Keisuke
Hirata
acd,
Kyota
Akasaka
ab,
Otto
Dopfer
*de,
Shun-ichi
Ishiuchi
*acd and
Masaaki
Fujii
*abd
aLaboratory for Chemistry and Life Science, Institute of Innovative Research, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama, 226-8503, Japan. E-mail: mfujii@res.titech.ac.jp; ishiuchi.s.aa@m.titech.ac.jp
bSchool of Life Science and Technology, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama, Kanagawa 226-8503, Japan
cDepartment of Chemistry, School of Science, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, Tokyo 152-8550, Japan
dInternational Research Frontiers Initiative, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama, 226-8503, Japan
eInstitut für Optik und Atomare Physik, Technische Universität Berlin, Hardenbergstrasse 36, 10623 Berlin, Germany. E-mail: dopfer@physik.tu-berlin.de
First published on 19th January 2024
Proton transfer (PT) is one of the most ubiquitous reactions in chemistry and life science. The unique nature of PT has been rationalized not by the transport of a solvated proton (vehicle mechanism) but by the Grotthuss mechanism in which a proton is transported to the nearest proton acceptor along a hydrogen-bonded network. However, clear experimental evidence of the Grotthuss mechanism has not been reported yet. Herein we show by infrared spectroscopy that a vehicle-type PT occurs in the penta- and hexahydrated clusters of protonated p-aminobenzoic acid, while Grotthuss-type PT is observed in heptahydrated clusters, indicating a change in the PT mechanism depending on the degree of hydration. These findings emphasize the importance of the usually ignored vehicle mechanism as well as the degree of hydration. It highlights the possibility of controlling the PT mechanism by the number of water molecules in chemical and biological environments.
Experimental evidence for the Grotthuss mechanism has been obtained by the photo-induced PT dynamics of the rhodopsin and yellow proteins.7,16 However, it is limited only to photo-initiated PT processes in the excited electronic state. Significantly, it is not straightforward to transfer this conclusion to ground-state PT phenomena, like those in aqueous solutions, because of its stochastic character. Indeed, many studies on ground-state PT in the condensed phase suggest the Grotthuss mechanism simply because of the low barrier for ion conductivity without presenting direct evidence.17–19 One of the approaches to characterize the PT mechanism is the assignment of intermediates. In the theoretical framework of the Grotthuss mechanism in aqueous solution, a proton of an H3O+ ion hydrated by three water molecules (Eigen ion, H9O4+)1 is transferred to its nearest H2O to form another Eigen structure via the transient Zundel ion (H5O2+).20–22 Two-dimensional IR spectroscopy revealed that an asymmetric Zundel structure can be formed in aqueous solution.21 However, the presence of such intermediates is not always directly connected to the Grotthuss mechanism. The general difficulty in determining the PT mechanisms in various molecular systems prevents a deeper understanding of the mechanism–property correlations at the molecular level.
PT in hydrated clusters consisting of a finite number of water molecules23–28 and other solvent molecules29–31 provides a well-defined benchmark to explore the so far elusive PT mechanism. For example, p-aminobenzoic acid (PABA, Fig. 1) is a prototypical molecule for microhydration-induced PT.23,24,32–34 PABA has two protonation sites, i.e., the O- and N-protomers resulting from protonation of the carboxy and amino groups, respectively (Fig. 1). Their relative stability varies with the surrounding environment. The O-protomer is more stable in the gas phase while the N-protomer is more stable in a polar solvent.35,36 Studies using ion mobility mass spectrometry37 and infrared photodissociation (IRPD) spectroscopy23 claim that for the hydrated clusters of protonated PABA, PABAH+(H2O)n, generated by electrospray ionization (ESI), the N-protomer becomes detectable at n = 6, despite the spectral ambiguity in both the arrival time distributions and the IRPD spectra. In subsequent studies, we re-determined the threshold size (n) for the O → N protomer switching as n = 5 using cryogenic double ion trap IR spectroscopy.24 For the PT mechanism, none of previous studies (including ours) have provided any experimental evidence whether the vehicle or the Grotthuss mechanism takes place.23,24,37 The previous studies using ion mobility and IRPD claim the Grotthuss mechanism because their density functional theory (DFT) calculations show a stable structure for n = 6, in which the O- and N-protonation sites are bridged by a H-bonded water chain. However, no experimental evaluation for this structure was presented. We also applied DFT calculations for n = 5 and found that a structure locally hydrated around the protonated amino group is significantly more stable than the ones in which the O- and N-sites are bridged by an H-bonded water chain, and thus the only observed hydration motif. Based on this exclusively observed hydration motif, we suggested the vehicle rather than the Grotthuss mechanism, although again we had no direct experimental evidence regarding which PT mechanism is in operation for n ≥ 5.
To determine the PT mechanisms operating in hydrated PABAH+ at the molecular level, we combine herein cryogenic double ion trap IR spectroscopy38–42 with isotopic labelling. Deuterated hydrated clusters of PABAH+ can be produced by introducing D2O vapour into the reaction trap. One can readily expect that, starting from the O-protomer, the Grotthuss-type PT gives a singly deuterated N-protomer (NH2D1) in the deuterated hydrated clusters because the proton source is D2O (in Fig. 1). The vehicle-type PT yields both deuterated (NH2D1) and nondeuterated (NH3) N-protomers via the transfer of HD2O+ (Fig. 1). Clearly, the detection of NH3 provides direct evidence for the vehicle mechanism. However, the appearance of NH3 can hardly be probed by conventional IRPD spectroscopy because its vibrational signatures are largely red-shifted and broadened by H-bonding with water molecules, causing spectral congestion.23,43
This problem is overcome herein by applying the novel collision-assisted stripping (CAS) method, developed recently by our group.24,27,43 In CAS experiments, one measures IRPD spectra of the PABAH+ monomer core after removing all hydrated water molecules of the cluster by soft collisions in the ion trap, allowing for the clear spectroscopic assignment of deuterated isotopologues of PABAH+, as the IR spectra of the solvent-free ions are not congested. Herein, we perform CAS-IRPD of PABAH+(D2O)n to determine the PT mechanisms as a function of the degree of hydration (see Experimental in the ESI†). We obtain clear evidence of the vehicle mechanism for n = 5/6 and suggestions of the Grotthuss mechanism for n = 7, respectively, suggesting a switch in the PT mechanism between n = 5 and n = 7. The consistency with the IRPD spectra of the hydrated clusters is also discussed.
The CAS-IRPD spectrum of PABAH+(D2O)5 shows new bands in the 3220–3320 cm−1 range, which are absent for n ≤ 4 (Fig. 2). The spectrum of these bands measured with a twice higher laser intensity is also shown as the inset. These bands are assigned to NH stretches of the N-protomer,24,44 indicating the appearance of the N-protomer core upon O → N PT. A further new band at 2416 cm−1 is attributed to an ND stretch of the N-protomer based on its frequency (see Fig. S1 in the ESI† for detailed band assignments), confirming the appearance of the N-protomer core at n = 5. The threshold size (n = 5)24 for O → N protomer switching does not change upon H2O → D2O substitution. Based on the given assignment, the other new band at 2620 cm−1 can be assigned to the OD stretching of the N-protomer core (νOD (N), Fig. S1†). The remaining two bands at 2585 and 2631 cm−1 are then attributed to νZ-OD (O) and νE-OD (O), respectively (Fig. S1†). Note that the OD stretches also appear for n = 4, indicating H/D exchange at the carboxylic group without protomer switching. One may wonder why H/D exchange happens at such low temperatures (T = 80 K), at which protonated glycine is not deuterated by clustering with D2O.45 This observation is rationalized here by the high acidity of the protonated carboxy group. Recently, the Johnson group reported IRPD spectra of hydrated deuterated isotopologues of the N-protomer (ND3+C6H4COOD(H2O)n) and found that the acidic COOD group is more susceptible to H/D exchange than the charged ND3+ group.46 The protonated C(OH)2+ group of the O-protomer may more readily replace H by D.
The CAS-IRPD spectra of PABAH+(D2O)6,7 also exhibit the ND/NH stretches of the N-protomer core in the 2380–2470/3220–3320 cm−1 ranges, but their band patterns are more complicated than that of n = 5 (Fig. 3a). To understand the PT mechanism, one must assign the vibrational bands of deuterated (NH2D1) and nondeuterated (NH3) N-protomers without ambiguity. For this purpose, we measured the IR spectra of nondeuterated (m/z 138), monodeuterated (m/z 139, NH3OD1 and NH2D1), and dideuterated (m/z 140, NH2D1OD1 and NH1D2) forms of the bare N-protomer in the NH stretch range, which are shown in Fig. 3b along with molecular structures. Details of the assignments are summarized in the ESI.† Briefly, three prominent bands in the spectrum of m/z 138 are assigned to NH3 stretching bands (coloured in green, ) of the N-protomer of the fully hydrogenated species (NH3). The new bands indicated by blue colour in the spectrum of m/z 139 are NH2 stretching bands of the amino-deuterated N-protomer (NH2D1). The bands due to the NH3 stretching (green) are overlapped with transitions of the coexisting N-protomer deuterated at the carboxylate group (NH3OD1). The newly observed band (purple) in the spectrum of m/z 140 is an NH stretch mode of the dideuterated amino group of the N-protomer (NH1D2). Among these bands, the vibrational transitions of NH3 () are the most important ones because they indicate the presence of the N-protomer (NH3) which can be generated only by the vehicle mechanism (Fig. 1).
Fig. 3 (a) CAS-IRPD spectra of PABAH+(D2O)n (n = 5–7) in the 3200–3350 cm−1 range (reprinted from the inset spectra in Fig. 2). (b) IRPD spectra of PABAH+ (m/z 138), singly deuterated protonated PABA (m/z 139), and doubly deuterated protonated PABA (m/z 140). These spectra are necessary to assign the CAS-IRPD spectra because the monomer core after the CAS event includes several deuterated isotopomers. The green-, blue-, and purple-colored bands are assigned to NH stretches of NH3, NH2D1, and NH1D2, respectively. The band marked by an asterisk is assigned to a combination band of the O-protomer (in-plane NH2 bend24,44 at ∼1660 cm−1 + aryl C–C stretch24,44 at ∼1600 cm−1). This type of combination band is also observed for an analog of PABA, ethyl p-aminobenzoate.43 |
Let us check the presence of (green), the marker bands of the vehicle mechanism, in the CAS-IRPD spectra of PABAH+(D2O)n with n = 5–7. Fig. 3a reproduces expanded spectra in the 3200–3350 cm−1 range. The green marker bands obviously appear in the spectra of n = 5 and 6, but are clearly absent in the n = 7 spectrum. This result directly indicates that the vehicle mechanism operates in the n = 5 and 6 clusters but not for n = 7. One cannot directly conclude from the CAS-IRPD spectra whether the Grotthuss mechanism operates in n = 5 and 6. The observed bands in the n = 7 spectrum correspond to N-protomers of NH2D1 (blue) and NH1D2 (purple). The former, NH2D1, can be generated by both the vehicle and Grotthuss mechanisms. As the absence of the vehicle mechanism for n = 7 is clear, the blue bands can only arise from the Grotthuss PT. The latter, NH1D2, cannot be generated by a single O → N PT reaction via either the Grotthuss or the vehicle mechanism. The observed multiple deuteration thus is ascribed to H/D exchange by D2O or the interconversion between N- and O-protomers by O → N and N → O back reactions (Fig. S5 in the ESI†). Here, we cannot fully exclude the possibility that the absence of , the signature of the vehicle mechanism, may arise from the H/D exchange reaction because such an H/D exchange is observed for n ≥ 4 (see above). However, if this scenario were the case, H/D exchange also modulates the intensities of in the spectra of n = 5 and 6 where the band is clearly observed. This observation means that the relative intensity of for n = 7 is rather weak in comparison to those for n = 5 and 6, and the vehicle PT mechanism is negligibly minor for n = 7.
The size-dependent PT mechanisms proposed by the experiment are consistent with the molecular structures of microhydrated PABAH+. The bridged structures in which the NH3+ and COOH groups are connected by an H-bonded water chain are more abundant than the unbridged structures for the N-protomer at n = 7, whereas the bridged structures become minor for n = 6 and are not observed for n = 5 (Fig. S6 in the ESI†).24 This strongly suggests that the water-bridged intermediates for the Grotthuss-type PT are stabilized at n = 7 (and partly stabilized at n = 6). In this sense, the Grotthuss mechanism may be involved in part in the PT at n = 6. These observations show that the reaction barriers of the Grotthuss mechanism can be altered by the number of water molecules. The Grotthuss mechanism with a lowered reaction barrier would account for the minor contribution of the vehicle mechanism at n = 7. This finding gives suggestions for the poorly understood properties of PT in various systems. For example, it is known that proton conductivity in porous materials such as MOFs is sensitive to humidity.5,6 Some MOFs show a drastic decrease in the conductivity below a critical humidity,47,48 however, the reason remains unclear. The poor conductivity at low humidity has been hampering the application of MOFs to proton-conducting membranes in fuel cells. Our conclusion that the PT mechanisms would be changed from Grotthuss to vehicle with decreasing number of water molecules would rationalize such a drastic change in PT by humidity.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3sc05455a |
This journal is © The Royal Society of Chemistry 2024 |