Combined Trapped Ion Mobility and Infrared Ion Spectroscopy Study of Protonation Sites in Aromatic Amines

Abstract

The protonation site of aromatic amines in the gas phase has been under substantial debate, as it involves a subtle competition between the higher electronegativity of the amine nitrogen and the better charge delocalization ability of the fused aromatic rings. Previous studies have unambiguously shown, especially by ion mobility measurements, that higher-energy tautomers are easily observed depending on the experimental conditions in the ion source, including voltage settings and the type of solvent used in spray sources. Here, we use a combination of ion mobility and ion spectroscopy and focus on the tautomeric structure after ion mobility separation, in particular for protonated 1-aminonaphthalene and 1-aminoanthracene. We employ an atmospheric pressure chemical ionization (APCI) source, with a direct insertion probe to avoid any solvent influence, mounted on an FTICR mass spectrometer with a trapped ion mobility (TIMS) unit and optical access to the ions to perform infrared (IR) multiple-photon dissociation spectroscopy using the Free-Electron Laser for Infrared eXperiments (FELIX). TIMS analysis indeed reveals the presence of both N- and C-protonated species, but the IR spectra recorded in the ICR cell also suggest that mobilization and scrambling of the proton occur after TIMS separation. We computationally investigate the energetics of tautomerization and experimentally explore ion activation after TIMS separation.

Introduction

The preferential protonation site of aniline and its derivatives has been a subject of much debate and extensive study. − In solution, the amine nitrogen is the most basic site in aniline, so protonation leads to the formation of an ammonium species (−NH₃ ⁺). In the gas phase of a mass spectrometer, however, the actual protonation site of aniline is often ambiguous. Theoretical studies at the AM1 and density functional theory (DFT) levels of theory show that particularly the carbon atom located para to the amine moiety forms an energetically competitive protonation site, about 8 kJ mol^–1 lower in energy than protonation on the amine nitrogen. , However, earlier theoretical studies using different levels of theory had proposed alternative energetic orderings, fueling the debate concerning the gas-phase protonation site. ,,,− Experimental studies suggest that several instrumental factors, such as ionization conditions, can yield either N- or C-protonated tautomers. − The N-protonated isomer of aniline can be kinetically stabilized in the gas phase ,,, depending on the experimental conditions, especially on whether ions are produced through gas-phase chemical ionization or through electrospray ionization, and in the latter case, on the solvent proticity and polarity. ,,,

Intrinsically, protonation on the amine or on one of the aromatic carbon atoms involves the competition between the higher electronegativity of the nitrogen atom and the better charge delocalization ability of the aromatic moiety. As the size of the aromatic system increases upon going from aniline to larger polycyclic aromatic amines, the efficacy of charge delocalization gradually improves, increasing the stability of ring-protonated tautomers relative to that of the amine-protonated tautomer. Figure displays this trend, presenting computed proton affinities of the amine nitrogen and the C atom located para to the amine moiety on the same ring, which gives the lowest-energy C-protonated tautomer in all cases. Note the close proximity of proton affinities for aniline and the observation that DFT and MP2 results show similar trends but with significant discrepancy in absolute value for the proton affinity of the amine, reversing the relative stabilities for aniline. Overall, these results align with computed proton affinities reported for 1-aminopyrene and 1- and 2-aminonaphthalene.

1.

Proton affinity of the amino group (blue) and the para-carbon atom (black) in amino acenes as a function of the number of aromatic rings, computed at B3LYP/6-311++G** (left panel) and MP2/6-311++G** (right panel) levels of theory.

In addition to these intrinsic molecular properties, various studies have shown that experimental factors can influence the actual protonation process, potentially favoring the amino group as the kinetically stabilized protonation site. To establish the protonation site experimentally, UV ion spectroscopy and ion mobility spectrometry (IMS) were mainly employed. Jouvet and Noble reported that aminopyrene preferably protonates on one of the skeletal carbon atoms based on absorption in the visible region, whereas the amino protonated tautomer displays absorption in the UV. Kumar and Attygalle recently employed traveling wave IMS to explore how conditions in the ion source can be adjusted to obtain different tautomeric structures of protonated 1- and 2-aminonaphthalene. The influence of ionization conditions on tautomer formation was also investigated for aminobenzoic and aminophthalic acids, where isomer abundances were probed by drift-tube IMS.

The facile tautomerization occurring in the ion source as revealed in these studies − also raises questions on the stability of structures under conditions of IMS. IMS separates and analyzes gaseous ions based on their mobility through a buffer gas, and the question whether and to what extent an IMS measurement could alter the structure of an analyte ion has often been addressed. − Ion heating can occur during or after mobility separation, affecting collision cross section measurements and possibly inducing conformational changes or even isomerization or dissociation. − Thermometer ions have been used to quantify ion heating during IMS, − providing insights into internal energy and effective temperature of the ions as they traverse the IMS mass spectrometer.

Conventional IMS instruments can detect structural changes induced in the ionization source prior to IMS analysis. However, post-IMS ion structures can only be probed by tandem-MS based methods, − which inherently limits the ability to directly characterize subtle structural differences. More sophisticated analysis of structural modifications arising post-IMS, e.g., through spectroscopy or a second IMS stage, requires mostly dedicated, nonconventional instrumentation. −

In this study, we combine ion mobility with infrared ion spectroscopy (IRIS) to probe the tautomeric structure of 1-aminonaphthalene and 1-aminoanthracene after separation by ion mobility. In particular, we employ trapped ion mobility spectrometry (TIMS) and IR multiple-photon dissociation (IRMPD) spectroscopy to analyze the site of protonation and determine whether structural changes may occur post-TIMS.

Methods

Mass Spectrometry

Experiments were performed on a modified FTICR MS instrument (SolariX XR, Bruker Daltonics, Bremen, Germany) equipped with a 7-T superconducting magnet (Maxwell magnet, Bruker BioSpin, Wissembourg, France). The instrument is equipped with a trapped ion mobility spectrometry (TIMS) stage and is coupled to the beamline of the Free-Electron Laser for Infrared eXperiments (FELIX), as schematically shown in Figure . This setup was installed recently and allows us to record IR spectra of TIMS-separated and m/z-selected ions.

2.

Layout of the FTICR-MS (Bruker SolariX XR) showing the locations of the TIMS unit, the collision cell, and IR laser beam used for IRMPD spectroscopy in the ICR cell.

1-Aminonaphthalene (99%) and 1-aminoanthracene (99%) were purchased from Sigma-Aldrich (St. Louis, USA) and used without further purification. Ions were generated with an atmospheric pressure chemical ionization (APCI) source equipped with a direct insertion probe (DIP). The solid samples were deposited in small quantities on the glass capillary and inserted into the DIP assembly, where the molecules sublimed and reached the corona discharge. This method resembles that used by Kumar and Attygalle and invokes protonation in the gas phase, avoiding any possible influence of solvent media on the site of protonation. Ions then travel through a glass capillary into the source vacuum housing. The conventional entrance funnel is replaced by a TIMS tunnel with ion funnels at the entrance and exit (TIMS 0.5, Bruker, Billerica, MA). This device can be used to mobility select the incoming ions, or it can be used as a standard ion funnel transferring the ions to an RF octopole ion guide, followed by a quadrupole mass filter. A National Instruments LabView program controls the TIMS settings together with an adapted pulse program that synchronizes its operation to the FTICR MS data acquisition sequence. When the TIMS unit is employed as a mobility filter, ions of a selected mobility accumulate in the collision cell, where ions can also be subjected to collision-induced dissociation (CID). After passing through a hexapole guide, ions are transferred to a dynamically harmonized ICR cell (ParaCell, Bruker Daltonics, Bremen, Germany), where the m/z-analysis takes place and where IRMPD spectroscopy can be performed with the FELIX beam reaching the ions through a ZnSe window mounted on the back UHV flange of the vacuum system.

IRMPD Spectroscopy

In the current experiments, ions were irradiated with 10 macropulses of FELIX; each macropulse is ∼10 μs long and has an energy of 20–130 mJ depending on the emission wavelength. Macropulses are emitted at a repetition rate of 10 Hz and with a bandwidth of 0.5% of the center frequency. IR spectra are recorded for mass (and mobility) selected features by acquiring a series of mass spectra as the laser frequency is tuned across the spectroscopic range of interest. Whenever the laser frequency is resonant with a vibrational transition of the selected ion, the absorption of multiple (10–100) photons occurs and induces unimolecular dissociation, as detected in the mass spectrum. An IRMPD spectrum is reconstructed from the individual mass spectra by relating the precursor ion intensity (P) to the total fragment ion intensity as a function of laser wavelength

$$\mathrm{Yield}=\frac{\sum _i{f}_i}{P+\sum _i{f}_i}$$ 1

where f _i is the intensity of each fragment ion i produced by the IR irradiation. The IRMPD yield ranges between 0 and 1. The fragment fluence is a better proxy for the absorbance and is derived as

$$\mathrm{f}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{fluence}=-\ln (1-\mathrm{Y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d})$$ 2

A linear correction is applied to the fragment fluence to account for wavelength dependent variations in laser power and for the number of laser pulses employed. Wavelength calibration is achieved with a grating spectrometer.

Trapped Ion Mobility Spectrometry

Ion mobility spectrometry separates and characterizes ions based on their mobility through a buffer gas, a property closely related to their 3-dimensional shape. Trapped ion mobility spectrometry (TIMS) constitutes the reversal of the classical drift cell analyzer: ,− instead of ions moving through a stationary gas as in a drift cell, TIMS is based on holding the ions stationary in a moving column of gas exerting a forward drag force while being slowed by an opposing electric field. An electric field gradient induces spatial separation of the ions along the tunnel axis that is dependent on their mobility, K, as ions are immobilized at the point where the forward drag force and the backward electric force cancel. By scanning the voltage V _hold that sets up the field gradient (and thus the electric force), ions of different collisional cross sections (CCSs) elute at different times so that a mobilogram is recorded. The instantaneous hold voltage at which ions of specific mobility elute (V _elute) is empirically related to the mobility K as

$$K=A+\frac{B}{V_{\mathrm{e}\mathrm{l}\mathrm{u}\mathrm{t}\mathrm{e}}-V_{\mathrm{o}\mathrm{u}\mathrm{t}}}$$ 3

where A and B are empirical parameters that are fitted using a linear calibration with an Agilent low-concentration tune mix. In this study, TIMS provides tautomer separation and aids to identify ions based on their CCSs. TIMS separations performed in this work used N₂ as a buffer gas at approximately 300 K and 1–2 mbar. Parameters for TIMS operation were optimized for maximum ion signal (in view of the ion spectroscopy to be performed subsequently) and are reported in the Supporting Information; these generic settings can be regarded as harsh.

Computational Methods

Theoretical IR spectra were generated for the possible protonation tautomers of 1-aminonaphthalene and 1-aminoanthracene using density functional theory (DFT) at the B3LYP/6-31++G­(d,p) and B3LYP-D3­(BJ)/6-31+G­(d,p) levels of theory employing the Gaussian 16 software package as installed at the Snellius supercomputer at SURFsara, Amsterdam. Furthermore, single-point MP2/6-31++G­(d,p) calculations were performed to verify the relative energies of the different structures. Geometry optimizations were performed with standard convergence criteria, and vibrational spectra were computed within the harmonic oscillator approximation. To account for anharmonicity, harmonic frequencies were scaled by a factor of 0.975 within the fingerprint region (2000–600 cm^–1). A revised scaling factor of 0.946 was applied to frequencies calculated for the C–H and N–H stretches; this factor brings the strongest N–H stretch band of mobility peak B of protonated 1-AA exactly in overlap with the calculated band for the C3-protomer (see below).

Transition state (TS) calculations were performed to elucidate the energetics of proton migration along the aromatic ring. TS geometry optimization and frequency calculations were performed following quasi-Newton synchronous transit (QST3) calculations. The stationary points identified along the reaction pathways were verified to be either first-order transition states or local minima by confirming the presence of one or zero imaginary frequency, respectively. Visualization of the corresponding normal mode verified that the TS structures are indeed saddle points connecting the reactant and the product. In addition, intrinsic reaction coordinate (IRC) calculations were performed to verify that the TSs connect the intended reactant and the product. Gibbs energies were used to calculate the activation energy barriers.

Results and Discussion

Table presents computed gas-phase site-specific proton affinities for 1-aminonaphthalene (1-AN) and 1-aminoanthracene (1-AA) at different levels of theory, enabling one to evaluate the most likely tautomeric structures formed upon protonation. Tautomers protonated at one of the carbon atoms are denoted as 1-ANC _n H⁺ or 1-AAC _n H⁺, with n indicating the C atom where protonation occurs (see Figure for atom numbering). For both species, carbon atom C3 is located para to the amine and has the highest proton affinity of all C atoms; it is also higher than that of the amino N atom. Furthermore, the energy gap between N- and C-protonated tautomers increases with the size of the aromatic system, as can also be seen in Figure . The MP2/6-31++G­(d,p) values deviate from the B3LYP/6-31++G­(d,p) values: at the MP2-level, the N-protonated tautomer is more stable than what is predicted at the B3LYP level. To investigate if this discrepancy is due to B3LYP’s lack of dispersion, we employed the D3­(BJ) dispersion correction, which gives an energy ordering identical with that for uncorrected B3LYP. Additionally, geometry optimization at higher levels of theory (B3LYP-D3­(BJ)/aug-cc-pVTZ) were performed for the lowest-energy isomers of 1-AN, giving again an energy ordering identical with that of B3LYP/6-31++G­(d,p). Single-point calculations at the CCSDT level also predict the C_para-protonated tautomer to be the most stable isomer, although the amino-protonated tautomer is now lower than the C1-protonated tautomer. Although the origin of the deviation between B3LYP and MP2 remains unclear, we conclude that it is likely not due to a different treatment of dispersion.

1. Relative Gibbs Energies (at 298 K) in kJ mol^–1 of Protonated 1-AN and 1-AA, Calculated at Different Levels of Theory.

1-ANH+
Level Basis Set	B3LYP 6-31++G(d,p)	MP2 6-31++G(d,p)	B3LYP-D3(BJ) 6-31++G(d,p)
1-ANC3H+	0	0	0
1-ANC1H+	16.0	18.6	15.9
1-ANNH3 +	37.9	11.9	35.3
1-ANC5H+	62.2	84.5	61.6
1-ANC7H+	66.9	89.7	66.9
1-ANC8H+	83.2	92.1	83.8
1-ANC6H+	98.5	108.6	99.5
1-ANC2H+	99.9	116.3	100.5
1-ANC9H+	116.6	124.3	114.2
1-ANC10H+	137.1	139.1	137.7
1-ANC4H+	163.4	164.1	161.5

1-AAH+
Level Basis Set	B3LYP 6-31++G(d,p)	MP2 6-31++G(d,p)	B3LYP-D3(BJ) 6-31++G(d,p)
1-AAC3H+	0	0	0
1-AAC1H+	12.7	12.7	11.3
1-AAC5H+	18.1	29.3	20.4
1-AAC12H+	38.1	37.3	38.8
1-AANH3 +	51.9	23.9	48.9
1-AAC7H+	61.5	86.0	65.7
1-AAC9H+	72.6	98.5	71.2
1-AAC10H+	81.8	90.5	82.8
1-AAC8H+	93.8	103.8	94.3
1-AAC2H+	94.5	109.9	94.9
1-AAC14H+	122.2	123.2	122.7
1-AAC13H+	133.9	142.1	131.6
1-AAC11H+	149.7	168.1	147.7
1-AAC4H+	164.4	169.5	162.3
1-AAC6H+	168.7	168.4	166.5

3.

Molecular structures of (a) 1-aminonaphthalene and (b) 1-aminoanthracene with carbon atom labeling.

Trapped Ion Mobility Spectrometry

Given the ambiguity of the protonation site reported in the literature and predicted by theoretical calculations by us and others, TIMS mobilograms of the [M + H]⁺ ions of 1-AN and 1-AA produced with the APCI source were recorded; see Figure . The DIP-APCI ion source does not involve any solvent; the sample sublimes directly from the solid to the gas phase, where protonation occurs during the corona discharge. The mobilograms reveal the presence of two and three distinct isomeric species for 1-AN and 1-AA, respectively, presumably corresponding to different protonation isomers (referred to as protomers in the following). According to CCS calculations and the IMS profiles previously reported, , N-protomers have higher CCS values compared to C-protomers as the delocalized charge in C-protomers, indicated for instance by their smaller dipole moment, leads to weaker charge-induced dipole interactions with the N₂ buffer gas. , Therefore, for both 1-AN and 1-AA, the peak eluting at the lowest TIMS hold voltage can presumably be assigned as the N-protomer, followed by a somewhat broader peak attributed to two or more C-protomers, whose CCS values are close and therefore cannot be separated further on our TIMS. For 1-AA, the relative integrated peak intensity associated with the C-protomers is higher than that for 1-AN, which is rationalized by the greater thermodynamic preference for C-protonation (Table ).

4.

TIMS mobilograms of mass-selected (a) 1-ANH⁺ and (b) 1-AAH⁺. The black, thick lines correspond to the recorded mobilograms. The colored thin lines are Gaussian fits of the experimental peaks. The dashed vertical lines correspond to the V _elute values employed to select specific tautomers for the IRMPD spectroscopy experiments.

Hence, in line with refs and , our experiments show that the protomers can be analyzed by ion mobility, but the question remains whether the selected isomer is retained until probed by IRMPD spectroscopy in the ICR cell. Ion mobility (in whatever form) relies on interaction with a buffer gas, which can possibly promote structural changes or even dissociation. ,,,− The tautomerization reactions connecting the different protomers of the amino-acenes involve barriers of intermediate energy, typically smaller than dissociation reactions but larger than conformational changes. Therefore, we spectroscopically characterize the ion structure after mobility separation to investigate whether such an isomerization occurs.

IRMPD Spectroscopy on Mobility-Selected Ions

To characterize the protonation site of 1-ANH⁺ and 1-AAH⁺ beyond theoretical calculations and TIMS data, IR spectra of the m/z features of interest are recorded with and without TIMS selection prior to IR interrogation. With the TIMS switched off, ions of all cross sections are transmitted to the ICR cell where their IRMPD spectrum is recorded; see Figure S1 in the SI. The mobility-selected IR ion spectra are obtained by fixing V _elute at a constant value indicated by the dashed lines in Figure , while stepping up the FELIX laser wavelength. For 1-AAH⁺, IRMPD spectra for TIMS-separated ions of peaks A, B, and C are shown in Figure b. For 1-ANH⁺ in Figure a, the C-protonated tautomers are not resolved in our TIMS (Figure a), so we recorded an IRMPD spectrum with V _elute at the midpoint of the broadened mobility peak, presumably corresponding to a mixture of at least two C-protomers.

5.

Experimental IRMPD spectra of TIMS-selected (a) 1-ANH⁺ and (b) 1-AAH⁺ ions (thick lines), compared to theoretical spectra of selected protomers (shaded plots): N- and C3-protonated for 1-ANH⁺ and N-, C3-, and C1-protonated for 1-AAH⁺ (top to bottom).

Before presenting a full analysis of the IR spectra based on DFT-predicted IR spectra for the different protomers, it is useful to qualitatively inspect the spectra in the 3 μm region (Figure ), where one expects to observe clear differences between N-protonated −NH₃ ⁺ species, having three NH-stretch modes, and C-protonated species (−NH₂), having two NH-stretches, as clearly displayed by the distinct computed spectra in Figure . For 1-ANH⁺, the two experimental mobility-selected spectra indeed have a distinct appearance, although the bands partially overlap. The IR bands of TIMS peak A corresponding to the N-protonated isomer appear more red-shifted compared with those of peak B, in agreement with the weaker N–H bonds in an ammonium moiety versus an amino moiety. In contrast to 1-ANH⁺, the mobility-separated ions for 1-AAH⁺ feature very similar spectral profiles in the N–H stretch range with just slight relative intensity differences. This common spectral signature resembles that computed for C-protonated species.

This qualitative spectral analysis leads to the preliminary conclusion that, for 1-ANH⁺, the mobility-separated species remain largely in their selected tautomeric state until they reach the ICR cell, where IR spectroscopic analysis takes place. In contrast, in mobility-selected 1-AAH⁺, interconversion of the protonation site occurs after TIMS separation so that the ions end up in the lower-energy C-protonated form when they are probed by IRMPD spectroscopy. In other words, N-protonated ions appear to remain kinetically trapped in 1-ANH⁺ but interconvert to C-protonated isomers in 1-AAH⁺. We perform a computational investigation of the underlying potential energy surfaces (PESs) to further address these observations.

Computed PES for Tautomerization

The PESs for intramolecular proton transfer in 1-ANH⁺ and 1-AAH⁺ are presented in Figure . The calculated pathways show that the initial proton transfer from the amino group to the ipso-carbon atom is the rate-limiting step for both systems. The energy of this transition state is ∼210 kJ mol^–1 in 1-ANH⁺ and nearly 10 kJ mol^–1 lower for 1-AAH⁺. Moreover, the endothermicity forming the ipso-C-protonated intermediate is almost 30 kJ mol^–1 lower for 1-AAH⁺ than that for 1-ANH⁺. From here on, the PESs of the two systems are qualitatively similar, although the relative energies of all barriers and intermediates are 10–20 kJ mol^–1 lower in 1-AAH⁺ versus 1-ANH⁺, which is rationalized by the greater charge delocalization for C-protonated isomers in larger aromatic systems. The meta-protonated intermediate is thermodynamically unfavorable due to the electron-donating character of the amino group. Our computed PES for 1-ANH⁺ shows good agreement with that reported by Kumar and Attygalle, who investigated kinetic and thermodynamic control of protonation in 1- and 2-aminonaphthalene.

6.

Potential energy surface at the B3LYP/6-31++G­(d,p) (black dashed lines) and MP2/6-31++G­(d,p) (red dotted lines) levels of theory for intramolecular proton hopping in protonated 1-AN (a) and 1-AA (b). Gibbs energies at 298 K in kJ mol^–1 of transition states and intermediates are relative to those of the N-protomer.

The lower TS-barriers, the lower endothermicity of the intermediates, and the greater the exothermicity of the final para-protonated product for 1-AA as compared to 1-AN qualitatively help to explain the observed differences in proton transfer efficiency between the two molecules. N-protonated 1-AA is more likely to undergo intramolecular rearrangement, whereas N-protonated 1-AN is more likely to remain kinetically trapped. Valadbeigi and Causon reported similar behavior occurring in the ion source (prior to ion mobility analysis) for the tautomers of polycyclic aromatic amines. For 2-aminonaphthalene, they reported a barrier for the first proton migration step of 230 kJ mol^–1, close to our value of 210 kJ mol^–1 for 1-AN.

IR Spectra at Different Collision Energies

The IRIS spectra recorded for the TIMS separated protomers suggest that intramolecular proton transfer around the aromatic moiety occurs after mobility separation, i.e., after the TIMS tunnel and before ions are irradiated in the ICR cell. Recently, Stroganova et al. found evidence for dissociation of multiply charged peptide oligomers occurring at the exit of the TIMS. Earlier, Morsa et al. noted dissociation of benzylpyridinium “thermometer” ions occurring after the TIMS tunnel and were able to clearly distinguish it from dissociation occurring before (likely in the ion source) or in the TIMS tunnel. These authors also argue that dissociation occurring in the TIMS tunnel does not result in a broadening of peaks in the mobilogram; rather, fragment ions are observed at the elution voltage of the fragment, i.e., as if they were created before the TIMS. Translating to our experiment, the spectroscopic observation of C-protomers while selecting the N-protomer in the TIMS must be due to tautomerization occurring after, and not in, the TIMS tunnel.

The rate-limiting steps in the proton migration pathways of the protonated amino-acenes (Figure ) are comparable with the dissociation energies of the benzylpyridinium thermometer ions used by Morsa et al. (125 to 225 kJ mol^–1). Depending on instrumental settings, a scenario where 1-AAH⁺ undergoes rearrangement but 1-ANH⁺ does not (or less so) is then plausible. The work by Morsa et al. further demonstrated that post-TIMS ion activation occurs especially in the collision cell. Therefore, we explored the effects of the collision cell voltage on the spectroscopically probed protonation site of 1-AA. We selectively transmit peak A (N-protonation) through the TIMS and manipulate the collision cell voltage to assess spectroscopically to what extent it influences tautomerization to the C-protomer.

The original IRIS spectra in Figure were recorded at a collision cell voltage of −3.0 V, recommended as a generic setting by the manufacturer. Figure compares IRMPD spectra recorded with this voltage set to −5.0 and +1.5 V, with other settings in the −8.0 V to +3.0 V range shown in Figure S2 in the SI. Upon going from −5.0 V (harsh) to +1.5 V (soft), we observe that the IR bands near 3400 and 3500 cm^–1, correlating with the C-protomer(s), decrease in intensity along with an increase of the band near 3300 cm^–1, which is due to the N-protomer. Increased formation of the C-protomer at lower collision cell voltages indeed suggests that ion activation and rearrangement occur in the collision cell.

7.

IRMPD spectra of TIMS peak A of 1-AAH⁺ for two different collision cell voltage settings: under harsh conditions (−5.0 V, purple thick line), more tautomerization to the C-protomer is observed than under soft conditions (+1.5 V, green thin line). The inset provides a quantitative view of the C tautomer population as a function of these collision cell settings.

To quantify the extent of tautomerization, we analyzed the IRMPD yield rather than the fragment fluence; see eqs and above and note that the IRMPD yield ranges from 0 to 1. Furthermore, we removed the lens from the laser beam path to obtain full overlap between laser beam and ion cloud. At 3400 cm^–1, we selectively excite the C-protomer (see computed spectra in Figure b) and verified that irradiation with 15 laser pulses removes all C-protomers from the ion population (see Figure S3 in the SI). The IRMPD yield of about 0.75 at a voltage of −5.0 V then indicates that 75% of the ions are in the C-protonated form after TIMS selection of the N-protomer. At a voltage of +1.5 V, this fraction drops to about 55%. The fact that, even at the mildest collision cell conditions, a significant fraction of the ions still converts to the C-protomer, likely suggests that ion activation also occurs elsewhere, for instance, in the exit funnel of TIMS, but this requires further investigation. Finally, we note that removal of the focusing lens improves the laser-ion cloud overlap but can suppress weaker bands in the IRMPD spectrum.

Spectroscopic Analysis

The above analysis suggests that scrambling of the TIMS-separated N- and C-protomers occurs for 1-AAH⁺ but less so for 1-ANH⁺, which aids in the analysis of the IRMPD spectra of the TIMS-selected ions. In the top panel of Figure a, the 3 μm spectral range suggests that mobility peak A of 1-ANH⁺ contains mostly the N-protonated species with some admixture of C-protomers suggested by the absorption band at 3385 cm^–1, coinciding with the strongest band in the bottom panel. In the 1000–2000 cm^–1 range, the experimental spectrum indeed matches reasonably with the computed spectrum for the N-protomer (1-ANNH₃ ⁺). The most intense experimental band at 1450 cm^–1 corresponds to the umbrella mode of the −NH₃ ⁺ group. In the blue, a moderate–intensity, significantly broadened feature is observed, which is due to NH bending, along with nearby weaker absorptions; the shoulder toward ≈1640 cm^–1 is not well accounted for in the spectrum predicted for the N-protomer and is speculated to be due to the admixture of the C3-protomer, which possesses its strongest band at this frequency, as seen in the bottom panel of Figure a.

The 3 μm spectrum of mobility peak B (bottom panel of Figure a) suggests a C-protonated ion, and we overlay the experimental spectrum with the predicted spectrum for the lowest-energy C3-protomer. The NH₂ umbrella mode at 1650 cm^–1 gives rise to the most intense band in this range of the spectrum, with the shoulder at 1588 cm^–1 assigned as the CC stretch mode. The strong, broadened band centered at 1475 cm^–1 is due to closely spaced CC stretch modes, combined with in-plane C–H and N–H bending; the close proximity of two strong bands may artificially enhance the intensity in an IRMPD spectrum. , A CH bending mode is present at 1416 cm^–1 and the CH₂ scissor mode produces the weak band at 1354 cm^–1. Between 1200 and 1300 cm^–1 and at 3300 cm^–1, more features are observed than those predicted for the C3-protomer. Indeed, the broadened profile of peak B in the mobilogram suggests the presence of multiple C-protomers. Inductive effects shift the NH stretch bands of the ipso (C10) and ortho (C1) protomers slightly to the red as compared to the other C-protomers (see Figure S4 in the SI), so their presence may explain the absorption near 3300 cm^–1. Several C-protomers have strong predicted bands near 1300 cm^–1, providing further evidence for their presence in mobility peak B. The broadened profile of this peak suggests that they are formed in the source, in line with refs and , and post-TIMS activation may lead to further scrambling.

Mobility peak A of 1-AAH⁺ ought to be the N-protomer. However, as argued above, the majority of TIMS-selected ions convert to C-protomer(s) in the collision cell. Indeed, also in the 1000–2000 cm^–1 range, the computed IR spectrum for the N-protonated ion does not match the experiment (top panel of Figure b). The most intense band in the observed spectrum is shifted by 70 cm^–1 from the predicted umbrella mode in 1-AANH₃ ⁺ and, moreover, falls at the same position as that in 1-AAC₃ H ⁺ in the middle panel.

The experimental IR spectrum of TIMS peak B (middle panel in Figure b) closely resembles the prediction for the lowest-energy protomer (C3) in both spectral ranges, confirming its assignment to the para-carbon protomer. The NH₂ umbrella mode is located at 1647 cm^–1 and the bands at 1605 cm^–1 and 1615 cm^–1 correspond to CC-stretch modes. The bands at 1455 and 1415 cm^–1 have mixed CC stretch and NH bending character and the band at 1370 cm^–1 has mixed CC stretch and CH bending character. The spectrum in the 3 μm range resembles that of C-protonated 1-AN closely, both experimentally and theoretically; even the small apparent shift of the symmetric and antisymmetric NH₂ stretches relative to theory is reproduced.

TIMS-peak C is a minor feature in the mobilogram of 1-AAH⁺. From its position in the mobilogram and its NH stretch spectrum (bottom panel in Figure b), we derive that it must correspond to one or multiple unresolved C-protomers. However, comparison of the experimental spectrum with that predicted for the second lowest-energy C-protomer, 1-AAC₁H⁺, is disappointing. The strong band at 1440 cm^–1 is likely associated with the NH₂ umbrella mode, but its position deviates by about 70 cm^–1 from the position predicted for the C₁-protomer. Predicted IR spectra for all C-protomers in addition to C1 and C3 are displayed in Figure S4 in the SI and show significant variation. However, none of them provide a decent match with the experimental spectrum. We note that the main peak, supposedly due to the NH₂ umbrella mode, overlaps exactly with that of the C3-protomer, i.e., mobility peak B in the middle panel. Moreover, experimental IR spectra for mobility peaks A and C are very similar and deviate only slightly from that for peak B, mainly in the band intensities between 1500 and 1700 cm^–1. We therefore speculate that the protomer(s) giving rise to mobility peak C tautomerize after the TIMS, just as the N-protomer. The small intensity deviations are then due to some admixture of other C-protomers in the ion population, such as 1-AAC₁H⁺ and 1-AAC₅H⁺, which are low in energy and display lower band intensities in the 1500–1700 cm^–1 range.

In the 3 μm region, vibrational bands related to CH stretching have not been observed in the experiments, in agreement with their low predicted oscillator strengths.

Conclusions

A novel FTICR-MS platform with combined TIMS and IRIS capabilities was applied to unravel the protonation sites of the amino-acenes 1-AN and 1-AA in the gas-phase. The TIMS mobilogram revealed the presence of both N- and C-protonated tautomers despite significant differences in their thermodynamic stability, analogous to what several ion mobility studies had shown recently as well. , Where these previous studies focused mainly on the use of ion mobility analysis to evaluate the distribution of protomers evolving from the ion source, our study employs ion spectroscopy to probe the protomer distribution after ion mobility selection. We find that proton migration may occur after the TIMS ion mobility stage, removing the protomer selection imposed and driving the species to the lower-energy C-protomers. We attribute this to ion activation occurring downstream from the TIMS as was previously observed in TIMS as well as traveling-wave and differential IMS instruments using “thermometer” ions. ,− Their dissociation thresholds are in the range of the rate-limiting transition states for N-to-C proton transfer (about 200 kJ mol^–1) in the protonated amino-acenes studied here. Tuning the conditions in the collision cell of the instrument indicates that tautomerization occurs here.

Referring to the conclusion of Kumar and Attygalle that a mobilogram may not reflect the actual distribution of amino-acene protomers in the ion source, we may add that the mobilogram may not even be representative of the distribution of protomers downstream from the IMS stage. In IM-MS investigations, where ion mobility separation is followed by MS­(/MS) analysis, this may remain unnoticed. While this may have no consequences for studies employing (T)­IMS as an analyzer, one should be cautious using it as a filter, even if no dissociation is observed in the MS analysis.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jasms.5c00164.

• Additional experimental details and data derived from TIMS measurements; calculations of the lowest-energy isomers of 1-AA at additional levels of theory; IRMPD spectra with and without TIMS isolation and equation relating individual mobility-selected IRMPD spectra to spectra recorded without TIMS selection; IRMPD spectra of the TIMS selected N-tautomer of 1-AAH⁺ recorded at different collision cell voltage conditions; computed IR spectra of the other possible tautomers for 1-ANH⁺ and 1-AAH⁺; XYZ-coordinates of the B3LYP optimized geometries of the tautomers and transition states involved in the proton scrambling (PDF)

The authors declare no competing financial interest.