m/e 266 is hard to explain. Yet, this peak is
reproducible.
176
2. Co(C01x(N0)y+ reactions with l-C
=0 l
H.
5 12
Ions formed as products of ion-molecule reactions in a
mixture of Co(C0)N0 and (26H12 are listed below, with their
precursors as identified by doubTe resonance.
m1; stoichiometry precursorLiL
101 Co(c3H6)1 Co1, Coco1
143 Co(C6H]2)+ Co1, Coco1, Co(c0)21
158 Coconoc3H51 c3H51
172 CoC0N0C4H7+ C4H7+
173 Coc0N0(c4H8)1 C4H 1, C6H12+,
+ CoN0(C6H]2)+ Co(80)2N01
201 CoCON0(C6H12)+ Co(c0)3)No1
Again, organic fragments C3H5+ and C4H7+ reacts with
Co(C0)3N0 neutral to form m/e l58 and 172 respectively.
m/e 173 is the mixture of Cocouoc4H81, and CoN0(C6H]2)+.
3. Ni(Cle+ reactions with l-C5fl12
Ions formed as products of ion-molecule reactions in a
mixture of Ni(C0)4 and C6H12 are listed below, with their
precursors as identified by double resonance.
177
gig stoichiometry precursor(sl
86 Ni(c2H4)1 Ni1
100 Ni(c3H6)1 Ni1, Nico1
112 Ni(c4H6)1 Ni', Nico
127 Ni80(c3H5)1 c3H51
142 Ni(CGH]2)+ Ni1, Nico1, Hi(c0)21,
Ni(c3H6)1
155 Ni(c0)2(c3H5)1 c3H51
17o Ni(CO)2(C4H8)+ C4H 1, CGHIZ, Ni(C0)2+,
+ NiC0(C6H]2)+ Ni(c0)31
198 Ni(C0)3(C4H8)+ c4H81, Ni(c0)31, Ni(c0)41
+ Ni(CO)2(C6H]2)+
4. Cr(C0)x+ reactions with l-Csfl12
Ions formed as products of ion-molecule reactions in
a mixture of Cr(C0)6 and C6H12 are listed below, with their
precursors as identified by double resonance.
gig_ stoichiometry precursor(s)
+ + + +
l34 Cr(CGH]0) Cr , CrCO , Cr(CO)2
+ + + +
l36 Cr(C6H]2) Cr , CrCO , Cr(C0)2
+ + +
164 CrC0(CGH12) Cr(CO)2 , Cr(CO)3
+, . . ~
5. Mo(CO)x react1ons w1th l-Csfl12
Ions formed as products of ion-molecule reactions in a
mixture of Mo(CO)6 and C6H12 are listed below, with their
precursors as identified by double resonance.
178
m/e stoichiometry precursoris)
l72 Mo(C6H8)+ Mo+, MoC0+, Mo(C0)2+
+ +
200 MoCO(C6H8) Mo(c0)2 . Ho(c0)31
204 MoCO(C6H]2)+ Mo(c0)21, Mo(c0)31
228 Mo(c0)2(c6H8)1 Ho(c0)31
+
232 Mo(C0)2(C6H Mo(C0)3
+
12)
B. Discussion
Table l8 summarizes the reactions of l-hexene with Fe+,
Co+, Ni+, Cr+, and Mo+ together with their product distribu-
tions.
Basically, the reactions of l-hexene with these transi-
tion metal ion can be explained similarly to that proposed
97
by Beauchamp , as follows:
141+ C6H12—>/@ ML,J +\\ -—_+H\;'q1/1._, l-E-Q
alkfil
V l +
ethnic M(2
H Shift [i=é1831333m
3\@ B-aligl , H\+/\
+
fl-H l{\144‘) F‘1n(czntt)
Shift H/ \/ 1HH(04H6)1|Mf=Fe;N1|
H\+
l79
m-fi$8~§ T1 Naawo
fifiaxemel
M «Axon E 8 WET.
+ we: “a.
8n +nm8v~211£8v§ NABVQ a: + «cameovuA8m81
8:91. ..A fixooxzihrwn. m2: . 8~ + “3883831 mos. .
onnu + Moznuszfizoo + +2. 8n + “fivamgvfiilu 3&4 + MABETnABE +3..
. 23308." 233803
c.” o..." 8~+m-=ou§~148¥411mz£8v=
8. 3 8m +~=~ ++Ao=o3~1xA8rJ
S. 9. .3 no. 8+~=~ lozouvnixABvxTi
8. . 8...: .woréjarii
S. oA 8. no. on. 8. o4 o4 8N ...AmfoovfluABET
.8. R. 2. a. 3. a. a. “N. n... a. Bramfisixaril
2. kNfizmovaBET.
a. . 8 +83 $1.3?ng
3. mm. 3. 8 +o=no 4%:n8848vzilllxm8v:
nNHnNH2n~a~H-~HHx
o.“ ~=~ + oncurj
8. u: + Aonzouvfll
.3. «A. 3. Aun=wo¥1|1
2. 8. mam. 182.6%]
am. am. co. can". iwxnurill
2. 3...”. JazwsxileMzooA ++=
£ .5 =— 8 o.—
A «found v 885:4 5:. 338$ .... on See
180
Note that Cr+ and Mo+ prefers to insert into C-H bonds
to eliminate H2. Mo+ can insert into C-H bonds to eliminate
two hydrogen molecules. The intermediates having C3H5+ or
four n electrons distributed on three carbons are reasonable,
because they are good n-donor ligands and exist as the
stable fragments in the mass spectrometer587’168.
VI. Reactions with Z-pentanone (Z'Csfliogl
A. Results
+ . .
l. FeiCO)x react1ons w1th 2-C5fl1og
Ions formed as formed as products of ion-molecule reac-
tions in a mixture of Fe(CO)5 and C5H100 are listed below,
with their precursors as identified by double resonance.
gig stoichiometry grecursor(s)
84 Fec2H41 C2H4+
114 Fe(03H60)1 Fe1, Feco1
14o Fe(c5H80)1 Fe1, Feco1
142 Fe(C5H100)+ Fe1, Feco1, Fe(80)21
17o FeCO(C5H100)+ Fe(00)21, Fe(00)31
181 Fe(c0)3c3H51 C3H 1, c3H71
l83 Fe(80)3c3H71 C3H7+
197 Fe(c0)5H1 c3H71
198 Fe(C0)2(C5H]00)+ Fe(c0)31, Fe(c0)41
211 Fe(C0)4C3H7+ c3H71, C5H100+
+
+ +
225 Fe(C0)3(CSH]00) Fe(CO)4 , Fe(C0)5
+ + +
181
m/e stoichiometry precursor(s)
0)1.
0)1
FeC0(C5H10
Fe(CO)2(C5H10
Note that C3H5+ and C3H7+ react with Fe(CO)5 neutral
to yield m/e l8] and 183 and 2ll respectively. Also, C3H7+
can protonate Fe(CO)5 to produce Fe(CO)5H+.
2. ngtQ)_LNO) + reactions with Z-CCH O
x y 0—10—
Ions formed as products of ion-molecule reactions in a
mixture of Co(CO)3NO and C5H100 are listed below, with their
precursors as identified by double resonance.
gig stoichiometry precursor(sl
117 Co(c3H60)1 Co1, Coco1
143 Co(c5H80)1 Co1, Coco1
+
145 Co(C5H]00)+ Co1, CoCO , Co(c0)21
+ +
l58 CoCONOC3H5 - C3H5
+ + +
l73 CoC0(C5H]00) CoCO , Co(C0)2
175 CoNO(C5H]00)+ Cocono1, Co(c0)2No1
+ + +
203 CoC0N0(C5H]00) 05H100 , CoCONO ,
Co(c0)2Ho1, co(c0)3N01
+ +
2l6 CoN0(C5H]0)(C3H5) C3H5
+ +
128 CoNO(C5H]00)(C3H7) C3H7
+ + + +
229 Co(CSH80)(CSH100) Co , coco , Co(C5H80)
+ + + +
23l Co(CSH]00)2 Co , CoCO ,~Co(C0)2 ,
Cocono1, Co(c0)2No1,
co(c0)3N01, C5H100+
182
Again, C3H5+ reacts with Co(CO)3NO netural to produce
m/e l58 and m/e 2l6. C3H7+ reacts with Co(C0)3N0 to pro-
duce m/e 2l8.
3. Ni(C0)x+ reactions with 2-C5fllog
Ions formed as products of ion-molecule reactions in a
mixture of Ni(C0)4 and C5H100 are listed below, with their
precursors as identified by double resonance.
gig stoichiometry Erecursor(s)
1oo Ni(c3H6)1 Ni1, Nico1
102 Hi(c2H40)1 Ni1, Nico1
ll6 Ni(c3H60)1 Ni1, Nico1
144 Ni(C5H]00)+ Ni1, Nico1, Ni(c0)21
155 Ni(c0)2c3H51 c3H51
157 Ni(c0)2c3H71 c3H71
172 NiC0(C5H]00)+ Ni(c0)21, Ni(c0)31
23o Ni2C0(C5H]OO)+ Hico1, Ni(c0)21, Ni(c0)31,
Ni(C5H]OO)+, Hi(c0)41.
NiC0(CSH100)+
. . + + . .
S1milarly, CBHS and C3H7 react w1th N1(CO)4 neutral
togive m/e l55 and m/e 157 respectively to displace 2 CD'S.
4. Cr(c0)x1 regctions with 2-c5glog
Ions formed as products of ion-molecule reactions in a
mixture of Cr(C0)6 and C5H100 are listed below, with their
precursors as identified by double resonance.
183
gig stoichiometry grecursor(s)
+ + +
l38 Cr(C5H]OO) Cr , CrCO
+ + +
l66 CrCO(CSH100) Cr(CO)2 , Cr(C0)3
+ +
224 Cr(C5H100)2 CrCO , Cr(C0)2+, or(c0)31
5. Mo(C0)x+ reactions with 2-csglog
Ions formed as products of ion-molecule reactions in a
mixture of Mo(CO)6 and CSHlDO are listed below, with their
precursors as identified by double resonance.
mie stoichiometry precursor(s)
+ +
l74 Mo(C5H60) M0
175 Ho(c5H80)1 Ho1, Hoco1
+ + + +
178 Mo(C5H100) Mo", MoCO , Mo(C0)2
204 Moc0(c5H80)1 Ho(c0)21, Mo(c0)31
+ + +
232 M0(C0)2(C5H80) MO(C0)3 , M0(C0)4
B. Discussion
Table 19 summarizes the Fe1, Co1, Ni1, Cr1 and Mo1
reactions with 2-pentanone, together with their product dis-
tributions. 1
The formation of all reaction products can be explained
87 et. al.
very similarly to the mechanism proposed by Freiser
as shown in Scheme XXIX.
Note that Fe+, Co+, Ni+, Cr+ and Mo+ can form stable
addition products, M(C5H100)+. Hi1 and Cr1 failed to elimi-
nate H2. However, Mo+ has extra energy to eliminate one
“- u u
+38 .2
3ofl88~£ll£8xz +moo~=n38§+|ooa=mu + NAB:
.9
fi.
Aoo~=n38~leooa=m8 + wh8v~=a|£8v= 38:. uoofmuxomchBTl oofizno dam—MES] can—Mu + 8
+
~23: .
+
493850 .
YNfievs .
woofnovaxll comm”. + moorMSQoo .loofnu + .888
£85 .
0
800.
Acommgootll corn". loormuvoofloouzmo + o .8
4 4
800 .
‘
. 133.qu 2; 388m
4 c; on. i. a.”
11 2.. B. B. o; oA
oA me.
an.
no.
3.
00H “0.
no.
no.
2.
mm.
.HH.
kw.
o; 8.
ma.
3.
4A oorMUVExABE
8w AoofmfiBWuABrMI
8 ¢ coranBAJBr +938 V:
8 Exam +moaz~37ufi8rol
.AoornovraarT
8N +Moofm3~148rtl
8 cmooaxnuvaiuABIT
8~ .m: documoanABril
8 .N: éomsmovanuaerol
8 +a=~u JawznovfliuABrI
8 +oa=~u +vo=noVaunA8rT
3.
nn.
3.
no.
mm.
3.
no.
an .
~a.
wane ies—mugs]
Carma 9 AmanoYT
Aces—MOYI
~=~ zoo—Moral.
u: .. Sonnet?!
ammo o Aowanoroll common.“ + z
... o
«I
8
p.
82.318 885a£-~ 5S, .8338
m. and.
185
fig ....
+30%"; : Tam“. (2,18
a m
mun-+4100
m
fund :
.85. E mam...
n8 Aoznov z AoamNuV z
fl\+zloo + +
«2.00.9muz
fl 8% al..“
m .M 3 z a
A114: 18 \—
Alum/3 ._1+:ti. /\/+:
CODhMU
2.38
a: . :25
£48.27: e258- m.
Aomznov z Aimwflli ....Hmnm TIQIQ 8:03."on /\/=\A|I/\J.\ + .w
+ +210 :\0 O
‘u
x _ xx oszum
186
more Hz to form Mo(C5H60)+, implying that Mo+ has the
strongest tendency to insert into a C-H bond. It is of
interest also to note that Ni+ is the only metal which can
insert into a carbonyl-carbon bond, producing Ni(C2H40)+
87 also reported to see FeCO+ and
and Ni(CBH6)+. Freiser
Fe(C2H4)+ in a trace amount (< l%) in the reaction of Fe+
with 2-epntanone, which is not observed in our experiment.
Another trace amount (6%) of the product ion, Fe(C4H8)+
reported by Freiser is not observed in our experiment.
VII. Reactions with sec-Bugylamine (s-BUNHZ, c451rg1
A. Results
l. Fe(001x+ reactions with sec-butylamine
(s-BUNH2i_£4fl11gi
Ions formed as products of ion-molecule reactions in a
mixture of Fe(CO)5 and sBUNH2 are listed below, with their
precursors as identified by double resonance.
gig. stoichiometry grecursorisl_
87 Fe(CHBNH2)+ Fe1, Feco1
98 Fe(c3H6)1 Feco1
11o Fe(c4H6)1 Fe1
113 Fe(c3H7N)1 Fe1, Feco1
114 Fe(C3H6NH2)+ Fe1, Feco1
125 Fe(c4H7N)1 Fe(c0)21, Fe(c0)31
127 Fe(c4H7NH2)1 Fe1, Feco1
128 FeCO(c2H6N)1 Feco1
Dis
129
157
185
202
stoichiometry
+
Fe(C4H9NH2)
+
FeC0(C4H9NH2)
Fe(50)2(c4H9NH2)1
+
Fe(C4H9NH2)2
187
ggecursor(s)
Feco1, Fe(C0)2+
Fe(c0)21, Fe(c0)31,
Fe(c0)41
Fe(c0)41, Fe(co)5+
Feco1, Fe(c0)21, Fe(c0)31,
Fe(CO)4+
2. Co(C0)x(NOly+ reactions with sec-bugylamine
sBUNHz, c4511gl
Ions formed as products of ion-molecule reactions in a
mixture of Co(CO)3N0 and sBUNH2 are listed below, with their
precursors as identified by double resonance.
mie
90
100
102
104
ll3
116
128
130
132
160
162
stoichiometry
Co(CH3NH2)1
Co(c3H5)1
Co(c2H5NH2)+
Co(c2H5NH2)1
Co(c4H6)1
Co(c3H7N)1
Co(c4H7N)1
Co(C4H7NH2)+
Co(c4H9NH2)1
Coc0(c4H9NH2)1
CoN0(c4H9NH2)1
precursorisl
Co+
CoC0+
Co1, Coco1
Co1, Coco1
Co+
Co1, Coco1
Co+
Co1, Coco1
co1, Coco1, Co(c0)21
Coco1, Co(c0)21
Coc0N01, Co(80)2Ho1
Co(c0)3No1
188
mie stoichiometty precursor(s)
+ +
190 CoC0N0(C4H9NH2) Co(CO)3NO
+ + + +
205 C0(C4H9NH2)2 Co , COCO , C0(C0)2
+ +
221 C02N0(C4H9NH2) Co
+ +
249 C02C0N0(C4H9NH2) COCO
3. Ni(C0)x+ reactions with sec-butylamine
(sBuNHz. c4511gl
Ions formed as products of ion-molecule reactions in a
mixture of Ni(c0)4 and sBUNH2 are listed below, with their
precursors as identified by double resonance.
mie stoichiometry precursor(gi
+ .+ . +
72 C4H10N N1 , NlCO
89 Ni(CH3NH2)+ NiCO
103 Ni(c2NH2)1 Ni1, Hico1, Hi(c0)21
. . . +
115 Ni(c3H5NH2)1 N11, NlCO
117 Nic0(CH3NH2)1 Ni(c0)21, Ni(c0)31
. + .+
127 N1(C4H7N) N1
. + +
128 N1C0(CH3CHNH2) CH3CHNH2
. + .+
l29 N1(C4H7NH2) N1
131 Hi(c4H9NH2)1 Nico1, Ni(c2H5NH2)1.
Ni(c0)21
132 NiC0(C4H6)+ Ni(c0)21, Ni(c0)31
1 Hi(c0)41
145 Ni(C0)2(CH3NH2)
159 Nic0(c4H9NH2)1 Ni(co)21, Ni(c0)31,
Ni(50)41
189
gig stoichiogetry precursor(gl
. + . + .
204 N1(C4H9NH2)2 Ni(c0)2 , N1(C0)3+,
Ni(c0)41, Ni(c4H9NH2)1.
Nic0(c4H9NH2)1
. + .+ . + . +
215 N12CO(C4H7NH2) N1 , N1C0 , N12(C0)2
4. Cr(C0)x+ reactions with sec-bugylamine
(sBUNHz, c4g11gi
Ions formed as products of ion-molecule reactions in a
mixture of Cr(C0)6 and sBUNH2 are listed below, with their
precursors as identified by double resonance.
gig stoichiometry grecursor(s)
112 CrCOCH3NH31 Crco1, Cr(c0)21
+
123 Cr(C4H7NH2) Cr1, Crco1, or(50)21
+
125 Cr(C4H9NH2) Cr1, Crco1, Cr(c0)21
+ + +
l53 CrC0(C4H9NH2) Cr(C0)2 , Cr(C0)3 ,
Cr(CO)4+
+ + +
l98 Cr(C4H9NH2)2 CrCO , Cr(C0)2 ,
+ +
Cr(C4H9NH2) , Cr(C0)3 ,
CrCo(C4H9NH2)+
5. Mo(C0)x+ reactions with sec-butylamine
stUNH2.§4flJ]fll
Ions formed as products of ion-molecule reactions in a
mixture of Mo(CO)6 and sBUNH2 are listed below, with their
precursors as identified by double resonance.
mie
151
163
165
191
217
219
190
stoichiometry
Mo(c4H7N)1
Mo(c4H7NH2)1
Mo(c4H9NH2)1
MoC0(C4H7
Mo(C0)2(C4H7N)+
Mo(C0)2(C4H7NH2)+
+
"”21
precursor(s)
Mo+, MoCO+
Mo1, Moco1
Moco1, Mo(c0)21
+ +
M0(C0)2 , M0(C0)3
Mo(CO)3+
Mo(CO)3+
6. WLC01x+ reactions with sec-butylamine
(sBUNHz, cqgllgi
N(C0)x+ reactions with sec-butylamine were also studied
by use of the CEC 2l-llO B double focusing mass spectrometer
without precursors identification.
are listed below:
mie
225
237
249
251
253
279
307
309
337
343
stoichiometry
Hc(CH3NH2)1
w(c3H5N)1
HC(c3H5N)1
H(c4H7N)1
W(C4H7NH2)+
Hc0(c4H7N)1
W(CO)2(C4H7N)+
H(c0)2(c4H7NH2)1
w(c0)3(c4H7NH2)1
W(CD)3(C4H9NH2)+
The reaction products
191
B. Discussion
Table 20 summarizes the Fe+, Co+, Ni+, Cr+ and Mo+
reactions with sec-butylamine and also their product distri-
butions.
Scheme XXX shows a proposed mechanism leading to the
formation of reaction products.
It is of interest to note that Ni+ undergoes a hydride
abstraction from sec-butylamine to yield C4H10N+ and that
Fe+ forms a high energy product, Fe(C3H8N)+ by losing a
methyl group after inserting into the C-C bond of sec-
butylamine.
Table 20 also shows a trend that Fe+, Co+ and Ni+
exhibit much richer chemistries in their reactions with sec-
butylamine. However, Cr+, Mo+, and N+ are mostly inducing
hydrogen molecule(s) elimination. Note that NC+ is a reac-
tive species, since it produces NC(CH3NH2)+. Also, w+ is
the only species which can eliminate both CH4 and H2 to pro-
duct W(C3H5N)+. In Scheme XXX, a process pathway leading to
the formation of M(CH3NH2)+ is also shown. Note that this
process needs an isomerization of sec-butylamine to isobutyl-
amine. Similar isomerization steps have been previously
reported91’97.
Also shown in Scheme XXX are pathways leading to the
formation of Fe(C3H6)+, Fe(C4H7N)+, Co(C3H5)+, Ni(CH3NH2)+
with a concurrent loss of the C0 ligand.
n + ...-21*” i “9.1
192
mm 20 . huuou with soc-m2 (bit-l2)
h
b I1
u—ol(u,-z).+ Gals
Mead-Y3 ski I.
amazing-,1» c211.
HKC‘IGYO .34 I2
“wai al.,
Hit-1,1131% a},
H‘ch‘ff1 a:
Exam-1‘s I,
«9.119.211
J37
.10
.03
.26
.23
.11 .19
.16
Knox—1
HKQ)‘.II‘Q gun. a
grabbing-1,11. can‘t m
~Km)u(flj.z Y. can‘o m
~10).(m,~,)'9 ‘3':
—.1(co)‘_1(a3-3y. can}. do
~9KQ)b1(P7I5I).o 0211... :2. w
«991.19.»? en,
Mews-5'12)“ can on
MOLJm‘usnzii of“ 2m
and. 1,4(95314 can). on
“(mk‘ficjl‘fi 03.24. in
«whom-.13 '3‘ I21 90
«mum-.33 If I.* w
antitank-1i an an
Mayday-3+ «1,. on
*wlhfiwfi :29 m
ammuwwlfo nzmn
Mm),.,(q.s,li+ a, . do
~(m),_2(wf+ 2.112 mm
~Km),_,(c.l,lit again
«away-1,11.
“wk-I‘vv'z" ‘”
H‘m)x-2(%-z).1 24D
an),,,(c.I,-z)’* ID
«90),; «mm-wqwii an
E‘muflcu'gizl.’ 2m
Km),-,n(c..l,-,)'i Jan
.16
.15
.25
.19 .fi
.‘6 .75
.61
.17
. 16
.19
.19
.16
. 35
.10
.18
.25
.60 .56 .50
.111
.36
T? .a
a” 1.0
.56
1.0 a,
.68
8 1'. action I
a 1 v
"1 W‘M-in-zF-Nwt
”((11); o
hung-awa5-23—a
No.51, )12 x e 3.4
0(0)}. muz—pwqugnzf—ouqnylzfz 230.1,!
cm «9 «(0)3n—H032m119 o-m-zdbzflqlflz‘ no
am. 4 Mm),p—omz(m)zflomz—mzm(6~l9flzy
Nani; --wnz-+u(c,n,-I,1—ua(c.u,w’, x = o. 1
111(c11i);...nuint2 «1.5-2141(wlz’:
W). “2.3.1.
94001; an. Q'io'lzf-‘mws'fl;
3.9.2). 331.2,)
E
n
......5... E ME 2.3” ......ae
mazes! Aninfir manor :4.
n A... . ./.\ new END ..
51w. . «a; uninzorfinflov immmll «sameneéinfi sf... .....5
\I \2 / \\
N...._ We... .— .a-
Enos-gig? a 3mm“; ......MM . . fl
3 c. .3: flaw: n 1 _ E or
w ...; m. ...... .... ...... a av... .2
.72. A N... no: fl
._ flawed mi EA .... fwd
sum. ....n o .w . .32 may. nwiuuwe hmfikanfi
CE. .2... .A n .— 75 55. fl
..- 43...; 32 2&5... : M233! .1 Q
new n in.“ Sandy \_ __ 5w... new new.
- NE - _ n x -
“Mu—”HA 6 «SJ. 5&0 5 NEthhzmu nBVu .n new”... arena
2... Sat8§\_. gates; coated
.3 one 0.0
:
EVI/ Tulle...
. xxx gunom
194
There are some new reaction products which are not seen
in metals alone. The formation of CrCO(CH3NH3)+ can be
explained similarly to the formation of M(CH3NH2)+ in Scheme
XXX except that one more B-H shifts from the left side to
give CrC0(C3H5)(CH3NH3)+ followed by a simple fragmentation.
The formation of FeCO(C2H6N)+ can be explained by the FeCO+
insertion between skeletal atoms C2 and C3 with CO acting as
a spectator (because CO is retained on Fe+). Note that the
loss of C3H5-radical which is also seen in CID spectra of
166
Co+ reactions with alkanes is a high energy process.
VIII. Conclusions
Based on our original goals, we have the following
results from this work:
1. All transition metal ions in this work insert into
a C-I bond.
2. All transition metal ions in this work induce a
8-H shift.
3. It is hard to tell if 5 or 6 membered rings are
preferred in the intermediates. It appears that
6-membered rings are the preferred intermediates
in the reactions with 2-pentanone.
Apparently, one cannot yet explicitly explain the
differences of metal ions in terms of the controlling fac-
tros which were assumed in the beginning of this section.
However, from this study, it seems that thermodynamics only
195
dominates the reactions disregarding geometrical considera-
tions.
From this study, we do see a certain pattern of how
different metals react in different ways, for instance, Fe+,
Co+, and Ni+ are more reactive than Cr+ and Mo+ in that Fe+,
Co+ and Ni+ have stronger interactions with organic sub-
strates and thus can rearrange the organic molecules to
different products. Cr+ and Mo+, on the other hand, seem to
prefer to eliminate one or more molecules of H2, which is
especially true for Mo+.
In order to gain some insight into the differences in
chemistry of metal ions, an ab initio calculation on CrCHz+
is being pursued and will be discussed in the next section.
AB INITIO CALCULATION
C. AB INITIO CALCULATIONS
1. Introduction
I. The Importance of Ab Initio Calculations:
In the last section (8.5), we have seen that
Fe+, Co+, and Ni+ exhibit similar chemistries in their reac-
tions with most of the organic compounds used, and Cr+, Mo+
and w+ fall into another group, which exhibit quite different
chemistries in their reactions with the same organic com-
pounds. However, we can not understand the bonding behavior
of these metal ions in terms of their orbitals, orientation,
size and symmetry as those listed at the beginning of sec-
tion 8.5 from these studies. Note that the reaction products
we have observed from ICR studies are mass peaks. Therefore
the assignments of ion structures of these products are
sometimes ambiguous, although the double resonance technique,
the CID (collisionally induced dissociation) technique and
the use of the labelled compounds can sometimes give helpful
information. It is important to do the ab initio calcula-
tions so that we can obtain insights into the nature of the
bond between the metal and substrate.
60, corresponding to an [Ar]
157 6
Fe+, has a ground state,
1 6 157
45 3d configuration Cr+ has a ground state 8,
corresponding to [Ar] 3d5. However, there is some
evidence““’148 for the presence in the gas phase of its next
excited state 6D (45 3d4) at l.47 eV. It is not surprising
that Cr+(6S) will form an addition complex with l2-crown-4
196
197
electrostatically. To form covalent bonds with ligands, it
presumably needs to use its 45 orbital. 0n the other hand,
Cr+ (GD) might be able to form covalent bonds with ligands
2
4.
by use of sp hybrid orbitals. Ni has a ground state D,
9 157
corresponding to the [Ar] 3d configuration and there is
no evidence that its next excited state 4F at leV (corres-
8
ponding to [Ar] 451 3d configuration) participates in the
gas phase chemistry. Thus Ni+ (20) will interact with
ligands electrostatically and form the addition complexes
with TDE, l2-crown-4 and l5-crown-5. It can also interact
with ligands covalently by use of its 5 and d orbitals+163
or sp hybrid orbitals by promoting to 4F statel123’164’165.
+ Ni+ has a ground state of [Ar] 3d9. To form a covalent
bond with a ligand, a d orbital has to be involved in bond-
ing, or it may mix with a s orbital to form a sd hybrids,
or Ni+ could be experiencing the configuration mixing of
[Ar] 3d9 and [Ar] 45 3d9, as a result of positive charge154
on the metal center. CID spectra of some product ions in the
reactions of Co+ with alkanes conducted in our laboratory166
also indicated that there are more than two different struc-
tures at the same mass unit, which might be derived from the
reactions of different electronic configurations of the metal
ion.
$ Note that Kunz concluded the 452 bonding to H by use of sp
hybrids for all the first transition metals. Later,
Walch et. al. correlated the s orbital with p orbitals, end-
ing up with a much d character in ScH with '2+ as its ground
state. However, its next excited state, 3A which is the
ground state from Kunz's calculation is only 4.6 kcal/mole
above 'Z+ ground state.
198
II. Review of Ab Initio Calculations of Transition
Metal Compounds.
From spectroscopic studies, P.R. Scott and
150
N.G. Richards concluded that three types of configura-
tions are important in the monhydride, and they may be
2 n-2
loosely described as s , 5dr"1 and d". In ScH and TiH,
2
the s d"'2 is important. From TiH to FeH, the sdn'1 config-
uration always gives rise to ground state; however, the d
electrons are largely non-bonding in these compounds. How-
ever, in NiH the s-orbital contribution is still not
negligible. In the second transition series, the d-orbitals
are relatively less stable, and PdH may be understood in
terms of simple covalent bonding between 4d- and ls orbitals
alone+.
151
However, recent ab initio calculation shows that ScH
has a tremendous amount of do bonding in the ground state.
151
S.P. Halch incorporated atomic correlation terms leading
2 1 3d9 separation for the Ni atom
into all electron MCSCF/CI calculations for the XZA state of
to an accurate 4s 3d8- 4s
NiH and found that although the bonding in NiH is predomin-
2
atly 4s1 3d9 like, 45 3d8 like configurations are found to
be important in the small R region, and he expects that
+ 3d orbital is only 30% size of 4s orbital. Hence, there
will be poor overlap of 3d with ligands; 45 on the other
hand will form bonds easily. In contrast, 4d has 60% size
of 55 orbital (electron density distribution), which then
makes it possible to bond by use of a 4dz orbital.
199
2 l
a mixture of 4s 3dn and 4s 3dn 1 to be important in
l 6
most transition metal bonds except for Mn where 45 3d is
152
too high to be accessible for bonding From other calcu-
lations on systems, such as NiH, NiCHz, NiCH3, Nico‘53,
2, CrH154, it was concluded that the s orbital is impor-
CrCH
tant in bonding with only a small contribution from d
orbital or a weak d bonding. To form a n bond, the ligand
has to get closer to the metal, and the strong repulsive
interaction with 45 orbital in the 452
3dn configuration
will avoid this energy lowering process, unless the 4s
orbital mixes with 4p to become two 4s 4p hybrids as in
NiCHz. Also, model studies of n-bonded organometallic sys-
tmes Mn-CZH2 and Mn-CZH4 are calculated to have small bond-
ing energy, < 10 kcal; Ni-CZH2 is bound by 2 20 kcal, and
Steigerwald and Goddard calculated the CpZTi(C2H4) and
ClzTi(C2H4) systems and found that metallocycle form is
more stable than n-bonded form by C2H4.
0n the other hand, MnCH2+ and CrCHz+155 were calculated
to have the very similar ground state molecular orbitals,
they all use M+ 45 to bond the methylene 3a1 orbital although
1 5 ground configuration, while Cr+ is 3d5.
154
Mn+ has a 45 3d
CrCH; seems to be similar
Further, the ligand has a strong influence on the mtal
center for its bonding to another ligand124’156. It turns
out that the change of the electronic configuration of a
metal center used for bonding might be crucial as we keep
200
putting ligands on it. In the solid state, if we keep add-
ing electrons into ZnS sytem to become GaSe then CuPbS and
then As system, we will see some bonds broken in the unit
cell to become different structures. Therefore, the
123
strategy that Beauchamp used in relating promotion
1 3dn'], to bond energy between first
energy, from 3dn to 45
transition metal ions and simple ligands has still to be
scrutinized as ligands are changed to the multifunctional
molecules, since the transition states or intermediates
of the metal ion molecule reactions might involve many
ligands to be bonded to the metal center.
III. Theory_of Ab Initio Calculations
Although there are a growing number of ab
initio calculations published in the past two decades‘72,
very few of them were devoted to the understanding of
univalent metal ion-ligand species. There are some books
available for discussing the theories in ab initio calcula-
173'177. However, a brief review will be discussed in
tions
the following.
From quantum mechanics, a total wave function of an
atom can sometimes be expressed by a Slater determinant.
For example, the lithium atom can be written as
¢1s (1) ¢is (l) ¢2s (1)
¢1s (2) ins (2) ¢2s (2)
¢is (3) ¢1s (3) ¢2s (3)
sill"
20l
or in a simpler form of
4" ¢l$ 371—5- “’25
where o's are one-electron wave functions (spin orbitals).
In most calculations the orbital m is expanded in terms
of a set of basic functions,
Our problem is to determine the coefficient Ci' The
Xi can be chosen to form an orthonormal set,
{xgdv=1
f X.X. dV
1 J 0 for i # j
but in practice it is more convenient if they are normalized
but not orthogonal, i.e.
but
Thus the wave function of the molecule W can be broken down
as follows:
W = and , these can be broken down into inte-
grals involving the orbitals, whichiriturn reduce, or
expand to atomic integrals involving the basic function
Xi'
In writing the electronic Hamitonian equations, there
is considerable advantage if one works in atomic units, in
which n (planck's constant devided by 2n), the electronic
charge e, and the electron mass me are all unity. Also,
the distance is given in bohr radii ao (bohrs), where l bohr
O
8 cm or 0.52918 A. The unit of energy is
is 0.529l8 x 10‘
the Hartree, equal to 27.2l ev or 627.5 kcal/mole and is
equal to ez/ao.
The rigorous mathematical expression of the molecular
orbital model is the Hartree-Fock(HF) approximation. For
closed-shell atoms and molecules the HF wave function is of
the form
we = A (n) ¢1(l)¢2(2)....¢n (n) (1.1)
and is conveniently written as a Slater determinant:
_1_
we = ,3. l ¢] (1) m2 (2) ...... ¢n and can be
expressed as,
N
E (H 25 + J
log loglo
+
°' =
2' £9)
9
where 513 is the energy of one electron in a molecule and
9 *
is defined as flog (l) H? log (I) dV1 and J is a coulomb
integral and is defined as
(2) dV1 dV =
2
* ‘k 'l -
ff] ———
Cg (1) log (2) r12 log (1) log
2 l 2
If log (1) -——- log (2) dV1 dV
r12 2
(assuming the orbitals are real).
More generally, the energy is a sum of one-electron,
coulomb, and exchange terms
E = 2e? + z m]? (l) —1—¢§ (2) dv1 av
i i1. (1)-
. 8scr
J J
m, (1)
or, in an even more compact form we may write
HSCF ¢l (1) = €§CF¢i (1)
Now if ¢i = Z Cin Xn
n
SCF _ SCF
then H i C1n Xn - 6i E C1n Xn
By multiplying both sides of this equation by Xm and
integrating over all space, we get
i Cin ("ifiF ' iCF Smn) = 0
or
.... Hggi - .§CF smn - o
If HgfiF and Smn could be calculated the secular deter-
minant could be solved directly for the eigenvalues, the SCF
SCF
orbital energies 8i However, both Hmn and Smn demand a
knowledge of the wave functions we are trying to find and
206
the solution has to be iterative, which makes the use of a
computer mandatory.
2
For example, in LiH, the structure for Li is ls 25 and
H is ls, and
NLiH; x 2
now
In outline, our procedure might be:
I. Guess some values of the 0'5
2. Calculate all the various atomic integrals and
. SCF
hence build up Hiojo and Sioj
3. Solve the determintal equation giving the possible
values of e§CF
4. Substitute these in the secular equations giving
new 0'5
5. Go back to stage I and repeat until the values of
SCF
8i
an arbitary threshold (m10'6) and then take the
or the C's converge to steady values within
values of the converged C's.
The result of this will be to provide SCF orbitals,
both occupied and unoccupied: that is to say the coeffi-
cients in
178
For open - shell SCF methods, both Roothaan's and
207
179
Nesbet's method are used in an attempt to eliminate the
SCF
ii
ted by means of unitary transformations.
off-diagonal terms a which cannot be completely elimina-
The secular equations may be simplified by the symmetry
considerations. The role of symmetry enters the calcula-
tions in the following way:
l. It enters the integral calculation and the basic
function transformation 188’189.
2. It enters SCF calculation. Linear combination of
basic functions for different irrepresentation species will
reflect the symmetry of the molecule. This will simplify
the large matrices by reducing them to a series of blocks
along the diagonal, one for each molecular orbital symmetry
type. Also, the matrices can be simplified by the use of
symmetry operator.
3. It enters MCSCF and CI calculations. By mixing
configurations having the same symmetry, the matrices in
both calculations can be simplified‘go.
Before going to the next section, some important con-
cepts used in ab initio calcualtions are summarized in the
following:
l. The solution of the HF equations including symmetry
and equivalence restrictions yields the restricted Hartree-
Fock (RHF) wave function. By removing the symmetry and
equivalence restrictions placed on RHF wave functions,
single-determinant wave functions of lower energy can fre-
quently be obtained..
208
2. Electron correlation. In the HF approximation,
the motion of each electron is solved for in the presence
of the average potential created by the remaining (n-l)
electrons. The contribution to the total energy due to
instaneous repulsions is called correlation energy180. The
most frequently used method for approaching the electron
correlation problem is configuration interaction (CI),
which is just a linear combination of configurations with
coefficients variationally determined. More precisely, the
CI wave function is of the form,
where the 6's are an orthonormal set of n electron configura-
tions. The coefficients C1. are determined to minimize the
energy fngewedz. Application of the variation principle
leads to
CI -
where ”ii = f¢;He¢jdz and is the matrix elements between
configurations. Note that matrix elements Hij between
different configurations i and j are zero if i and j are of
different symmetry. Therefore, the secular equation is
greatly simplified by only considering configurations which
have the total symmetry of the particular electronic state
being investigated.
209
3. Multiconfiguration SCF (MCSCF). Solution of the
eigenvalue problem in (1.3) will yield the optimun values
of C1, C2, and C3. However, the wave function has not been
variationally determined yet unless the spin-orbitals used
in the configurations have also been varied to minimize the
orbital energy. The MCSCF wave function is the best
(lowest energy) wave function that can be obtained by simul-
taneously varying both the orbitals ¢ and the CI coef-
ficients C. This regards to MCSCF caluclations, which is
again usually solved by the interative techniqueIe].
4. Basis sets. Molecular calculations are generally
carried out in terms of basis functions centered on each
atom in the molecule. A primary consideration in evaluat-
ing the reliability (i.e. in agreement with reality) of an
electronic structure calculation is the basis set, in which
both slater and gaussian functions are most frequently
employed. Slater basis functions have the radial form
A rn-l e-(r
where A is a normalization factor, n is the principal quan-
tum number, and E is the orbital exponent or screening
parameter. Guassian basis functions have the radialy form,
- 2
B r" e or
In principle one would probably prefer to use slater
functions in all molecular calculations. However, for
210
nonlinear polyatomic molecules the two-electron integrals
(such as coulomb and exchange integrals) in terms of a
slater basis set are extremely difficult to compute, while
the same integrals in terms of gaussians may be evaluated
comparatively simply and rapidly.
There are many different basis sets used depending on
the systems we are dealing with.
a. The minimum basis set. It includes one function
for each SCF-occupied atomic orbital with distinct n and t
quantum numbers. The functions of a minimum basis set are
usually slater functions. The atom-optimized minimum basis
182 may be used directly molecular
set developed by Clementi
calculations.
b. Double Zeta and extended basis sets. Usually, the
minimum basis sets yield SCF energies above the HF energies.
For this reason electronic structure calculation are fre-
quently carried out with larger basis set. A double zeta
includes exactly twice as many functions as the minimum
basis. Any basis set of slater functions larger than
double zeta are referred to be an extended basis set.
c. Contracted functions. As mentioned above, for big
and nonlinear molecules, the use of gaussian functions are
utilized to speed up computing the multicenter integrals,
despite the fact that the number of gaussian functions is
more than tiwce the number of slater functions. In order
to save the computation time, the use of contracted
211
gaussina5183"185
, i.e, linear combinations of gaussians
with fixed coefficients is suggested. In solving the SCF
equations, then, only the coefficients in each SCF orbital
of the contracted functions must be determined. A notation
of (105 5p/35 2p) means a basis set of ten 5 and five p (for
example six ls functions are used for ls orbital, two ls
functions are used for 25 orbital and two ls functions are
used for 35 orbital and three 2p functions are used for 2p
atomic orbital and another two 2p functions are used for 3p
orbital) has been contracted to three 5 and two p functions.
d. Polarization functions. In order to approach the
HF energy limit, functions with higher 2 values (d,f....)
must be added to the basis when we are working on molecules.
These functions with higher 2 values being added are called
186
polarization functions A basis including (4s 2p ld) on
each first row atom and (25 lp) on each hydrogen is referred
187
to as double zeta plus polarization Note that the
double zeta level for a first row atom is (4s 2p).
The reliability of each basis set depends on the
molecular pr0perties we are dealing with. Therefore, the
choice of the basis set is often based on experience.
2. Use of CrCH2+ As An Example of Ab Initio Calculation
In this section, a study of CrCH; will be given as an
example of an ab initio calculation.
212
Assume Cr+ interacts with CH2 in a C2v symmetry in the
following coordinates:
y
b H
L L
Cr+ ./
a Z
1 HR
b
b] R
where the various orbitals can be associated with the irre-
ducible representations of C2v character table
191,
C2 ov(xz) o;(yz)
2v
A1 1 1 ’ 1 z " x2,y2,22
-A2 l -l -l Rz xy
81 -l l -l x, Ry x2
82 -l -l l y, Rx yz
I" o o 2
1
By using the formula, a. = %-EX(R)Xi(R) where h is the
R
group order, X(R) is the character of the matrix correspond-
ing to operation R in the reducible representation, and
Xi(R) is
symmetry
any irreducible representation, we can get the
species for the bonding orbitals bL and bR.
(2 + 2)
l i.e., there is one A1 species
n+a
(2 + 2) l i.e., there is one 82 species
A+a
213
To determine orbitals correspondong to A1 and 82, we use a
projection operator‘gz.
3A1 HR = (1)1511R + (1) 62 HR + (1) 8v(xz)HR + (1) o;(yz)HR
= HR + HL + HL + HR = 2 (HR + HL)
A32 x A A A1
P HR = (l) E HR + (-l) 02HR + (-l) ov(xz) HR + ov(yz)HR
= HR - HL - HL + HR = 2 (HR - HL)
Hence, the sum of HR and HL belongs to the A1 irreduci-
ble representation and the difference of HR and HL to the
B2 irreducible representation. By consulting the character
table of C2v symmetry, we have the following correspondence:
lsc -ila1 2P5
A1 2sC —> 3:11 32 ——>lb2 81:2pg —1b1
2p: 1s2 - 15E
. -92a1
ls? + lsé}
Since two orbitals having almost the same energy and symme-
try can form a bond, the 2p; may form a bonding orbital with
H C
lsL + ls: and 2p.y may form another bonding orbital with 153-
lsE. Therefore the electronic configuration of CH2 in the
3
ground state 81 can be described as
2 2 2 3
la1 2a1 3a1 lbz lb1 ( 8])
214
2 251 2p; 2p; 2p; configuration. 0n the
2 2 l l . .
y sz conf1guration,
we will have the electronic configuration of CH2 as,
where carbon has ls
other hand, if carbon uses ls 25
29
2 2 2 2
la1 2a1 3a1 lb2
which is the first excited state of 1A1 symmetry.
Keeping this idea in mind, we will have the following
species:
species ground state first excited state
or+ 1522522p63523p6, l522522p63523p6,
3d5(65) 4s‘3d4(5o)
‘ CH2 la$2a$lbg3a1lb1 la$2a$3a$lb§(]A1)
3
(3,)
There are 23 electrons in Cr+, l8 electrons are in the Ar
core and 5 electrons are valence. There are 8 electrons in
CH2, 6 electrons are in the "core" and there are only 2
"valence" electrons. Therefore, the symmetry adapted core
molecular orbitals can be described as
"core" orbitals Cr+ contributions £52 contribution
la1 ........ .7a1 ls,25,2pz,3s, 3pz lsc, bR + bL
lb1 2b1 2px, 3px none
and leave 7 valence electrons for bonding between Cr+ and
CH2.
215
By symmetry, we have the following states when we
leave Cr+ in C2v symmetry.
states of Cr+ in an states of Cr+ in C2 symmetry
isolated atom V
6 4 6 6 6 6
D (sd ) 2 A1 + 81 + 82 + A2
Note that there are five states in 02v symmetry when
Cr+ is in its first excited state, because 0 has 2 = 2, and
the number of states is given by 22 + l = 5. In other
words, there are five different ways of distributing one s
and d orbitals, each gives corresponding state.
Now, if we leave Cr+ and CH2 in 02v symmetry at w
separation, we will have even more possible states, which
are summarized below:
Cr+ CH2 (Cr - CH2)+ at w separation193
6 3 8 6 4
6s 1111 6111
l l l 2 2
6 l 6 6 6 6
0 A1 2 A1 + 81 + 82 + A2
Since the 1A1 state of CH2 has an electron pair located on
the z axis, when CH2 approached to Cr+, in an effort to make
a bond with it, these paired electrons might experience
strong repulsion form the d electrons of Cr+. We think that
this interaction will be repulsive and it is therefore
216
reasonable to disregard any states arising from the combina-
1
tion of A.' of CH2 with ground state Cr+.
If we consider the first situation, 65 + 3
8B], 6B1 and 48], we will have
81 and the
states which evolve from the
no bond, one bond and possible double bond respectively.
881 - there are 7 unpaired electrons with the same
spin, and it is unlikely to be bound, i.e. no
bond will form.
81 - there are five unpaired electrons, i.e., there
is one singlet couples pair. Hence, the pos-
sible chemical bond at large separation must
be due to the orbital overlap between 3 do of
Cr+ and o orbital of CH2.
B1 - there are three unpaired electrons, i.e., there
are two singlet coupled pairs and hence a pos-
sible double bond exists.
Therefore, the calculation will focus on the 6B1 and
4 6 3
81 states. However, we have to remember that D +
1
B1 and
A1 asymptotes also contribute states of these symme-
tries. Nhile these are noninteracting at w separation, they
may interact strongly as the distance R decreases to the
equilibrium value.
The 681 state arises from one bond formed by Cr (65)
and CH2 (38]). Consider the valence electrons of both Cr+
and CH2 in this state.
217
Cr+ (core) 3d5 l—- 1 1 1 1
M31 b1 b2 a1 a2
1 1
CH2(core) 3— El
1
where do denotes dzz; d
mx denotes dxz and dfly denotes d
YZ’
dA+ denotes dx2_y2 and dA- denotes the dxy orbital with
|m£|= 0,1,2 respectively. Also, the symmetry of each
orbital can be found from the character table. Now, if a
sigma bond is formed between do of Cr+ and po of CH2 (both
have the same irreducible representation a1), then by tak-
ing into account the core orbitals, the molecular orbitals
of CrCH2+ at infinite separation can be described as:
2
core 2 +
8a](do + po) 3b](CH2) 9a](dA )
1a2(dA') 4b1(dnx) 4b2(dfly)
In order to carry out reliable ab initio calculation,
good basis functions have to be chosen for each atom. The
Cr+basis functions used in the CrCH; calculation are from
Nachters, i.e., (l4 s, llp, 6d/ 55 4p 3d). The carbon
basis (lls, 6p/3s 2p) is from Duijneveldt and the hydrogen
basis (45/25) is from Huzenaga. In all 120 primitive gaus-
sians are used to form 48 contracted orbitals over 4 centers.
For convenience, each basis function is given by a number
218
so that they are easy to read. The basis functions used
in CrCH2+ are the following.
l 15 21 xz 4l yc
2 25 22 x2' 42 yé
3 35 23 xz" 43 2C
4 4s 24 yz 44 zé
5 55 25 yz' 45 sHR
6 x 26 yz" 46 SAR
7 x' 27 xx 47 sHL
8 x" 28 xx' 48 SfiL
9 x"' 29 xx"
10 y 30' yy
11 y' 31 yy'
12 y" 32 yy"
13 y"' 33 22
14 z 34 zz'
15 z' 35 22"
16 z" 36 5C
17 z“ 37 sé
18 xy 38 56
19- xy' 39 xC
20 xy" 40 xé
where x, x', x" and x"' basis functions are used to stand for
the px orbitals of chromium, and y, y', y", y"' are for the
py orbitals of chromium, etc° xy, xy', xy", xz, xz', xz"---
219
22" are used to build up the 3d orbitals of chromium.
Number 36 to 44 are used for s, px, p and p2 wave functions
y
of carbon and SHR’ SHR are used for 5 wave functions of
Hydrogen atom HR and SHL’ SAL are used for 5 wave functions
of Hydrogen atom HL‘
Moreover, these basis functions can be catagorized to
four irreducible representations in C2v character table
according to their symmetry characters by the use of the
same method used on page 212. Hence, we have the following
arrangements
1 A1 Symmetry (22) g 82 Symmetry (11)
1 1 23 10
2 3 24 11
4 4 25 12
5 5 26 13
6 14 27 24
7 15 28 25
8 16 29 26
9 17 30 41
10 2 x 33 - 27 - 30 31 42
11 2 x 34 - 28 - 31 32 45 - 47
12 2 X 35 - 29 - 32 33 46 - 48
13 27 - 30 3 Bl Symmetry (9)
14 28 - 31 34 6
15 29 - 32 35 7
16 36 36 8
220
17 37 37 9
18 38 1 38 21
19 43 39 22
20 44 4o _23
21 45 + 47 41 39
22 46 + 48 42 4O
4 A2 symmetry (3)
43 18
44 19
45 20
Thus, the 48 basis functions are reduced to 45 symme-
try orbitals after taking into account the symmetry in a
C2v system. Note that the first and third column label the
orbital numbering which is the computer readable form. For
example, orbital 28 is acutally the basis function 25 (yz')
and is 6b2 in C2v symmetry (Note that each symmetry species
starts at 1). Similarly, orbital 21 is actually the sum of
basis function 45 and 47, namely 5HR SHL’ and is 21 a1 in
C2v symmetry.
Therefore, the molecular electronic configuration of
CrCHZ+ described on page 217
2 2
can be written in an orbital numbering form by referring to
the symmetry orbitals given above as
221
(1 2 3 4 5 6 7 34 35 23 24 25)2 8 8 36 9 43 37 26
or in a ordered form:
CORE: 1 2 3 4 5 6 7 23 24 25 34 35
VALENCE: 8 8 9 26 36 37 43
Table 21 lists the results of ab initio calculations
of CrCH2+ at different electronic states.
3. Discussion
Figure 10, and 11 summarize the energy levels of
different states from the ab initio calculations. At this
level of calculations none of these states seem to form a
bond between Cr+ and CH2 as that reported by Beauchamp148
(65 i 7 kcal/mole). As we can see from these figures that
681 only forms a 12.5 kcal/mole bond at (2 x 2) SCF level,
which includes the o bond correlation. 6
A2 forms a 22.0 kcal/
mole bond relative to excited state of Cr+(60) and ground
state of CH2(381). It doesn't co-relate to ground states of
separated species and is thus not the ground state. On the
other hand, 481 gives a tremendously high energy at SCF
level, reflecting the inadequacy of the SCF theory for this
state. Correlating o and n configurations to allow elec-
trons more freedom, the energy is lowered. In fact, the
lowest energy of this sytem is formed at bond separation
r = 1.85 A. However, its energy, —1082.02832 H shows no
evidence of the presence of strong bond between ground
222
A 83238.8“. 8888
.....smmxisme ...s s.
Nflmmmhfienafieoamefl 88o v +
u wfivflofioov .68: m Hmmodmo? «wed «definaannfimmaflmmawmmfi 280 Z.
93mm 5.3030
5 m 588 . 831 «93 we Nearer.”
{Hammafiarfl .88 T
833.58 98$ :3 a 988.821 «88 8m? x NV
.3 m 88o . $31 «$5 Nflfanpmfieoafm
wemNAanmnmmfifmfaf? 1 1 fig
MI
«0
833.58 83 5+5... x 359.831 «SJ 5m? x 3
Npefemflfoapn
H83 8m x 038.831 «83 memfiwpmmpmmfifmafir? 1 1 Had
9:25.60 “Mono M 533950 ”mm“
mama .8 8033.38 2.35 2 no 39mg . Hm e38
223
m mmmmm.amoa1 «cm.a eumfin x mv
83.3968 82:. + m :8 a 838.831 «mm; 8m? x C
A “hand 265 0....
805033“ mo acumen.“ 23 3033
8838.48 83: + m + 5m x 88.831 «woo 8m? x 3
833488 883 Team x 39.831 «88 cam? x 3
N3 «.5 ado
x «88.831 «8.4.. «Ammfmfampnmfimfifiem 1 1 1 44$
an:
mamfmfnfieoaadmwma 88 Y.
Namfafnffiaamwamfi .88 Y.
«maflpmfipmammawamfi eumm1wm4
vcoesoo a m coapgmcoo N450
9.83 S .538.
4O
36
Figure 10. Energy Levels Of 631 And
224
CrCH§(SBI and 6.4.2)
(Fixed oi Schoefer's Geometry)
C '1089.98I66fi
, 0*(601 +CH2(3B 1’
r- 1 \
I 1
1- \ 34.4 MHEZZchol
I 1
: \
, 140329019;
L ‘ ‘
_ 32.1 MH=24.311ca1 \
1' \
: \ 11
.. mosaousoe 6A
L'IOBQOIQYOfi 2
r Cr"(551+c1-12(38,1\ ; 108202550
: A
: SCF \\ 19.5 MHEIZ.5kcal
L
1- \
F \
t Ugagoaerr 63'
E (2:12) SCF
1-
6.42 States Of
“0112*
225
manage
38m wemv ...: an: 8 :33 9.8 .: paw:
38:
go
.5.
”a
_
59.58:: an .m.
“8
Us
Us
_ m...
. .on
. m9.
_Bxuaiz ..nn , .3.
H 9.
— “No
....mamzufiooro“ on
308.39. .8
1483105
(HWBV
B. 222.8. mom
mom 3.». ..{uzu ..aoiu
...
_
_
.
_
_
.
ILILII
l
3 2.938.-”
_
_
_
_
..mn.«:u+.€+o
111098.59-
lllllllllllllLllllJ
0v
qumxucu
(HWEIV
226
state of Cr+(65) and CH2(3B]), which is in contrast with
Beauchamp's suggestion that CrCHz+ has a double bond.
Beauchamp did suggest that if metal ions promote one elec-
tron into the s orbital, they will form a stronger bond with
a ligand. This is borne out in this calcualtion since 6A2
forms a 22 kcal/mole bond with respect to Cr+(6D) and
CH2(BB]). One should however include d electron correla-
tions at the MCSCF level and then compare with the same
MCSCF level energy of separated species, i.e. Cr+(65) and
3
CH2(uB]).
6 6
Neither the B], nor the A2 state is predicted to have
comparable bond energy to the experimental value, although
6A2 has the most energy lowering relative to excited state
of Cr+(60) and CH2 (38]), which somewhat matches Beauchamp's
suggestion that essentially the bonding between the first
row transition metal ions and ligands is via the metals 5
electron. However, he pointed out that CrCHZ+ doesn't fit
in his plot of promotion energy (coordinate) versus bonding
energy (abscissa). Perhaps the experiments he did do not
refer to the bond energies of these high-spin electronic
states of CrCH2+. However, since the calculation of 431
indicates a failure to form strong bonds, the experimental
value is hard to refer to either ground state or the first
excited state of Cr+ and ground state of CH2. This may sug-
gest another possible structure of CrCH2+, for instance,
+
The
H-Cr+-C-H instead of C v symmetry structure of CrCH2
2
227
possibility of forming bonds from excited state of CH2 has
been overlooked, because when it gets closer to Cr+, the
paired electrons on CH2 will experience a strong repulsive
force. However, to check if it is possible to undergo a
back bonding from d electrons of the Cr+, a 6
A1 state,
derived from excited states of both Cr+(60) and CH2(1A]) is
being calculated. Other states derived from Cr+(60) and
CH2(3B]) have not been considered in this work. The reason
being that other arrangements of four d electrons in five d
orbitals of Cr+ would not make the bonding dramatically dif-
ferent, neither would correlate the d electrons in MCSCF
calculation, since the bonding is mostly contributed from
45 orbital of the Cr+ and 3a1 orbital of the CH2.
Ab initio calculations provide considerable insight in-
to the nature of the chemical bond since we can not only
obtain the bond energies, but also orbital occupancies.
Calculations of this sort might help us to understand the
nature of transition metal insertion processes. For exam-
ple, in order to understand how Cr+ can split up H2 mole-
cule (D(H-H) = 104.2 kcal/mole) to yield a low energy pro-
dUCt’ crH+(D(Cr+-H) = 35 i 4 kcal/mole)148
after putting
the kinetic energy into Cr+. Ab initio calculations of
CrH+ and CrH2+, are in progress.
APPENDICES
APPENDIX A.
Schematic Diagram For ICR Voltages Controls
There are two operating modes in ICR: normal drift mode
and trapping mode.
In the normal drift mode, we want all drift plates and
trap plates of the cell to be +15 V to -l5 V (dc) adjustable.
However, we need to pulse one trapping plate in both the
source and the analyzer regions. Ions will have long
residence times for reaction during positive pulse and are
swept out of cell during the negative pulse. Also, the same
pulsing will form a modulated signal out of the marginal
oscillator, and then is input into the lock-in amplifier to
be in phase with reference wave and consequently enhance the
S/N ratio. This can be done by setting the mode switch to
normal drift mode in Fig. l2. In this case, the monostable
vibrator 74l2l doesn't get a triggering pulse and Q will
output a +5 V dc. As a result, both 2N404A PNP and 2N3053
NPN are switched off and output a +l5 V do as a +GATE pulse.
On the other hand, a will send out a -5v dc on a dar-
lington pair of transistors on the other side, which will
conduct both 2N404A PNP and 2N3053 NPN to output a -lS V dc
as a -GATE pulse. Note that FETs need more than 10 Vs to
drive, a regular TTL voltage cannot serve this purpose.
228
229
Hence, the FET at the bottom in Fig. 13 will always be
ON and the drift potentials (adjustable) relative to the
ground will appear on a drift plate after a voltage follower
741. Four of the same circuit will allow us to adjust four
drift plates separately.
On the other hand, the comparator Bll in Fig. l4 will
send out a pulse train whose pulse width is determined by
voltage inputs to trigger the 7412l so that output 6 will
give a TTL pulse train with pulse width adjustable by
external RC values. When the pulse is at +5 V, both PNP
and NPN transistors don't conduct and will output a +l5 V
at B, which will switch on N-channel FET (2N38l9). Conse-
quently, a reverse setting will then appear on the pulsed
trapping plate (remember we put a unit_gain of inverting
amplifier in the front of 2N3819). When the pulse is at
0V, both PNP and NPN transistors will conduct to output a
-l5 V at point B, which will then turn on p-channel FET
(2N3820) so that the voltage setting will appear on the
pulsed trapping plate. In this way, we can have a voltage
put on the pulsed trapping plate. Meanwhile, the voltage
will always appear on the constant trapping plate.
118, there are three pulsing
In the trapping mode
sequence: trapping, detection and quench. Quench sequence
can be combined in the detection step. For the trapping and
detection period, we need to keep the trapping plates posi-
tively charged on both the source and analyzer. However, in
230
the trapping period, the source drift plates have to be at
ground, and the analyzer drift plates have to be at negative
potentials which can be done by adjusting the trapping poten-
tial as shown in Fig. 13. In the detection period, opera-
tion is set back to drift mode by putting a positive
potential on the top drift plates and a negative potential
on the bottom drift plates by adjusting the drift potential
in Fig. 13. The necessary pulse timing for trapping and
detection is provided again, by the circuit in Fig. 12. In
order to be more versatile, input to pin 3 of the 311 com;
parator can be either a slow ramp for varied trapping and
detection times or a constant voltage as shown in Fig. 16.
When pulse from Q output is at 0V, it will switch on
both PNP and NPN transistors to give -15 V. When the pulse
is at +5 V, it will switch off the PNP and NPN transistors
to give + 15 V at +GATE output as shown in Fig. 16. When
the +GATE reaches +15 V, it will switch on the bottom FET in
Fig. 13 and at the same time the -GATE is at ~15 V to switch
off the top FET. As a result, the drift potential will
appear on drift plates for detecting ions. On the other
hand, when +GATE is at -15 V, it will switch off the bottom
FET. Meanwhile, -GATE will be at + 15 V to switch on the
top gate and the trapping potentials will appear on the
drift plates for trapping the ions.
The trapping voltage in this mode can be set by switch-
ing the timing circuit in Fig. 14 to ground, which will
231
switch on both PNP and NPN transitors and will send out -15
V dc to point B which will drive P-channel FET 2N 3820 only
so that a constant voltage appears on the pulse trapping
plate. 50 now we have the same voltage going to both trap-
ping plates. To be flexible, we can make two of the same
circuits for both source and analyzer.
In the trapping mode, we also need a circuit to pulse
the electron energy, this is done in Fig. 15. 0 output
(pin 6) of a 74121 will send out a pulse train to switch
on and off the 2N 5415 PNP transistor so that it can pulse
the electron energy and form ions, producing a modulated
signal out of detector for S/N enhancement as described
earlier.
As we can see from these cirucit diagrams, all the tim-
ing pulses can be replaced by computer software programming
using DACS. The computer setup for on line data acquisition
is shown in Fig. 17. Where the setting cirucit on the left
bottom is for the mass calibration and circuit after ADC
0817 is for amplifying the signal and tailoring the analog
signal to the safe values (5 5.12 V for PET). The software
has been developed in our laboratoryllg.
232
.eoflomm wfimglhflgoflo 58 8H 3&5 NH 25mg
..o. t. l a: nan
AL“ What ‘8 .n £03.sz 0084
B; so-
233
-GATE
IN
+ 15v +59A
1
-15v 0
(4:5)“) 10K C
2N38L9 74, . To Cell Plate
6 (One for each
TRAPPING 'K ‘ drift Plate)
POTENTIALS ”“3" 975
. 4
$50K
+6ATE
-lSV 3 l :pfio {-ISV
+15v o 1" 459A
; 2~3319
(1;)? (OK (>
u: j".— , L—
J-
I
~ ‘HSV
To - ift plate
-L tap some(UIS)
Figure 13. Mapped Ion Cell Circuitry-m
Drift Section.
234
. 338m gmafillggoflo Sumo :3 damage .3” gm?"
3mg
M535
«539.501
I;
B. 91014»)
x8 w
(«$2. :
@
<33: 2
28a \
J ’1’.) IIPU‘1 b
no. 0
o 1 ..I._
x8 :8. L w“. _ 5.1M“
as.
.06
5....
)9. H3358 .. a
_ a
h...
u
H
...-P
...R
3.1
w
o
>h1.
amfim
235
mean vamsuafim.:ouuooflm .ma onnwflh
@fim >7 9.: who... 8.03
>nal ... an mmofiou
no.5 owdvag no can:
H >2.
g. .n c:
v
. I“. j
mszN 9 . .
. ¢ ”7.! 1
- .. L
r ..nm - .
m8. ...... >3- ...
To «...... .
9:05.38 on. 7 T6
I “a W
H? mm H ...
152 no «339?...
.8925 coupooam >n+
9.8m and...
236
fast lamp input to pin 2 of 311
WWI/AV l/ 1/1/1/1/1/ A /1 {3.311.313.33...
5VIl‘llllllilllli'll.outputofjll
V
O,
H n H n H H H H H output of 74121
Blow ramp
output of 311
II” H H ' H H H Qoutputof74121
o_. .' - -0
Hsv drift(detection)
”i H i i '1‘ n '
-..--1__ _ __-_______4 +GA'IEoutput
* Z. i‘
ppins
Figure 16. Timings 0f tapped Ion Cell Circuitry
In Figure 12
237
.25.“.6523
-o~ .28
.
6H .8 990m Beganfié .S 835
NMO. lv. >._.2
24m 5% xwm
.24me
$0092 >5
.giggafi
7.25 x 0%.
OVON Emu
.
241! Q-
venom NOmm-
92.5 .35 .030 1
.OON hung
659:4 5.-..qu
d 0N. mda
m uoEcm>072 o
003$ 55>
. ..bs.
3.8.8.... .25....
3.3. 0022.9;
.
ooQ..> Egon—u 65> .N.
€32.33.
Appendix B.
Marginal Oscillator Setup In ICR Experiment
The tank circuit of a marginal oscillator is shown
below:
V
R L C
I
—LICR
;J-Cell
The principle of detecting ions is in references 10 and 11.
238
APPENDIX C.
[he Realtionship Between Magnetic Field
And Mass In The ICR Experiment
-22-_-
wc - mc 2nvc
v an
. _ c
B - e
At resonance, Vc = Vm.o
for m = 100 and if vm.o = 153 kHz
then
B = 1.53 x 105 Hz x 100 amy x 1.67 x 10‘27 kg/amu x 2w
1.5 x 10“9 coul
1.0028 W/mz = 10028 c = 10.028 kG
Similarily,
for m = 350, 8 = 3.510 11m2 = 35.1 kG
Therefore, by setting 0 = 153 kHz, at 10 k0 we can
m.o
get a peak corresponding to m/e 100 and at 35 k0, we can
get a peak corresponding mass unit of 350 amu.
239
APPENDIX D.
Alternate CID Circuit For Conventional ICR
Double resonance techniques in ICR can be used to
unambiguously identify the precursors of products. Yet,
sometimes it is hard to determine the structure of product
ions. For example,
N'++CH ——-—>N°(CH)++H
‘ 410 ‘48 2
many possible structurescan be assigned to Ni(C4H8)+:
N142
Max
.0
11—~i+—H
Substitution reactions can solve some of these questions.
However, collisional-induced dissociation (CID)120’121 is a
good technique for distinguishing one structure from another.
In conventional ICR, one can design a circuit which will
excite or eject an ion while scanning the magnetic field.
If that ion we are exciting is the product ion, it will gain
energy from a radiofrequency (r.f.) and will actively collide
240
241
with neutral molecules and then dissociate. We also can
excite any reactant ions to determine how kinetic energy
imparted to them affects the reaction. Fig. l8 shows this
design which was constructed. Note that a ramp (0-l0v) from
the magnet controller is fed into the op-amp to reduce the
ramp to 0-5v, which is the useful range that can be fed
into the VCG of a Wavetek frequency generator. In this way,
the angular frequency of the ion will be linked with magne-
tic field while scanning the magnetic field. A 10 turn pot
is used for slope adjustment. The circuit in the upper
section is for offset about 5v as shown below
10 V
ramp from magnetic controller
0
<
x;
+ offset{i Slope and offset adjustment
- offset{:
The fine adjustment offset value is done by use of a 500 k
pot.
242
6 V Zener Diode
Low Leahge
005 ”f 005 ”f
' l H
k . m: l
on” I
*15" 10k ‘ 15v
+lOv 51 ‘
VHS 5 1% 'Do
..‘v Uave’oek
902 LHoo'ro-oH *5" I»!
K
1
O-lOv .
from
magnetic
controller
Figure 18. Schematic Circuit Diagram Of CID For
Conventional ICR.
Appendix E.
Calculation Of The Collision Frequency For Co+ And .v/O\,,
(l) polarizability16
e(ahc) = %(§TA)2A3
where N is total electrons of molecule
a is polarizability
TA is atomic hybrid components obtained from
Table I in Ref. l6.
[(1:C + 3TH + Tc + 21H)2 + TOJZ
#43
N
0‘c H
4 100
[(1.294 + 3x0.314 + 1.294 + 2x0.314)2 +
#4:-
N
1.290]2
= %7 (92.275) = 8.7889 A3 expt'l = 8.73 A3
(2) Langevin collision rate17
KL = Zne (95)”2 where u is the reduced mass
u
u = ggfigg = 35.63758
-24 3
KL = 2x3.l4x4.8xl0'10x( 8.73xlO cm _24 )1/2
35.63758xl.67xlO g
243
244
l.lS45xl0'9cm3/molecules
(3) collision frequency is KL.N
But
'. N
at
23
% = P(t°rr)X5-°2x‘° = 3.24155x1019 molecules/l2
760x0.082x298
= Px3.24155x1016 mo1ecu1es/cm3 = 3.24155x1016
3
molecules/cm at 25°C
P = 5xlO'6 torr for ethyl ether
N = 5x10'5x3.24155x1016 = 15.20775x1010
molecules /cm3
'. collision frequency = KLN = 1.1545x10‘9x15.20775x
1010 = 187.11898s'1
1 1 3
(4) time between collisions ? = 8 .1 898 = 5.34419xl0
S
Appendix F.
Branching Ratios Of Fe+ Reactions with .~,0\,,
Fe+ +xc,0\,r > Fe(CH20+) + CH
The absolute intensity of each peak and the absolute power'
4
+ +
~———+ Fe(C2H6O ) ~1F-+> (CH20)F9 (CZHGO)
+
drop in the double resonance spectra are arbitary but they
have to be in the same scale.
First, we have to take the peak height for each peak
in question (m/e 86, l02, l32, and l33 in this case)
gig_
85
102
132
133
corrected
ion peak height peak height
Fe+(CH20) .80 9.30
Fe+(C2H60) 2.75 27.0
(CH20)Fe+(C2H60) .58 5.15
(CH30)Fe+(CzH60) 2.53 19.8
where corrected peak height = peak height/mass of the ion.
+ + _
Total Fe (CZHGO) from Fe - 1102+ + 1132+ + 1133+
245
°. % of m/e 133
Double
246
resonance spectra of m/e l32
PUT30
55+
+
'. % of m/z l32
.20
1.35\[ .2dffi \[1 15
84+ 102+ 112+
contribution from Fe(C2H60)+ =
.07
(.30+l.§§+.
Double
20+l.15)
resonance spectra of m/3 l33+
_U:50
50+
+
.60
T.50+4.2+.60+.
50V ADV
4.2\’
84+ 102+ 112+
contribution from Fe+(C2H60) =
10) = 01]
'. total contribution of m/3 132+, l33+ from m/e 102+ is:
1132+(.07)+ 1133+(.ll)
'. branch ratio of m/e 132+
'. branch ratio of m/e l33+
= 5.l5 x .07 + 19.8 x .ll =
34 + 2.19 = 2.54
= .34/2.S4 = .l4
= 2.19/2.54 = .86
'. total contribution of Fe+(C2H60) from Fe+ = 1102+ +
1132+ + 1133+
= 27.0 + (.07) x 1132+ + ( 11) x I133+
= 27.0 +
.07 x 5.15 +
.ll x 19.8
247
= 29.5
The double resonance spectra of m/e 86 appears as
2%]
follows:
+
56
and m/e 102+:
{HAT
l.6
56+ 34+ .31 %contribution from Fe+
_ .15 _ 09
’ .ls+l.5' ' °
m/e 132+, l33+ should be from successive reaction by m/e
102+ and m/e 85+
Fe+ + \,0\,.—r———9 Fe(CH20)+ + 0H4
m/e 86
4————a Fe(C2H60)+ —:lia»m/e 132
m/e 102 ‘86! m/e 133
. . +
Total contr1but1on from Fe = 186+ + (.09) x 1102+
9.302 + .09 x 29.5 = 9.30 + 2.54
11.8
'. branching ratio of m/e 85+ = 9.3/11.8 = .79
'. branching ratio of m/e 102+ = 2.54/11.8 = .21
However, if we assume all m/e 132+, 133+ are from successive
reaction of m/e l02+, then;
Total Fe+(C2H60) from Fe+ = I102+ + 1132+ + 1133+ = 27.0
+ 5.15 + 19.8 = 51.9
248
‘. % of m/e 132+ contribution from Fe+(C2H6O) =
5.15/(5.15 + 19.8) = .21
'. % of m/e 133+ contribution from Fe+(CzH60)
19.77/(5.15 + 19.8) = .79
i.e. 102+ '21; 132+
'79; 133+
Total contribution from Fe+ = 186* + (.09)I]02+ = 9.30 +
(.09) x 51.9 = 13.8
'. branching ratio of 85+ = 9.30/13.8 = .58
‘. branching ratio of 102+ = 4.46/13.8 = .32
.68
i.e. Fe+ +V0v Fe(CH20)+ + CH4
.32
+ .21
Fe(C2H60) -~—— (CH20)Fe+(C2H60)
.79
_—— (CH30)Fe+(C2H60)
REFERENCES
LIST OF REFERENCES
1. T.A. Lehman and M.M. Bursey, Ion Cyclotron Resonance
Spectrometry, J. Wiley and Sons, New York (1976).
2. J.D. Baldeschwieler and 3.5. Woodgate, Acc. Chem. Res.,
.1. 114 (1971).
3. J.D. Baldeschwieler, Science, 1 9, 263 (1968).
4. H. Hartman, K.H. Lebert and K.P. Nanczek, Fortschr.
Chem. Forch., 43, 57 (1973).
5. R.C. Burnier and 8.5. Freiser, J. Chem. Educ., 56, 687
(1979).
6. R. T. McIver, Jr., Sci. Amer., 243, 186 (1980).
7. J.L. Beauchamp, Ann. Rev. Phys. Chem., 22, 527 (1971).
8. J.M.S. Henis, Ion-Molecule Reactions, Vol. 2, J.L.
Franklin, Ed., Plenum Press, New York (1972).
9. D. Hobschall, Rev. Sci. Instr., 36, 466 (1965).
10. R.T. McIver, Jr., Rev. Sci. Instr., 44, 1071 (1973).
11. A. Warnick, L.R. Anders and T.E. Sharp, Rev. Sci.
Instr., 45, 929 (1974).
12. R.T. McIver, Jr., R.L. Hunter, E.B. Ledford, Jr.,
M.J. Locke and T.J. Francl, Int. J. Mass Spec. Ion
Phys., 32, 65 (1981).
13. J. Wronka and D.P. Ridge, Rev. Sci. Instr., 5}, 491
(1982).
14. 1.0. O'Haver, 0. Chem. Educ., 49, A131 (1972); ibid,
19, A211 (1972).
249
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
250
. Goode, A.J. Ferrer-Correia and K.R. Jennings,
. J. Mass. Spec. Ion Phys., 5, 229 (1970).
--'7< H63
0 :0
dc; (+0
. Miller and J.A. Savchik, J. Amer. Chem. Soc.,
___, 7206 (1979).
G. Cioumousis and D.P. Stevenson, J. Chem. Phys., 29,
294 (1958). ‘—
Heats of reaction can be calculated from the group
values for Hf. See s.w. Benson, Thermochemical Kine-
tics, 2nd ed., J. Wiley and Sons, New York(l976).
There are three ways to product metal ion source:
(1) electron impact on metal carbonyls, (2) thermo-
ionic emission (see J.P. Blewett, Phy. Rev., 59, 464
(1936), and D.H. Smith and J.A. Cater, Int. J. Mass
Spect. Ion Phys., 49, 211 (1981), (3) laser desorp-
tion (see R.C. Burnier, J. Amer. Chem. Soc., 191,
7127 (1979), and R.B. Cody and R.C. Burnier, et al.,
Int. J. Mass. Spect. Ion Phys., 33, 37 (1980)). Some
unstable ions can be detected by varying drift volt-
ages on method (1) and (2). Laser ionization techni-
que sometimes gives metal ions having higher kinetic
energies, thereby giving more reactions in some cases.
P.B. Armentrout and J.L. Beauchamp, J. Chem. Phys.,
74, 2819 (1981).
M.S. Foster and J.L. Beauchamp, J. Amer. Chem. Soc.,
9], 4814 (1975).
G.A. Somorjai, Chemistry in Two Dimensions: Surfaces,
The George Fisher Baker Non-Resident Lectureship in
Chemistry at Cornell University, Cornell. Univ. Press.
198] O
R.F. Heck, Organotransition Metal Chemistry, Academic
Press, New York (1974).
H.F. Schaeffer III, Acc. Chem. Res., 19, 287 (1977).
J.R. Anderson, Structure of Metallic Catalysts,
Academic Press, London (1975).
M.G. Thomas and B.F. Freiser, J. Amer. Chem. Soc., 98,
1296 (1976).
E.L. Mutterties, Chem. Eng. News., Aug. 30, 28 (1982).
w.A. Goddard, S.P. Walch, A.K. Rappe, T.H. Upton and
C.F. Melius, J. Vac. Sci. Technol., 14, 416 (1977).
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
251
R.B. Brewington, C.F. Bender and H.F. Schaeffer, J.
Chem. Phys., 66, 905 (1976).
R.R. Schrock, Science, 212, 13 (1983).
R. Burch, Acc. Chem. Res., 16, 24 (1982).
G.A. 02in and A. Vander Voet, Acc. Chem. Res., 6, 313
(1973); ibid, 19, 21 (1977).
Myron L. Corrin, J. Chem. Educ., 66, 210 (1978).
W.E. Billups, Mark M. Kanarski, Robert H. Hauge and
John L. Margrave, Tetra. Lett., 61, 3861 (1980).
P.B. Armentrout and J.L. Beauchamp, Chem. Phys., 48,
315 (1980). ‘T
T. Kametani and K. Fukumoto, Accts. Chem. Res.,‘g, 319
(1976).
P.W. Tiedemann and J.M. Riveros, J. Amer. Chem. Soc.,
‘26, 185 (1974).
C.H. Depuy, J.J. Grabowski and V.M. Bierbaum, Science,
218. 955 (1982).
S.E. Buttril1, Jr., J. Chem. Phys., §2, 6174 (1970).
P.S. Brateman and R.J. Cross, JCS Dalton, 657 (1972).
W. Mowat and G. Wilkinson, J. Organomet. Chem., 36,
C35 (1972).
W. Mowat and G. Wilkins, JCS Dalton, 1120 (1973).
W. Mowat and Shortland, G. Yagupsky, and G. Wilkinson,
JCS Dalton, 533 (1972).
U. Klabunde and G.W. Parshall, J. Amer. Chem. Soc.,
94, 9081 (1972).
Y. Ishii and M. Tsutsu, Fundamental Research in Homo-
geneous Catalysis, Plenum Press, New York (1978).
C.T. Mortimer, Reaction Heats and Bond Strength:
Pergamon Press, New York (1962).
D. Ialage, S. Brown, J. Connor, and H.A. Skinner,
J. Organometal. Chem., 61, 403 (1974).
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
252
K.W. Egger, J. Organometal. Chem., 65, 501 (1970).
M. Brown and R. Puddephatt, J. Chem. Soc. Dalton,
1613 (1974).
M.R. Litzow and T.R. Spalding, Mass Spectrometry of
Inorganic and Organometallic Compounds. Elsevier
Scientific Publishing Co., New York11973).
J.L. Franklin, J.G. Dillard, H.M. Rosenstock, J.T.
Herron, K. Draxl, and F.H. Field, Ionization Poten-
tials and Heats of Formation of Gaseous Positive Ions,
NSRDS-NBS, 26 (1969).
R.E. Winters and R.W. Kiser, J. Phys. Chem., 66, 3198
(1965).
A. Foffani, S. Pignataro, G. Distefane, and G. Innorta,
J. Organometal. Chem., 1, 473 (1967).
J. Charalambous, Ed., Mass Spectrometry of Metal Com-
pounds, Butterworths, Boston (1975).
C.S. Kraihanyel, J.J. Conville, and J.E. Sturm, Chem.
Comm., 159 (1971).
J.B. Westmore, Chem. Rev., 16, 695 (1976).
S.M. Schildscrout, J. Phys. Chem., 66, 2834 (1976).
J. Muller, Angew. Chem. Int. Ed. Eng., 11, 653 (1972).
D.E. Games, A.H. Jackson, and K. Taylor, J. Organo-
metal. Chem., 66, 345 (1975).
R.D. Macfarlane and D.F. Torgerson, Science, 1_l, 920
(1976).
M.A. Posthumus, P.G. Kistemaker, and H.L.C. Meuzelaar,
Anal. Chem., 66, 985 (1978).
J. Pierce, K.L. Busch, R.A. Walton and R.C. Cooks, J.
Am. Chem. Soc., 196, 2583 (1981).
F. Marcel and J.G. Roustan, Org. Mass. Spect., 11,
173 (1982).
W.P. Anderson, N. Hsu, C.W. Stranger, and B. Munson,
Organometal. Chem., 9, 249 (1974).
. Hunt, J.W. Russel, and R.L. Torian, J. Organo-
J
D F.
metal. Chem., 46, 175 (1972).
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
253
R. Dunbar and J. Hutchinson, J. Amer. Chem. Soc.,
3816 (1974).
G. Distefano, J. Res. -NBS-A. 74A, 233 (1970).
J. MUller, and K. Fenderl, Chem. Ber., 165, 2199
(1971).
9_6_.
M.S. Foster and J.L. Beauchamp, J. Amer. Chem. Soc.,
93, 4924 (1971).
M.S. Foster and J.L. Beauchamp, J. Amer. Chem. Soc
97, 4808 (1975).
J. Wronka and D.P. Ridge, Int. J. Mass. Spect. Ion
Phys., 66, 23 (1982).
R.C. Dunbar, J.F. Ennever, and J.P. Fackler, Jr.,
Inorg. Chem., 16, 2734 (1973).
S.M. Schildcrout, J. Amer. Chem. Soc., 66, 3846,
(1973).
R.R. Corderman and J.L. Beauchamp, Inorg. Chem., 1
665 (1976).
M.S. Foster and J.L. Beauchamp, J. Amer. Chem. Soc
21, 4814 (1975).
F. Schumacher and R. Taubenset, Helv. Chim. Acta.,
1525 (1966).
J. Muller, Advances In Mass Spectrometry, Vol. 6,
_5.
11.
A.R. West Ed., Appl. Science Publishers, Ltd. England,
823 (1974).
J. MUller, W. Holzinger and W. Kalbfus, J. Organomet.
Chem., 2], 213 (1975
J. MUller and K. Fenderl, Chem. Ber., 166, 2207 (l
971).
0. MUller and w. Goll, Chem. Ber., 129, 1129 (1973).
R.R. Cordeman and J.L. Beauchamp, J. Amer. Chem. Soc.,
96, 3998 (1975).
R.R. Corderman and J.L. Beauchamp, J. Amer. Chem.
66, 5700 (1976).
R.D. Wieting, R.H. Staley, and J.L. Beauchamp, J.
Chem. Soc., 97, 924 (1975).
Soc.,
Amer.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
254
J. Allison and D.P. Ridge, J. Amer. Chem. Soc., 66,
7445 (1976).
J. Allison and D.P. Ridge, J. Amer. Chem. Soc., 1_l,
4998 (1979).
8.0. Radecki and J. Allison, submitted to J. Amer.
Chem. Soc.
R.C. Burnier, G.D. Byrd and B.S. Freiser, J. Amer.
Chem. Soc., 166, 4360 (1981).
J.S. Uppal and R.H. Staley, J. Amer. Chem. Soc., l_6,
4144 (1980).
J.S. Uppal, D.E. Johnson and R.H. Staley, J. Amer.
Chem. Soc., 166, 508 (1981).
J. Allison and D.P. Ridge, J. Amer. Chem. Soc., 66,
35 (1977).
J. Allison and D.P. Ridge, J. Amer. Chem. Soc., [_9,
163 (1978).
J. Allison, Ph.D. Dissertation, University of Delaware,
1976.
6.0. Byrd, B.G. Burnier and B.S. Freiser, J. Amer.
Chem. Soc., 165, 3565 (1982).
P.B. Armentrout and J.L. Beauchamp, J. Amer. Chem.
Soc., 166, 1736 (1980).
P.B. Armentrout and J.L. Beauchamp, J. Amer. Chem.
Soc., 166, 784 (1981).
P.B. Armentrout and J.L. Beauchamp, J. Amer. Chem.
Soc., 166, 6628 (1981).
P.B. Armentrout, L.F. Halle and J.L. Beauchamp, J.
Amer. Chem. Soc., 166, 6624 (1981).
R.W. Jones and R.H. Staley, J. Amer. Chem. Soc., 1_g,
3794 (1980).
R.C. Burnier, G.D. Byrd, and B.S. Freiser, Anal. Chem.,
66, 1641 (1980).
M. Lombarski and J. Allison, Int. J. Mass Spect. Ion
Phys., 66, 281 (1983).
255
101. R.T. Morrison and R.N. Boyd, Organic Chemistry, 3rd
Ed., Allyn and Bacon Inc., Boston, Massachusetts
(1975).
102. R.A. Bartsch, Acc. Chem. Res., 6, 239 (1975).
103. G.A. Olah, Acc. Chem. Res., 6, 41 (1976).
104. F.A. Carey and R.J. Sundberg, Advanced Organic
Chemistry, Plenum Press, New York (1977).
105. A. Pullman, Giessness-Prettre and Yu. V. Kruglyak,
Chem. Phys. Lett., 66, 156 (1975).
106. G. Wipff, P. Weiner and P. Kollman, J. Amer. Chem.
Soc., 196, 3249 (1982).
107. E.E. Astrup, Acta. Chem. Scand. 919, 85 (1980).
108. J.D. Dunitz and P. Seiler, Acta. Cryst., 696, 2739
(1974).
109. M. Dobler, J.D. Dunitz, and P. Seiler, Acta. Cryst.,
696, 2741 (1974). ‘
110. P. Groth, Acta. Chem. Scand., 666, 279 (1981).
111. J. Dale, Israd. J. Chem., 69, 3 (1980).
112. J. Dale, Tetrahedron 99, 1683 (1974).
113. D.K. Gabbiness and D.W. Margerum, J. Amer. Chem. Soc.,
91, 6540 (1969).
114. D.A. Phipps, Metals and Metabolism, Oxford University
Press, (1977).
115. J.L. Hall and-D.A. Baker, Cell Membranes and Ion
Transport: Integrated Themes In Biology, Longmans,
Harlow, U.K., (1977).
116. A.I. Popov, Pure & Appl. Chem” 61, 101 (1979).
117. N.S. Poonia and A.V. Bajai, Chem. Rev. 16, 389 (1979).
118. T.B. McMahon and J.L. Beauchamp, Rev. Sci. Inst., 96,
509 (1972).
119. S. McElvany, P. Hardebeck and J. Allison, 30th Annaul
Conference on Mass Spectrometry and Allied Topics,
p. 848 (1982).
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
256
R.B. Cody, R.C. Burnier and B.S. Freiser, Anal.
Chem., 66, 96 (1982).
R.B. Cody, R.C. Burnier, C.J. Cassady and B.S.
Freiser, Anal. Chem., 69, 2225 (1982).
A. Tsarbopoulos, J. Allison, unpublished results.
L.F. Halle, P.B. Armentrout, J. Amer. Chem. Soc.,
196, 6501 (1981).
M.M. Kappes, R.W. Jones and R.H. Staley, J. Amer.
Chem. Soc., 199, 888 (1982).
W.B. Pearson, Structure Rgports, Vol. 28, p. 666,
Int. Union of Crystallography (1963).
M. Davis and 0. Hassel, Acta. Chem. Scand., 11, 1181
(1963).
T. Bjorvatten, Act. Chem. Scan., g;, 1109 (1969).
0. Hassel and J. Hvoslef, Acta. Chem. Scan., 8, 1953
(1954). ’
6. Hassel and J. Hvoslef, Acta. Chem. Scan., 6, 873
. Dekok and C. Romers, Rec. Trav. Chim., 66, 313
. Elie1 and Sr. M.C. Knoeber, J. Amer. Chem. Soc.,
3444 (1968).
E.L. Eliel, Acc. Chem. Res., 6, 1 (1970).
R.R. Whitney and D.A. Jaeger, Org. Mass Spect., 15,
343 (1980).
Based on the AHf (12-crown-4) = -150.8 kcal/mole as per,
from ref. 136; and AHf(1,4-dioxane) = -85 kcal/mole as
per, from ref. 135.
All thermodynamical values, unless otherwise noted were
obtained from H.M. Rosentock, K. Draxl, B.W. Steiner,
J.T. Herron, J. Phys. Chem. Ref. Data, 6 (1977).
K. Bystrom and M. Mansson, J. Chem. Soc. Perkin II,
565 (1982).
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
257
See for example, (1) P. Groth, Acta. Chem. Scan., §_A,
721 (1981), ibid, 35A, 541 (1981), ibid, 669, 109
(1982). (2) J.D. Dunitz and P. Seiler, Acta. Cryst.,
999, 2733 (1974).
J.D. Dunitz and P. Seiler, Acta. Cryst., 2739 (1974).
P.E. Arte, J.F. Dupont, J.P. Declereq,
Merrssche, Acta Cryst., 35B, 1215 (l 79
G.M. Van
rc
(1
R.B. Freas, D.P. Ridge, J. Amer. Chem. Soc., l_6,
7129 (1980).
G.
)0
P.J.F. Dupont, E. Arte, J.P. Decle G.G.M. Van
Meerssche, Act. Cryst., 35B, 1217 ).
0.0
79
J.L. Franklin and J.G. Dillard, NSRDS-NBS, 26 (1969).
S.K. Huang, Y.C. Lee, J. Allison and A.I. Popov,
Spect. Leet., 16, 215 (1983).
S.K. Huang, J. Allison, accepted for publication,
Organometallics.
Hydrogen shifts may be sequential, while we prefer
to simultaneous double insertion. Sequential inser-
tion would avoid intermediates with unusually high
formal oxidation state for the metal.
Based on the calculation of Pauling's formula. See
B.E. Douglas and D.H. McDaniel, Concepts and Model;
of Inorganic Chemistry, p. 109, Blaisdell Publ Co.,
Néw York (1965).
L. Pauling, The Natureof the Chemical Bond, Cornell
University Press, Ithaca, New York (1960).
L.F. Halle, P.B. Armentrout, and J.L. Beauchamp, J.
Amer. Chem. Soc., 199, 962 (1981).
Y.C. Lee, A.I. Popov, and J. Allison, Accepted for
publication, Int. J. Mass. Spect. Ion Phys.
P.R. Scott and W.G. Richards, Molecular Spectroscopy.
Vol. 4, Specialist Periodical Reports; the Chemical
Society, Burlington House, London WIV OBN (1976).
C.W. Bauschlicher Jr., and S.P. Walch, J. Chem. Phys.,
16, 4560 (1982).
S.P. Walch and W. Bauschlicher Jr., Chem. Phys. Lett.,
86, 66 (1982).
169.
170.
171.
172.
173.
174.
175.
176.
177.
178.
179.
180.
181.
182.
183.
184.
185.
258
R.V. Hodges, P.B. Armentrout and J.L. Beauchamp,
Int. J. Mass Spect. Ion Phys., 66, 375 (1979).
J.S. Uppal and R.H. Staley, J. Amer. Chem. Soc., 104,
1229 (1982). 7"
T.H. Morton and J.L. Beauchamp, J. Amer. Chem. Soc.,
.6], 2355 (1975).
W.G. Richards, P.R. Scott, E.A. Colbourn and A.F.
Marchington, Bibliography of Ab Initio Molecular
Wave Functions, Oxford University Press (1978).
H.F. Schaeffer III, The Electronic Structure of Atoms
and Molecules: A Survey of Rigorous Quantum Mechanil
cal Results, Addison-Wesley Pub. Co. (1972).
H.F. Schaeffer III, Applications of Electronic Struc-
ture Theogy (Modern Theoretical Chemistry Vol. 4):
PTenum Press, New York (1977).
H.F. Schaeffer III., Methods of Electronic Theory
(Modern Theoretical Chemistry Vol. 3)., Plenum Press,
New York, (1976).~
W.G. Richards and J.A. Horsley, Ab Initio Molecular
Orbital Calculations for Chemists, Oxford University
Press (1970).
D.B. Cook, Ab Initio Calculations in Chemistry, Butter-
worths, Londfin (1974).
C.C.J. Roothaan, Rev. Mod. Phys., 66, 179 (1960).
R.K. Nesbet, PrOC. R. Soc., 230A, 312, 322 (1955).
P.0. Lowdin, Adv. Chem. Phys., 63 207 (1959).
J.H. Wilkinson, The Algebraic Eigenvalue Problem,
Oxford University Press, London (1965).
E. Clementi and A. Veillard, J. Chem. Phys., 69, 3050
(1966).
E. Clementi and D.R. Davis, J. Comput. Phys., 63 223
(1967).
J.L. Whitten, J. Chem. Phys., 66, 359 (1966).
C. Salez and A. Veilland, Thev. Chim. Acta., 11, 441
(1968).
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
259
All the references can be found from : W.G. Richards,
Bibliography of Ab Initio Molecular Wave Functions,
Oxford University Press (1978).
S.K. Huang, A. Lavarado, J. Allison and J. Harrison,
31st Annual Conference of Mass Spectrometry and Allied
Topics, Boston (1983).
M.A. Vincent, Y. Yoshloka and H.F. Schaeffer III, J.
Phys. Chem., 66, 3905 (1982).
W.C. Swope and H.F. Schaeffer III, M01. Phys., 69,
1037 (1977).
C.E. Moore, Atomic Energy Levels: U.S. Government
Printing Office, Washington, D.C. 1952, Vol. 2, NBS
Circular 467.
J.S. Uppal and R.H. Staley, J. Amer. Chem. Soc., 1 4,
1235 (1982).
R.W. Jones and R.H. Staley, J. Amer. Chem. Soc., 196,
2296 (1982).
M.M.
1913
as and R.H. Staley, J. Amer. Chem. Soc., 104,
M.M. Kappes and R.H. Staley, J. Amer. Chem. Soc., 196,
1819 (1982)
J.S. Uppal and R.H. Staley, J. Amer. Chem. Soc., 196,
1238 (1982).
K
(
S.P. Walch and W.A. Goddard III, J. Amer. Chem. Soc.,
66, 7908 (1976).
L.F. Halle, P.B. Armentrout and J.L. Beauchamp, Organo-
metallics. l, 963 (1982).
A.B. Kunz, M.P. Guse and R.J. Blint, J. Phys.,_66,
1358 (1975).
J. Allison, et al., unpublished results.
0.8. Jacobson and B.S. Freiser, J. Amer. Chem. Soc.,
196, 736 (1983).
See for example, J.D. Lay Jr., and M.L. Gross, J.C.S.
Chem. Comm., 970 (1982); R.D. Bowen, D.H. Williams,
H. Schwarz and C. Wesdemiotis, J. Amer. Chem. Soc.,
191, 4681 (1979).
260
186. R.K. Nesbet, Rev. Mod. Phys., 66, 272 (1960).
187. R.K. Nesbet, J. Chem. Phys., _6, 3619 (1964).
188. R.C. Raffenetti, J. Chem. Phys., 96, 4452 (1973).
189. R.C. Raffenetti, Chem. Phys. Leet., 69, 335 (1973).
190. K. Ruedenberg, L.M. Cheung, and S.T. Elbert, Int. J.
Quan. Chem., 16, 1069 (1980).
191. For more complicated molecules, please refer to ref.
176, chapter 10.
192. F.A. Cotton, Chemical Applications of Grogp Theory,
2nd Ed., Wiley-Interscience, New York (1971).
193. Direct Product of 65 (6A in C2v Symmetry) and B1 will
give B1 state, see ref. 192.
194. K.A. Kalmbach, D.P. Ridge, D. Peake and M.L. Gross,
31st Annual Conference on Mass Spectrometry and
Allied Topics, Boston. (1983).
195. J. Allison and D.P. Ridge, J. Organomet. Chem., 66,
C11 (1975).
196. J. Allison and D.P. Ridge, J. Amer. Chem. Soc., 96,
7445 (1976).
197. ?.V. 6odges and J.L. Beauchamp, Anal. Chem., 16, 825
1976 .