David V. Dearden

Analysis of enantiomeric excess using mass spectrometry: fast atom bombardment/sector and electrospray ionization/Fourier transform mass spectrometric approaches

Abstract The utility of fast atom bombardment (FAB) ionization on a sector mass spectrometer, and of electrospray ionization (ESI) on a Fourier transform ion cyclotron resonance mass spectrometer, for enantiomeric excess measurements was explored. Both methods involved the same host–guest system: (R,R)- or (S,S)-dimethyldiketopyridino-18-crown-6 (host) and α-(1-naphthyl)ethylammonium (guest). Both use an achiral amine (benzylamine for the FAB experiments, cyclohexylamine for the ESI experiments) as an internal reference compound and involve competitive complexation of the achiral and chiral amines with the chiral host. The FAB experiments are shown to give stable, reproducible results, but exhibit a smaller degree of enantiodiscrimination than the ESI experiments. The ESI experiments, which involve measurement of apparent guest exchange equilibrium constants, show a linear relationship between apparent equilibrium constant and enantiomeric excess. The apparent equilibrium constant is shown to be a composition-weighted average of the equilibrium constants for the two pure enantiomers. Enantiomeric impurities as small as about 2% can currently be detected.

Supramolecular Modification of Ion Chemistry : Modulation of Peptide Charge State and Dissociation Behavior through Complexation with Cucurbit[n]uril (n = 5, 6) or α-Cyclodextrin

Electrospray Fourier transform ion cyclotron resonance mass spectrometry, ion mobility spectrometry, and computational methods were utilized to characterize the complexes between lysine or pentalysine with three prototypical host molecules: alpha-cyclodextrin (alpha-CD), cucurbit[5]uril (CB[5]), and cucurbit[6]uril (CB[6]). Ion mobility measurements show lysine forms externally bound, singly charged complexes with either alpha-CD or CB[5], but a doubly charged complex with the lysine side chain threaded through the host cavity of CB[6]. These structural differences result in distinct dissociation behaviors in collision-induced dissociation (CID) experiments: the alpha-CD complex dissociates via the simple loss of intact lysine, whereas the CB[5] complex dissociates to yield [CB[5] + H(3)O](+), and the CB[6] complex loses neutral NH(3) and CO, the product ion remaining a doubly charged complex. These results are consistent with B3LYP/6-31G* binding energies (kJ mol(-1)) of D(Lys + H(+)-alpha-CD) = 281, D(Lys + H(+)-CB[5]) = 327, and D(Lys + 2H(2+)-CB[6]) = 600. B3LYP/6-31G* geometry optimizations show complexation with alpha-CD stabilizes the salt bridge form of protonated lysine, whereas complexation with CB[6] stabilizes doubly protonated lysine. Complexation of the larger polypeptide pentalysine with alpha-CD forms a nonspecific adduct: no modification of the pentalysine charge state distribution is observed, and dissociation occurs via the simple loss of alpha-CD. Complexation of pentalysine with the cucurbiturils is more specific: the observed charge state distribution shifts higher on complexation, and fragmentation patterns are significantly altered relative to uncomplexed pentalysine: C-terminal fragment ions appear that are consistent with charge stabilization by the cucurbiturils, and the cucurbiturils are retained on the fragment ions. Molecular mechanics calculations suggest CB[5] binds to two protonated sites on pentalysine without threading onto the peptide and that CB[6] binds two adjacent protonated sites via threading onto the peptide.

Macrocyclic chemistry without solvents: Gas phase reaction rates

Both size selection and macrocyclic effects are observed for crown ethers reacting with alkali metal cations in the absence of solvent species. The efficiencies of the radiatively-stabilized 1 :1 1igand:metal complexation reactions scale directly with cation charge density, suggesting that ion-induced rearrangement of the ligand into a favorable binding conformation is more efficient for smaller cations. The 1 :1 reaction efficiencies are also greater for the macrocycles than for the corresponding acyclic ligands, reflecting preorganization of the binding cavity in the case of the crowns. The efficiencies of the subsequent reactions of the 1 :1 complexes to form 2:l 1igand:metal species are dominated by size effects: when the cations are small enough to enter the macrocycle cavity, reaction is very slow, but the reactions occur readily when the cations cannot enter the crown cavity. Acyclic ligands undergo the 2:l complexation reaction much more slowly and less selectively than do the macrocycles, again evidencing macrocyclic effects in the absence of solvation.

One Ring to Bind Them All: Shape-Selective Complexation of Phenylenediamine Isomers with Cucurbit(6)uril in the Gas Phase

We examined complexes between cucurbit[6]uril and each of ortho-, meta-, and para-phenylenediamine using computational methods, Fourier transform ion cyclotron resonance mass spectrometry, and ion mobility spectrometry. These fundamental gas phase studies show that the lowest energy binding sites for ortho- and meta-phenylenediamine are on the exterior of cucurbit[6]uril, whereas para-phenylenediamine preferentially binds in the interior, in a pseudorotaxane fashion. This conclusion is based on reactivity of each of the complexes with tert-butylamine, where the ortho- and meta-phenylenediamine complexes exchange with tert-butylamine, whereas the para-phenylenediamine complex undergoes two slow additions without displacement. Further, under sustained off-resonance irradiation conditions, the ortho- and meta-phenylenediamine complexes fragment easily via losses of neutral phenylenediamine, whereas the para-phenylenediamine complex fragments at higher energies primarily via cleavage of covalent bonds in the cucurbituril. Finally, ion mobility studies show ion populations for the ortho- and meta-phenylenediamine complexes that primarily have collision cross sections consistent with external complexation, whereas the para-phenylenediamine complex has a collision cross section that is smaller, the same as that of protonated cucurbit[6]uril within experimental error. In agreement with experiment, computational studies indicate that at the HF/6-31G* and B3LYP/6-31G*//HF/6-31G* levels of theory external complexation is favored for ortho- and meta-phenylenediamine, whereas internal complexation is lower in energy for para-phenylenediamine. In contrast, MP2/6-31G*//HF-6-31G* calculations predict internal complexation for all three isomers.

Gas phase studies of ammonium–cyclodextrin compounds using Fourier transform ion cyclotron resonance

Abstract Cyclodextrins (CDs) are cyclic oligosaccharides composed of 6, 7, or 8 glucose molecules (α-, β-, or γ-cyclodextrin, respectively) which are used widely in industry due to their ability to form inclusion complexes with a variety of molecules in aqueous solution. Much speculation has been made as to whether inclusion complexes form as a result of hydrophobic interactions between guest molecules and the inner hydrophobic cavity of the CDs in water. Fourier transform ion cyclotron resonance (FTICR) mass spectrometry was used to study adducts of cyclodextrins with various amines in the gas phase. Protonated cyclodextrins were generated using electrospray ionization, and were allowed to react with neutral amines. Adducts of each amine studied were observed to form with all three cyclodextrins. Equilibrium constants were measured for the exchange of neutral amines on protonated CD molecules. Size and shape dependent trends, especially with bulkier amines, suggest inclusion complex formation. Molecular modeling studies also support the formation of inclusion complexes rather than nonspecific adducts, and suggest that solvation of the charged guest by the CD host provides a large driving force for the formation of inclusion complexes, which are then stabilized by van der Waals interactions between the host and the guest. A second series of experiments was performed using gas phase hydrogen/deuterium exchange of protonated cyclodextrins and cyclodextrin–amine complexes with D 2 O. The protonated cyclodextrins have a rapid rate of exchange that slows by more than a factor of 10 when an amino guest is added. The amino groups of the guests are expected to have significantly higher gas phase basicities than the hydroxyl sites on the cyclodextrins or the deuterating agent, accounting for the observed decrease in exchange rates for cyclodextrin–amine complexes. Observed differences in the α- versus β- versus γ-cyclodextrin exchange rates suggest an exchange mechanism dependent upon the size of the cyclodextrin ring and its gas phase conformation.

Metal Separations Using Emulsion Liquid Membranes

Abstract Emulsion membrane systems consisting of an aqueous metal salt source phase, a toluene membrane containing the macrocyclic ligand dicyclohexano-18-crown-6 (DC18C6) (0.02 M) and the surfactant sorbitan monooleate (3% v/v), and an aqueous 0.05 M Li4P2O7 receiving phase were studied with respect to the disappearance of metal from the source phase as a function of time. The salts Pb(NO3)2, Sr(NO3)2, TINO3, and LiNO3 were studied both singly and in mixtures of Pb(NO3)2 with each of the other salts. In all mixtures studied, Pb2+ was transported first, followed by the second cation (except Li+ which was not transported). An excess of a second salt with a common anion enhanced the transport of Pb2+. Modeling of these systems was discussed. Source phases containing basic (pH 11) K[Al(OH)4] solutions were studied using the same membrane and a 0.15 M H3PO4 receiving phase. K+ and Al(III) (as aluminate anion) were both found to transport in this system, but no transport of Al(III) and little transport of K+ w...

Reactions of multidentate ligands with ligated alkali cation complexes: self-exchange and “sandwich” complex formation kinetics of gas phase crown ether–alkali cation complexes

Abstract Natural abundance isotopic labeling has been employed to study the reactions of labeled LM + complexes with L (L = triglyme, TG; 12-crown-4, C4; 15-crown-5, C5; 18-crown-6, C6; or 21-crown-7, C7; M = Li, Na, K, Rb, or Cs) in the gas phase using Fourier transform ion cyclotron resonance mass spectrometry. Reaction efficiencies for both ligand exchange and the formation of 2:1 ligand:metal “sandwich” complexes were determined. For a given ligand, self-exchange rates generally decrease with increasing metal size, while the sandwich complex formation rates show strong dependence on the relative sizes of the metal ions and ligand cavities. Acyclic TG complexes undergo self-exchange more rapidly than the analogous cyclic C4 complexes, whereas the sandwich complex formation rates are faster for the C4 complexes. Sandwich formation rates show a weak positive pressure dependence, as increased pressure leads to increased collisional stabilization of the complexes. Extrapolation of the rates to the zero pressure limits still yields significant rates, reflecting radiative stabilization. The self-exchange reactions have weak, negative pressure dependences, suggesting they are in direct competition with sandwich complex formation. Analysis of the sandwich complex formation radiative association kinetics yields estimates of binding enthalpies for attachment of the second ligand. Trends in the binding enthalpies, like the kinetics, show strong dependence on the relative sizes of the metal ions and ligand cavities. For a given metal, binding of a second TG is weaker than binding of a second C4. Binding enthalpies for the second ligand are in every case substantially less than calculated binding enthalpies for the first ligand to attach to a given metal.

Fundamental factors controlling the exchange of multidentate ligands: displacement of 12-crown-4 and triglyme from complexes with divalent alkaline earth cations

Abstract In an effort to shed light on the factors that influence the recognition of alkaline earth cations in natural systems, we have studied intrinsic recognition of these cations by well-ordered synthetic ionophores such as crown ethers (12-crown-4 [C4] and 18-crown-6 [C6]) as well as the acyclic analog of C4, triglyme (TG), in the gas phase. We have employed electrospray ionization (ESI) to generate gas phase crown and glyme alkaline earth complexes, and have used Fourier transform ion cyclotron resonance mass spectrometry to measure rate constants for displacement of the original ligands by C6. ESI of mixtures of C4 and TG with alkaline earths primarily produces sandwich complexes of the doubly charged cations, (C4)2M2+, (C4)(TG)M2+, and (TG)2M2+. We find that the ligand exchange reactions are generally very efficient, with rates approaching or exceeding the Langevin collision rate in most cases. Trends in rates as metal size varies can be understood in terms of the degree of encapsulation of the metal by the ligands when the coordination shell is partially filled (smaller metals are more thoroughly encapsulated and tend to react more slowly) and in terms of the polarizing power of the metal cation when the metals are either “bare” or completely coordinated (smaller metals have greater charge density and tend to react more rapidly). Efficiencies for most of the reactions studied fall off in the order Mg2+ > Ca2+ > Sr2+ > Ba2+, consistent with decreasing charge density as the cation radius increases. Interestingly, TG is displaced more efficiently than C4 by C6, despite the fact that the total binding energy of the glyme is greater than that of the crown. This is consistent with a mechanism wherein the rate-limiting step involves breaking O–M2+ electrostatic bonds, and where the bonds to the oxygens of TG can be broken one at a time, whereas the more rigid ring structure of C4 requires concerted breaking of multiple bonds. Molecular dynamics simulations of this process for complexes where M2+ = Ca2+ give support to this interpretation: in all observed dissociation events, TG oxygens were removed from the metal one at a time, whereas displacement of C4 oxygens occurred in pairs.

Complexes of p-tert-butylcalix[4]arene with mono- anddipositive cations in the gas phase

Abstract The host-guest chemistry of p-tert-butyl-calix[4]arene in the gas phase was studied using Fourier transform ion cyclotron resonance mass spectrometry. Gas phase complexes with laser-desorbed alkali metal cations (Na+, K+, Rb+ and Cs+) were generated. The calixarene was observed to complex with up to two metal cations at the same time with the accompanying loss of a proton, to yield species with a net charge of +1. Complexes with Sr2+ and Ba2+, with accompanying loss of a proton to maintain a net charge of +1, were also observed. Relative rate constants and reaction efficiencies for the formation of the 1:1 calixarene/alkali metal complexes vary inversely with cation size, probably reflecting differences in the polarizing capabilities of the cations. Reaction of the alkali metal and alkaline earth complexes with 15-crown-5 results in adduction of the crown in cases where the metals are large, suggesting the larger cations bind in exposed positions on the calixarene. Finally, complexes of the calixarene with benzylammonium cation were observed. Equilibrium measurements indicate ΔG°350 for transfer of benzylammonium from the calixarene to 15-crown-5 is −9.1 ± 0.4 kJ mol−1.

Collision cross sectional areas from analysis of Fourier transform ion cyclotron resonance line width: a new method for characterizing molecular structure.

We demonstrate a technique for determining molecular collision cross sections via measuring the variation of Fourier transform ion cyclotron resonance (FTICR) line width with background damping gas pressure, under conditions where the length of the FTICR transient is pressure limited. Key features of our method include monoisotopic isolation of ions, the pulsed introduction of damping gas to a constant pressure using a pulsed leak valve, short excitation events to minimize collisions during the excitation, and proper choice of damping gas (Xe is superior to He). The measurements are reproducible within a few percent, which is sufficient for distinguishing between many structural possibilities and is comparable to the uncertainty in cross sections calculated from computed molecular structures. These techniques complement drift ion mobility measurements obtained on dedicated instruments. They do not require a specialized instrument, but should be easily performed on any FTICR mass spectrometer equipped with a pulsed leak valve.