Gas-phase ion-molecule chemistry of borate and … recent review of the Hammett substituent con stantsindicatesthat - BXzgroupshaverelatively small polar/field and resonance interactions
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Gas-Phase Ion-Molecule Chemistry of Borateand Boronate Esters
Jeffrey P. Kiplinger*Department of Chemistry, Indiana University, Bloomington, Indiana, USA
Catherine A. Crowder," Daniel N. Sorensen} and John E. BartmessDepartment of Chemistry, University of Tennessee, Knoxville, Tennessee, USA
Borate esters B(OR)3 and boronate esters RB(OR)z undergo ion-molecule reactions to yieldboth addition products (by an implied radiative emission mechanism), ligand exchange, andproton transfer products, in both positive and negative ion modes. Although an acidity forCH 3B(ORh could not be determined, HOB(OR)z has an acidity between acetaldehyde andnitromethane, In light of the negligible polar electron acceptor properties of the ~B(OR)zgroup, that functionality must therefore be one of the best resonance electron acceptorgroups known, almost half again as effective as the nitro group. (j Am Soc Mass Speetrom1994,5,169-176)
1The lack of accessible d orbitals for bonding in
boron makes this resonance interaction even more intriguing. Calculations on vinylboronate esters show ahigh degree of resonance through the carbon- boron
The ability of boron to act as a Lewis acid via itsempty p orbital is well known [1). The ability ofthis empty orbital to act as a stabilizing function
ality for Brensted acids, by charge delocalization froman adjacent anion, as shown in 1, has not been asextensively explored. The isoelcctronic relationship thatthe boron atom has to electron-deficient carbon (i.e.,R3Band R3C+, RB(OH)2, and RC02H, etc.) provides auseful rationale for this ability. Behavior of this lattertype has not been widely investigated in solution,because solvents polar enough to solvate such anionicspecies are in general also good Lewis bases and willcomplex with the boron, thus destroying the possibility of internal charge stabilization by resonance.
xI x-
CH -B--R +-3 I
R
R RIx~ 1
CH -B ---+ -CH -B3 \ z \
R I ~II
XH + CH 2=B-\R
bond [Ib], predicting the strong electrophilic reactivityof the resulting electron-deficient double bond.
Boron-stabilized carbanions have been used in solution as synthons for a-hydroxy carbanions [2-6). Theseare generally produced irreversibly; other than thesimple bracketing based on how strong an alkyllithiumbase is necessary to deprotonate these, there are fewsolution phase pKa data [7]. Two borinate groups willstabilize a carbanion about as much as three phenylgroups (3). Only one case is known where a carbanionis stabilized by a single borinate group [8). Such anionshave been observed in the gas phase as well, both fromlow energy electron impact on triethylborane (9) andby deprotonation of trimethylborane by various an~
ionic bases [10). In the latter study, a rough bracketingof the acidity of (CH3)3B proved possible; and thevalue obtained (~Gacid = 360 ± 5 kcal /rnol) [11) illustrates that boron-containing functionalities can exhibitgreater electron accepting ability than is indicated bytheir polar substituent parameters. Likewise, alkylborane 2 gives an (M - H)- ion on deprotonation byvarious bases, with a bracketed acidity of ~Gactd =
362.2 ± 4.0 kcalyrnol [12).
Me CH=CH\1 \
C BCH1 \ 1-3
Me CH=CH
2
• Current address: Pfizer Inc., Eastern Point Road, Groton, CT 06340.t Current address: EG & G Idaho Inc., P.O. Box 1625, Mailstop 4107,
Idaho Falls, ID 83415.t Current address: Department of Chemistry, Pikeville College,
Pikeville, KY 41501.Address reprint requests to John E. Bartmess, Department of Chem
istry, University of Tennessee, Knoxville, TN 37996-1600.
The polar effect of a boron group must also beconsidered. Boron is more electropositive than carbon,but slightly less so than hydrogen. Based on 19 F NMRshifts, a polar substituent constant UI of -0.056 isreported for - BMez, and - 0.07 for - B(OMe)2 (13).
Received July 8,1993Revised September 30,1993
Accepted October 5, 1993
170 KIPLINGER ET AL. J Am Soc Mass Spectrom 1994. 5. 169-176
The 0'1 substituent parameter for the - B(OnBu)zgroup, as derived from the infrared frequencies of the0-H stretch in hydrogen bonding from phenol tothe borate ester, is comparable to hydrogen with (T ~0.0, and slightly electron accepting relative to methyl[14]. The - B(nBu)2 group is between hydrogen and- B(OEt)z in its polar effect [14]. The polar effect ofthese groups is thus small and slightly electron donating; any major stabilizing effect of boron attached to ananion is not due to a polar effect.
If other gas-phase acidities [11] for carbon acids ofthe CH 3-G type, where -G is an electron acceptinggroup such as nitro, cyano, carbonyl, etc., are correlated with the substituent constants 0;, (polarizabilitv),(TF (field or inductive effect), and (TR (resonance) [15,16], eq 1 is obtained.
By using (TF(-BMe z) ~ -0.056 from above and0;, = - 0.4 (comparable to - NMe z), then the abovecited gas-phase acidity for trimethvlborane gives(TR( - BMez) = 0.27, comparable to the largest knownfor any - G, and half again the size of the value for- NO z. It is evident that the 11" electron acceptingability of boron containing substituents can be quitelarge.
A recent review of the Hammett substituent constants indicates that - BXz groups have relatively smallpolar/field and resonance interactions «(Tr- and (TR <0.1) with a benzene ring for X ~ - Me and -OMe,and larger electron accepting effects «(IF and (IR 0.17to 0.23) for X = halogen [17]. (The parameters (TI and(Ip are essentially equal numerically.)
There have been several reports [18-20] of borateand borinate ion chemistry in the gas phase. There areseveral revealing differences between the chemistrythat we observe and that reported by others [19],which we believe are due to the different time scales ofdrift versus trapped cell ion cyclotron resonance (ICR)spectrometry. There have been other studies concerning the ionic Lewis acid chemistry of boranes in thegas phase [2l], but little systematic work on theBrensted chemistry. Borate and boronate derivativeshave been used in mass spectrometry [22]. To furtherexamine the ability of boron to stabilize adjacent anions, as well as to examine reactivity related to theLewis acidity of such species in the gas phase, we haveinvestigated the gas-phase ion chemistry of a numberof borate and boronate esters by using pulsed ICRspectrometry.
Experimental
All experiments were carried out by using an ICRspectrometer which has been previously described [23].Much of the work in this study was carried out byusing a computer controlled system for the ICR spectrometer, operating under either intermediate passage
conditions [24] for quantitation of peak intensities, orFourier transform conditions for mass identification.The ICR cell is a cubic cell utilizing a capacitancebridge detector [24]. The Fourier transform capabilitiesof this instrumentation involve a partial IorSpec datasystem, interfaced to an IBM 8088 PC-l computer bymeans of a Tecmar Labmaster board. We are currentlylimited to 32K transforms due to memory limitationsof the Pc. Typical conditions for data acquisition include 7 V p-p chirp excitation from a Rockland frequency synthesizer over the frequency range corresponding to the masses of interest, digitization of thesignal at up to 2 MHz as appropriate by the Nyquistcriterion, and fast Fourier transform with the MicroWay 87FFT software (Program 87FFT™, Ver. 1.1.2,MicroWay Inc., Kingston, MA), after one zero fill andBlackman-Harris apodization [25]. Neutral gas pressures were generally in the range of 5 X 10 -7 to 5 X
1O~6 torr. Spectra are presented in magnitude mode.In general, the instrument is tuned by using rapid scanconditions, for real-time feedback of the effect ofchanging parameters, then Fourier transform spectraare acquired to survey what reactions are occurring.Finally, rate and equilibrium constants are determinedby using rapid scan detection, which is more accuratefor signal intensities than Fourier transform detection[26].
Trimethylborate and triethylborate were obtainedcommercially (Aldrich Chemical Co., Milwaukee, WI)and were distilled under nitrogen prior to use. Thealkylboronate esters 4 and 5 were synthesized by thegeneral method of Matteson et aL [5] and were usedafter distillation of the samples over a 0.5 °C range(compound 4, b.p. 132.5-133.0 DC; compound 5, b.p.151.5-152.0 DC). Alkoxide anions were generated bydissociative electron attachment to the correspondingnitrite esters, which were prepared in situ [27]. Hydroxide and amide ions were generated by electronimpact on water and ammonia, respectively. All compounds were subjected to the usual freeze-pump-thawcycles on the ICR foreline immediately prior to use. Allreaction sequences were confirmed by double resonance techniques [28].
Results
Positive [on Chemistry
Borate esters. Conventional mass spectrometric studies of trimethylborate [14] and triethylborate [29] havedocumented the unimolecular fragmentation patternsof these compounds under electron impact. At moderately low pressures (ca. 5 x 10-7 torr) and short reaction times (50 msec). we observe essentially the sameions as seen by Hettich et al. [18]. In the ICR spectrometer, the unquenched spectrum of trimethylborate, corresponding roughly to long reaction times [30], showsonly two ion-molecule products. The major production is the cluster of masses from m/z 175 to 177,
J Am Soc Mass Spectrom 1994,5, 169-176
corresponding to two borons or (2M - 31)+. By doubleresonance, this is the result of addition of the parentradical cation to the neutral molecule with subsequentstabilizing loss of a methoxy radical. The minor product ion, (2M - 45)+, is 2% ofthe size ofthe (2M - 31)+peak. It is the result of an analogous process in whichthe primary fragment ion (M - 15)+' adds to the neutral borate, followed by loss of a CH 20 group. Allother primary fragment ions disappear by rapid reaction via charge transfer to the neutral borate. This issimilar to the pattern of reactivity seen in high pressure chemical ionization mass spectrometry usingtrimethylborate as a reagent gas [22].
Triethylborate has a primary fragmentation spectrum somewhat more complicated than the methylester [18]. Double resonance indicates that the processes giving rise to these peaks are all very similar:each primary fragment ion A+ or the parent radicalcation adds to the neutral borate, followed by loss ofsome neutral fragment, usually the ethoxy radical.Again, we observe essentially the same reactivity pattern as Hettich et al. [18].
The reactions of neutral borate esters with a varietyof ionic reagents are summarized in Table 1. Twomajor reaction pathways are observed: proton transfer,and nucleophilic attack at boron, with accompanyingloss of HOR. This latter pathway is effectively a ligandexchange at boron, but with the charge transferred as a
ION CHEMISTRY OF BORATE AND BORONATE ESTERS 171
proton to the boron compound. Reaction with CH~
results in proton transfer only, because the conjugatebase of the cationic acid is nonnucleophilic. Protonatedwater, methanol, and ethanol give both simple protontransfer forming (RO)2B - O+(H)R, and ligand exchange plus proton transfer. Acids weaker thanprotonated ethanol give only ligand exchange withthe proton ending up on the most basic site, usually the new boron species. Even the neutral borateester can act as a ligand source: B(OMe)3 reacts with[(EtO)2B]20Et+ with ligand interchange.
It is not trivial to determine the gas-phase basicityof the borate esters, because of this facile ligand interchange. MNDO calculations [31] predict that the gasphase basicity of B(OMe)3 should be comparable tothat of methanol, and the basicity of B(OEt)3 should becomparable to that of ethanol. Thus, reaction 2 involving both proton transfer and ligand exchange shouldbe approximately thermoneutral for any R' group:
(ROhB + R'OHi ~ (ROhBO(H)R'++ ROH (2)
We observe appreciable amounts of both protonatedborate and protonated alcohol for R = R' = Me or Etin reaction 2. For B(OMe)3 and methanol, the equilibrium appreciably favors the protonated borate ester,though an exact equilibrium constant could not beobtained. When acetonitrile, comparable in basicity to
Table 1. Reaction products of B(OR)3with various ions
a Neutral plus proton or hydride.b Neutral B(OR)3 + ion - ROH.CTertiary ion from reaction of (MeO)2BO- with neutral.
172 KIPLINGER ET AL. I Am Soc Mass Spectrom 1994, 5,169-176
7.06.0
O-CHI \ 2
R-B CMe 2\ /0-CH2
4, R ~ Me
5, R = Et
1.0 2.0 3.0 4.0 5.0
P~e6su~e. microtorr
0-C(Me)2
Me\ I0-C(Me)2
3
.o+---~--~--~-~--~--~-~. c
2.0
Figure 1. Bimolecular rate constants for formation of [M Mj+of 3, as a function of pressure.
At the low neutral pressures used in the ICR spectrometer, simple addition of an ion to a neutral speciesrequires some method of disposal of the excess energyof bonding or clustering to stabilize the product withrespect to unimolecular dissociation [34]. Such relaxation of the addition complex must be either a collisional or a radiative process [35]. The observedboronate addition complexes exhibit bimolecular rateconstants for formation that are not appreciably affected by changes in the neutral pressure, as shown for5 in Figure 1. At 10-6 torr, there is an ion-moleculecollision on the order of every 30 msec. Although wecannot observe emitted radiation directly, it has previously been shown that this lack of pressure dependence may be taken as good evidence for radiativerelaxation [35]. For a radiative relaxation process tooccur, the complex must have a rate of decompositioncomparable to or slower than its rate of radiative
Scheme I
other end to the boron, as shown in Scheme I.
u 1.~
"m"'"0o~ 1.00E-,c
" .11
8~
these alcohols [11], is added to the ICR vacuum system, the ligand exchange reaction 3:
Boronate esters. The reactions of the cyclic boronateesters 3-5 can be summarized as shown in Scheme I:simple addition of the ion M+ . to the neutral boronate,as well as ions arising from 1055 of fragments (C3HsOand C6HI20 for 3, CH 20 and CSH 100 for 4 and 5)from the addition product. For 4 and 5, there are alsoions from comparable pathways, from the precursorion (M - 15)+ from 4 and (M - 29)+ from 5. Theobserved products arise from pathways comparable tothose of the borate esters' reactions with neutrals, butwithout the complications of ligand exchange. Themajor difference between the positive ion chemistry ofthe boronates and the borates is that stable ion-molecule addition products M· M+ - are observed with theboronates, presumably because these compounds, unlike the berates, cannot lose a simple alkoxy radical tostabilize the addition complex. Cleavage of a B-Obond can occur, but the ligand is still attached via its
is observed. Similarly, reaction of NHt or nPrNHtwith B(OMe)3 results in ligand exchange and loss ofMeOH. This is intriguing in that endothermic protontransfer (approximately 33 kcaljmoI) from the propylammonium ion to the borate ester must be the firststep in such a mechanism, followed by nucleophilicattack at boron, then loss of methanol. This endothermic proton transfer must be driven by the complexation energy of the ion and neutral; for the protontransfer to occur, part of the Lewis acid/base complexation energy must be present at the end of protontransfer, or the reaction would be too slow to observe.AMI calculations [32] indicate that the complexationenthalpy of 1-aminopropane to protonated B(OMe)3 isca. -54 kcal z'mol. Although considerably larger thanmost ion-molecule complexation energies, this is consistent with the covalent N - B bond formed in thisstep.
Use of protonated benzene as a cationic acid with arelatively nonnucleophilic conjugate base results inestablishment of equilibrium with no ligand exchange,with the borate ester 1.1 kcalyrnol more basic thanbenzene. This yields a ,iGbase for B(OMe)3 of 175.7kcal Zmol (proton affinity = 181.9 kcaljmol, derivedfrom a ,iSba", calculated from only the change insymmetry numbers) relative to benzene (gas-phasebasicity = 174.6 kcalyrnol [9]). This is consistent withthe above results using methanol as abase/nucleophile; methanol's gas-phase basicity of 174.1kcal /mol is only 1.6 kcaljmol weaker than that of(MeO)3B. It also is consistent with the reported protonaffinity for (HO)3B of 175.6 kcaljmol [33].
J Am Soc Mass Spectrom 1994, 5, 169-176
emission. In the case of the boronate esters, a longcomplex lifetime is easily rationalized. Assuming thatformation of the addition complex is accompanied bycleavage of a B- 0 bond, this creates a large andunconstrained group which adds additional torsionsand internal rotations to the complex, greatly increasing the number and density of vibrational states. Inaddition to this, a complex of this type should have arelatively fast radiative emission rate; theory indicatesthat compounds having intense infrared absorbancebands in the 1000 to 1300 ern-I region [35] are likely tohave large radiative emission rates. The B-O bondhas one of the strongest known infrared intensities(large matrix dipole moment) and is in the region of1350 cm- I . The complexes formed with the boronates3-5 have three B-O bonds each, and are thus likelycandidates for radiative relaxation.
Negative Ion ChemistryBorate esters. Trimethylborate and triethylborate reactwith methoxide to yield both (M + H)- ions and(M + 31)- addition products, as given in Table 1. Theformer arise from hydride transfer from CH 30 - orHNO- to the boron atom; this has been observed inreactions between methoxide and other Lewis acids[12, 36]. In regard to the (M + 31)- ions, as discussedin the preceding section, ionic addition reactions arerare in the pressure regime of the ICR spectrometer,because there is no convenient sink for the excessenergy. The attractive forces between the ion and theneutral dipole are great enough that a deep well existson the potential surface for the complex, so additioncomplexes are likely to be highly excited species withshort lifetimes. MNDO calculations [31] indicate thatthe methoxide affinity of B(OMe)3 is 39 kcal /rnol, atthe AMI level [32], the B-O bond in [(MeO)3B OMe]- is calculated to be 57 kcal /mol. We have nobasis for choosing one of these as being more accurate;however, as noted below, AMI calculations yieldacidities for R2BCH3 carbon acids 25 kcaljmol tooweak, and acidities for R2BOH oxyacids 16 kcalyrnoltoo strong. We note that MNDO calculations agree towithin 5 kcaljmol with the bond energies for [MeBH2- OMe]- and [MeOBH2 - OMe]- calculated at the4-31G level [19], and also agree with the experimentalbond energies [37] for [BH3 - CN]- and [Et38 - H]to within 8 kcaljmoL Also, MeO- ... HOMe transfersmethoxide to B(OMe)3[20], indicating that experimentally the B-O bond strength must be at least 29kcaljmol [38]. We have no reliable experimental upperlimit on this bond energy at present. The enolate ofmethyl acetate has a methoxide affinity of 47 kcaljmol[11] (i.e., methoxide bound to ketene). It does nottransfer methoxide to B(OMe)3' but rather yields anion corresponding to (MeO)2BCH=C(OMe)O-,equivalent to a Claisen condensation. This cannot betaken as an upper limit to the methoxide affinity ofB(OMe)3' therefore. It seems unlikely, however, that
ION CHEMISTRY OF BORATE AND BORONATE ESTERS 173
this would be as strong as the fluoride bond in BF4- at71 kcaljmol [11]. We thus say that the methoxideaffinity of B(OMe)3 is probably 50 ± 15 kcaljmoL
Based on these estimates of the energy of bondformation, the initially formed complex must thereforebe highly excited vibrationally, and requires somemethod of disposing of this energy to live long enoughto be detected by the ICR technique. The rate constantsfor appearance of the addition product of methoxideplus trimethyl borate reveal a slight negative pressuredependence: The apparent bimolecular rate constantdecreases from 3.0 x 10- 10 cm3/mol· sec at 9.0 x 10-8
torr, to 1.0 x 10-10 cm3/mol' sec at 5.3 x 10- 7 torr,with uncertainties in the rate constants of ca. 25%. Weare not sure whether the negative pressure dependence is an artifact, possibly due to more completeinternal thermalization of the reactant methoxide atthe higher pressures, but a similar negative pressuredependence was seen for alkoxide addition to alcohols[35]. Arguments for this process being in part radiativemay be made along the same lines as for the positiveion complexes:good infrared radiative structures existin the four B-O bonds in the adduct.
It is noteworthy that in a drift cell ICR experiment[19], similar adducts were noted but were attributed tocollisional stabilization at the relatively high pressuresused (I X 10-5 torr). Adduct formation was found inthat case to be considerably more facile for alkoxidesadding to borate esters than for the same alkoxidesadding to trialkylboranes [19]. This was attributed tothe reduced Lewis acidity of B(OR)3 relative to BR3. Inlight of the present study, the radiative ability of(RO)4B- with four B-O bonds versus that ofRO(R')3B- with only one may be a more importantfactor.
For B(OEt)3 reacting with methoxide, an (M + 45)ion is observed, which double resonance studies indicate is derived from the (M + 31)- addition product,and not from an ion of m/z 45-. This indicates that anexchange of alkoxide ligands can occur on the collisionof the addition product (EtO)3B-OMe with anotherneutral borate molecule. Similar ligand interchangewas observed in the drift cell study [19] for borateesters, but not for trialkylboranes. If methanol is present in the vacuum system, then methoxy and ethoxyligands are both available for exchanges into the addition products on collision with methanol or with theborate. Figure 2 illustrates the unquenched spectrum[30] obtained in this experiment; it shows a series ofpeaks with a spacing of 14 mass-to-charge ratios in therange of mrz 135- to m/z 191-. When no methanol ispresent, only (EtO)3B(OMe)- at mjz 177- and(EtO)4B- at m r z 191- are observed. As the methanolpressure is increased, the relative intensity of the highermass peaks decreases and (EtO)2B(OMe); at mjz163, EtOB(OMe); at mrz 149-, and (MeO)4B- atm/z 135- grow in. The spectrum shown representsapproximately equal pressure of methanol and triethylborate. Scheme II illustrates the sequential exchange
174 KIPUNGER ET AL. JAm Soc Mass Speclrom 1994, 5, 169-176
Figure 2. Unquenched mass spectrum of reaction of (EtO)3Bwith methoxide. Scale is in daltons.
in the spectrum at long reaction times, indicating thatit is a very minor product, formed slowly. This agreeswith the results from drift cell ICR investigations [19]of this system, where small amounts (1-3%) of(RO)2BO- product are produced, from either displacement (R ~ Me) or elimination (R = Et) reactions, witha variety of alkoxide reactants. In our trapped cellexperiments, when 0-· is used as the primary ion withB(OMe)3' (MeO)2BO- becomes the most prominentsignal in the spectrum, ca. 62% of the ionic products. This ion reacts with (MeO)3B to give anaddition product, (MeO)2BOB(OMe)3' The acidityof the conjugate acid of (MeO)2BO- is bracketedbetween tBuCH=NOH (~Gacid = 355.8) and pyrrole(LlGacid = 350.9) [11], yielding ~Gacid = 353.4 ± 4.5kcaljmol for (MeO)2BOH. By using statistical mechanics to calculate ~Sadd [11], a ~Hacid of 360.3 ± 4.8kcaljmol is found, compared to a value from theMNDO method (31) of 361.1 kcaljmol.
The acidities of XOH compounds of this type havebeen shown by Taft et al. [15) to be correlated with amaster equation:
f.Gacid = 23.4· IT. - 73.4· IT p - 72.8' ITR + 384.5 (5)
where IT., ITp, and ITR are polarizability, field/inductive, and resonance parameters, respectively. A IT p
value of - 0.07 for the - B(OMe)2 group is availablefrom 19F NMR shift data [13], and the polarizability of- B(OMe)2 is taken as equal to that of the corresponding alkyl group, for IT" = -0.72 [15). This latter isjustified because the bulk polarizability of (EtO)3B iswithin 2% of that for (EtO)3CH and Et'lCH [39). Byusing the ITR for - BMe2 of 0.27, derived at the beginning of this article, a LlGacid of 353.1 kcaljmol iscalculated from the equation, agreeing very well withthe experimental value. This may be fortuitous in thatthe - B(OMe)2 group would be expected to be not asgood in stabilizing negative charge by resonance as- BMe2 (smaller ITR), due to electron donation fromthe oxygens to the boron reducing its accepting ability.Considering the uncertainties in the values for the fieldand polarizability parameters, we view this applicationof eq 5 as semiquantitative, but in agreement with theexperimental results.
Reaction of NH2 with B(OMe)3 results in production of (MeO)2BNH-, presumably by an addition/elimination/deprotonation sequence of reactions.MNDO calculations predict ~Hacid = 374.5 kcaljmolfor the neutral amine; no attempt was made to bracketthe acidity directly.
Boronate esters. The initial impetus for this work wasto produce and examine the properties of carbanionsgenerated Ci. to boron atoms. MNDO calculations yieldan anion proton affinity (f.Hacid) of 380.0 kcaljmol for4. Because MNOO calculations yield a value for theanion proton affinity of trimethylborane of 360.7kcaljmol, within 1 kcaljmol of the experimentallydetermined gas-phase number [11], we are reasonablyconfident of their general accuracy in this case. The
'9'
177
IBO
163
(MeO) 38(08) m/z 121or M~OH . -OB(OMc) 2
(EtC) 2B(OMe) (OH) m/z 149or aeoa- --OB(OEt) 2
(EtO) )B(OH) m/z; 163or EtOH· -08(OEt) 2
160
149
0/'
140120
Scheme II
11-(EtO) ROMe
+M$OH +B(OEtf 3~EtOH
11-(EtO) ,SOMe
-xeoa +B(OEt) 3-EtOH
mjz 149 (EtO)B(OMe)3 - CK 2=CH2 +
m/z ~Js.
equilibria involved in the generation of such a spectrum. Exchanges of the ligands may occur on collisionwith either the neutral borate or methanol.
One problem dealing with structural assignmentsinvolves the ion at mrz 149-. Nominally EtOB-(OMe)3by Scheme II, this might also arise as shown in theelimination/addition reaction 4:
However, this ion should also undergo the ligandexchange reaction described above with the methanolpresent, culminating in HOB-(OMe)3 at 121-. No ionof that mass is observed in the spectrum, implying thatthis elimination/clustering mechanism is not operating. The presence of a small peak at m/z 117- inFigure 2 indicates that some elimination occurs, butwithout the clustering shown. This peak is not present
J Am Soc Mass Spectrom 1994,5,169-176
acidity calculated for 4 is comparable to that ofmethanol C~Hacid = 380.6 ± 2.3 kcaljmoO [11]. Wethus expect that methoxide thermochemically shouldbe able to deprotonate boronate esters in the lCRspectrometer.
Boronate ester 3, used in solution as a source ofa-boron carbanions [2-4], in the gas phase displays analternate reactivity on treatment with strong anionicbases. An (M - H)- ion is observed when 3 is reactedwith methoxide. However, this (M - H)- ion does notexhibit the expected two deuterium exchange productswhen CH30D is admitted to the system [411, an observation which is consistent with the elimination product shown in reaction 6. The threshold for this elimination process is at approximately JiGacid = 370 ± 5kcaljmol (bracketed as occurring between methoxideand neopentoxide: JiGacid methanol = 374.0 ± 2kcalymol, and JiGacid neopentyl alcohol = 366.2 ± 2kcalymol [11]). A similar elimination product has beenproposed elsewhere [19].
MeO-+ 3
0--> MeB~ (6)
OC(Me)2C(Me)=CH2 + MeOH
In contrast to 3, boronate ester 4 does not havereactive sites that could lead to an elimination reactionlike 6 [41]. On reaction with methoxide, 4 yields anionsat mf z 127 and mJz 159. The ion at m/z 159 corresponds to the methoxide addition product of 4. The ionat m/z 127 is not the desired (M - H)- ion, however,but rather iodide. The iodide ion is ubiquitous innegative ion mass spectrometry whenever methyl iodide is used in the synthesis of a compound, as wasdone for 4 and 5. Evidence for this assignment includes a lack of deuterium exchange into 127- whenMeOD is present, and the complete lack of reactivity of127- with any other reagent tried. Boronate ester 5,with the nominal (M - H)- ion shifted to 141 u, doesnot yield an ion at that mass on treatment withmethoxide, but instead only an (M + 31)- ion, plusiodide.
If hydroxide is used as the anionic base, the expected (M - H)- elimination product on reaction withthe boronate 3 is observed. There is also a low intensity (M + 1)~. ion, attributed to hydroxide addition toboron, followed by methane loss, to give a cyclic(RO)2BO- ion. Addition of hydroxide to 3 is alsoobserved, yielding an (M + 17)- ion, presumably viaan addition/radiative stabilization mechanism, as forthe borate esters. When deuteroxide, from D20, isused as the base, no deuterium exchange or mass shiftis seen for any of these ions, save that the (M + 17)becomes an (M + 18)- ion. Similar reactivity is observed for 4 and 5 when deuteroxide generated fromD20 is used: replacement of the R group on boronwith -0-, but no evidence of a deuterium exchange,as in reaction 7.
ION CHEMISTRY OF BORATE AND BORONATE ESTERS 175
[HO,- /OR']
HO- + RB(OR'h -> /lk....R OR'
-> [R- ... HOB(OR'h]
-> RH + -OB(OR'h (7)
Similar addition/elimination mechanisms have beenproposed for nucleophilic addition to silicon esters[42]. Based on MNDO heats of formation, the loss ofmethane by this process is calculated to be essentiallythermoneutral from the addition complex, and theaddition complex of 4 plus hydroxide is ca. 50 kcaljmolexothermic from reactants, due to the covalent bondformation. There is thus sufficient energy present inthe intermediate to drive this process. The equivalentprocess is not observed when methoxide is the base;the elimination of two methyl groups as ethane fromthe complex of borate plus methoxide is calculated torequire about 7 kcaljmol more energy. This is stillthermochemically allowed, but may involve a moreelaborate mechanism. The mJz 129- ion undergoesfurther reaction, adding to the neutral boronate ester ina (presumably radiative) method, to yield an (M +129)- ion. If deuteroxide is used as the primary anion,no deuterium incorporation is observed for m/z 129or the (M + 129)- ions, consistent with the proposedmechanism.
The reaction of NDi, generated by electron ionization of ND3, with 3, 4, and 5 does not produce any(M - H)- ions, with the exception of the eliminationproduct from 3. Only (1) addition products (M + D)and (M + ND 2)-, (2) addition/fragmentation products such as (RO)2BND-, and (3) hydrogen exchangespecies from NH3 or H 20 contaminants into the aboveions are observed in the unquenched spectra [30].
Conclusions
The borate and boronate esters investigated here allundergo Lewis acid/base reactions involving nucleophilic attack at boron, for both negative and positiveions. All attempts to deprotonate at a C-H a to theboron were unsuccessful; only certain reactions involving Lewis attack at the boron first, followed by aproton transfer, resulted in such o-boronate anions.This gives (MeO)2BO- and (MeO)2BNH-. Such reactivity is in contrast to the trialkylboranes, where simple proton transfer reactions appear to be facile [19].
The Lewis basicity of the borate and boronate estersdominates their ion-molecule reactivity. Lewis basicity is likewise important for the chemistry of gaseoussilicon and titanium esters [20]. For the titanium esters,however, only ligand interchange was observed, andno pentacoordinate titanate anions. For silicon, ligandinterchange and pentacoordinate siliconate ions compete fairly evenly. For the borate and boronate esters inthis study, the anionic products always involve theboron species; while ligand interchange occurs, theligands that are lost are neutral alcohols. There thus
176 KIPLINGER ET AL. J Am Soc Mass Speclrom 1994, 5, 169-176
appears to be a range of Lewis basicity for theseelements.
AcknowledgmentsWe thank the National Science Foundation for support of thiswork, and Prof. D. S. Matteson for a sample of 3.
References1. (a) Progress in Boron Chemistry, Vol. 1-3; Steinberg, H.;
Melosky, A. L.; Brotherton, R J., Eds.; Pergamon Press: Oxford, 1964; (b) Matteson, D. S. ibid; Vol. 3, pp 118-134; (c)Brown, H. C, Boranes in Organic Chemistry; Cornell Univ.Press: Ithaca, NY 1972; (d) Borane Real?ents; Peltier, A.; Smith,K; Brown, H. C, Eds.; Academic Press: New York, 1988.
2. Matteson, O. S.; Kramer, E. ]. Am. Chem. Soc. 1968, 90,7261-7266.
3. Matteson, D. S.; Moody, R]. Organometallics 1982, 1,20-28.4. Matteson, D. S.; Sadhu, K M.; Leinhard, G. E. f. Am. Chem.
Soc. 1981, 103, 5241-5242.5. Matteson, D. S.; Moody, R. J.; Jesthi, P. K. f. Am. Chem. Soc.
1975, 97,5608-5609.6. Pelter, A.; Bugden, G.; Rosser, R Teirohed. Lett. 1985, 26,
5097-5100.7. Matteson, D. S. Synthesis 1975, 147-158; Rathke, M. W.; Kow,
R f. Am. Chem. Soc. 1972, 94, 6854-6856; Kow, R; Rathke,M. W. j. Am. Chem. Soc. 1973, 95,2715-2716.
8. Matteson, D. 5.; Wilson, J. Orl?anometallics 1985, 4 1690-1694.9. [ohlman, C. L.; Ijames, C. F.; Wilkins, C. L.; Morton, T. H. f.
Org. Chem. 1983, 48, 2628-2629.10. Murphy, M. K.; Beauchamp, J. L. }. Am. Chem. Soc. 1976, 98,
1433-1440.11. NTST Negatil}e Ton Energetics Database, SRD Database 19B;
Bartmess, ]. E., Ed. Version 2.1, April 1990. Distributedthrough Office of Standard Reference Data, National Institutefor Standards and Technology, Gaithersburg, Maryland; Lias,S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin,R. D.; Mallard, W. G. J. Phys. Chern. Ref Data, 1988, J7, Suppl.1.
12. Sullivan, S. A. Ph.D. Thesis, Cal'Tech, 1977. Reported aciditiesadjusted to new absolute acidity scale in ref 11.
13. Taft, R. W.; Price, E.; Fox, I. R; Lewis, I. c, Andersen, K K.;Davis, G. T. J. Am. Chem. Soc. 1963, 85,709-724; Taft, R. W.Private communication.
14. Pola, ].; [akoubkova, M.; Chvalovsky, V. Coil. Czech. Chern.Commu n. 1979, 44, 3688-3694.
15. Taft, R W.; Koppel, I. A.; Topsom, R. D.; Anvia, F. J. Am.Chent. Soc. 1990, 112, 2047-2052; Taft, R. W.; Abboud,J. L. M.; Anvia, F.; Berthelot, M.; Fujio, M.; Gal, J.-F.;Headley, A. D.; Henderson, W. G.; J. Am. Chem. Soc. 1988,110,1797-1800.
16. Bartmess, J. E.; Caldwell, G.; Rozeboom, M. D. f. Am. Chern.Soc. 1983, 105, 340-344.
17. Hansch, C.; Leo, A.; Taft, R. W. Chern. Rev.1991, 91, 165-195.J8. Hettich, R. L.; Cole, T; Freiser, B. S. Int. f. Mass Spectrom. Ion
Proc. 1987, 81, 203-215.19. Hayes, R.N.; Sheldon, 1- c, Bowie, J. H Organometallics 1986,
5,162-167.20. van der Wei, H.; Nibbering, N. M. M.; Sheldon, J. C; Hayes,
R. N.; Bowie, J. H f. Am. Chem. Soc. 1987, 109, 5823-5828.21. Blair, L. K.; Isolani, P. c.; Rivcros, J. M. f. Am. Chem. Soc.
1973, 95, 1057-1060; Haartz, 1- c, McDaniel, D. H. J. Am.Chern. Soc. 1973, 95, 8562-8565; Krivtsov, N. V.; Titova, K. V.;Rosolovskii, V. Ya. Russ. J. Inorg. Chem. 1977, 22, 374-377;Murphy, M. K.; Beauchamp, J. L. Tnorg. Chern. 1977, 16,
2437-2443; Larson, 1- W.; McMahon, T B. J. Am. Chern. Soc.1985, 107, 766~773; Larson, J. W.; Szulejko, J. E.; McMahon, T.B. J. Am. Chern. Soc. 1988, 110, 7604-7609; Veljkovic, M.;Neskovic, 0.; Zrnbov, K. F.; Borshchevsky, A. Y.; Vaisberg, V.E.; Sidorov, L. N. Rapid. Commun. MassSpectrom. 1991, 5,37.
22. Dolthun, J. J.; Wiebers, J. L. Org. Mass Spectrom. 1970, 3,669-681; Swning, H.; Yaozu, c.; Lonfei, J.; Shuma, X. Org.Mass Spectrom. 1985, 20, 719-723; Rothnell, A. P.; Wood,K. V.; Srebnik, M.; Cole, T E. Org. Mass Spectrom. 1986, 21,165-167.
23. Mciver, R T., Jr. Rev. Sci. lnstrum. 1970, 41, 555-558; Mciver,R. T., Jr. ibid. 1978, 49, 111-118; Mciver, R. T., [r.; Hunter,R. L.; Ledford, E. B., Jr.; Locke, M. 1-;Francl, T. J. Int. f. MassSpectrom. Ion Phys. 1981, 39, 65-84.
24. Hunter, R. L.; Mciver, R. T., Jr. Chern. Phys. Lett. 1977, 49,577-585; Hunter, R L.; Mciver, R T., Jr. Am. Lab. 1977,9,13;Hunter, R. L.; Mciver, R. T., Jr. in Lecture Notes in Chemistry,Vol. 31; Hartmann, H.; Wanczek, K.-P., Eds.; Springer-Verlag:Berlin, 1982; pp 464-483.
25. Lee, 1- P.; Comisarow, M. B. Appl. Spectros. 1987, 41, 93-98;ibid. 1989, 43, 599-604.
26. Grimm, D. T.; Bartmess, J. E. Proceedings of Ihe 39th AnnualConference of the American Society for Mass Spectrometry,Nashville, TN, May 19-24, 1991; p 1521.
27. Caldwell, G.; Bartmess, J. E. Org. Mass 5pectrom. 1982, 17,456-457.
28. Lehman, T.; Bursey, M. M. IO>l Cyclotron Resonance Spectrometry; Wiley: New York, 1976.
29. Wada, Y.; Kiser, R. W. J. PIIYs. Chem. 1964, 68, 1588-1590;Hammerrum, 5.; Djerassi, C. Org. Mass Spectrorn. 1974, 8,217-227; Fallon, P. J.; Kelly, P.; Lockhart, J. C. Int. J. MassSpectrom. IonPltys. 1968, 1, 133-139.
30. Hunter, R L.; Mciver, R. T, Jr. Anal. Chern. 1979, 51, 699;Bartmess, J. E.; Caldwell, G. Int. J. Mass Spectrom. Ion Phys.1981, 41, 125-134.
31. Dewar, M. J. S.; Thiel, W. f. Am. Chem. Soc. 1977, 99,4899-4907; Dewar, M. J. S.; Rzepa, H. S. f. Am. Chem. Soc.1978, JUO, 784-790; Program No. QCMPOO5, Quantum Chemistry Program Exchange, Indiana University, Bloomington,Indiana; Dewar, M. J. 5.; McKee, M. L. J. Am. Chern. Soc. 1977,99,5231-5241.
32. Dewar, M. J. 5.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. P. J. J.Am. Chern. Soc. 1985, 107,3902-3909; Dewar, M.]. 5.; Dieter,K. M. J. Am. Chern. Soc. 1986, 108,8075-8086.
34. Bartmess, J. E. J. Am. Chern. Soc. 1980, J02, 2483-2484.35. Caldwell, G.; Bartmess, j. E. J. Phys. Chem. 1981, 85,
3571-3577; Herd, C. R; Babcock, L. M. J. Phys. Chern. 1987,91,2372-2376; Kofel, P.; McMahon, T. B. f. Phys. Chern. 1988,92, 6174-6176; Herd, C. R; Babcock, L. M. J. Pltys. Chem.1989, 93,245-251.
36. Ellis, H. B.,Jr.; Ellison, G. B. J. Chern. Phys. 1983, 78,6541-6558.37. Workman, D. 8.; Squires, R. R. lnorg. Cltem. 1988, 27,
1846-1848.38. Meet-ncr, M.; Sieck, L. W. f. Phys. Chelll. 1986, 90, 6687-6690.39. Handbook of Chemistry and Physics, 58th ed.; Weast, R. C; Ed.;
CRC Press: West Palm Beach, FL, 1977.40. Stewart,]. H.; Shapiro, R H; DePuy, C. H; Bierbaum, V. M.
f. Am. Cltem. Soc. 1977, 99, 7650-7653; DePuy, C. H.; Bierbaum, V. M.; King, G. K; Shapiro, R H f. Alii. Chern. Soc.1978, 100, 2921-2922; Hunt, D. F.; Sethi, S. K. f. Am. Chem.Soc. 1980, 102, 6953-6963; Lloyd, J. R; Agosta, W. c.. Field,F. H. f. Org. Chem. 1979, 45, 3483-3492.
41. Bartmess, J. E.; Hays, R L.; Khatri, H. N.; Misra, R. N.;Wilson, S. R. J. Am. Chem. Soc. 1981, 103, 4746-4751.
42. Angelini, G.; Johnson, C E.; Brauman, ]. I. Tnt. f. MassSpectrom. Ton Proc. 1991, 109, 1-14.