Intramolecular Inhibition by Imidazole in Acid‐Catalyzed Hydrolysis of Protected Carbohydrates

Abstract

The present study reveals an unexpected anomaly observed in the acid‐catalyzed hydrolysis of the 5,6‐O‐isopropylidene group in 3‐O‐protected D‐gluco‐ and D‐allofuranose derivatives. Although the removal of the 5,6‐O‐isopropylidene protecting group is typically rapid and quantitative under acidic conditions, an unexpected inhibition of this reaction is observed for the two C3‐epimers, 3‐O‐imidazole sulfonyl moiety. X‐ray data show a two‐faced imidazole ring orientation in the crystal, while solution state NOE data reveal a critical interaction type between the isopropylidene and the imidazole rings. Advanced conformational searches coupled with ab initio molecular modeling illuminate and explain the NMR and kinetic data and lay the groundwork for the most plausible mechanism of this unprecedented inhibition. These results provide valuable insights into the cross‐coupling of carbohydrate O‐protecting groups and shed light on how specific ring orientations and steric effects can trigger the inhibition of an otherwise easily feasible reaction, such as an acid‐catalyzed hydrolysis.

Unexpected inhibition of the acid‐catalyzed hydrolysis of the 5,6‐O‐isopropylidene group in 3‐O‐protected D‐gluco‐ and D‐allofuranose derivatives is observed. Structural and conformational analyses elucidate critical interactions between the isopropylidene‐protected dioxolane ring and imidazole rings, providing insights into the interplay of carbohydrate O‐protecting groups and steric effects on reaction mechanisms.

Introduction

The isopropylidene group is commonly used in organic chemistry as a selective group to protect vicinal diols and/or adjacent hydroxyl groups. For example, the successful acid‐catalyzed cleavage of the 5,6‐O‐isopropylidene group of D‐glucofuranose has been reported using various reagents such as 80 % aq. acetic acid, Nafion‐H, aq.HCl, HClO₄*SiO₂, etc.[ ¹ , ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² ] Under mild acidic conditions, this hydrolysis proceeds according to the regular mechanism shown in Figure 1.

Figure 1.

Regular mechanism of mild acidic hydrolysis of an isopropylidene‐protected vicinal diol A to the resulting vicinal diol B plus acetone.

The mechanism of such hydrolysis has been well studied and, according to the literature, is well understood.. ^[13] We were therefore surprised to find molecular conditions that prevent isopropylidene derivatives from being hydrolyzed as expected. We found a pair of imidazole containing stereoisomeric carbohydrate moieties where the protecting isopropylidene group cannot be successfully removed.

Thus, the current anomaly and failure of a “simple” acid‐catalyzed hydrolysis prompted the question: “What exactly is happening during this poorly performing and low‐yield reaction? What is the molecular background that causes this anomalous behavior?”

Results and Discussion

In order to perform a comprehensive analysis, we designed a molecular scaffold containing two pairs of chiral vicinal diols, both of which could be protected as isopropylidene derivatives, plus a fifth hydroxyl group between them to be esterified. The second glycolic subunit was chosen to have a built‐in reference group and a reference reaction within the same moiety. A prerequisite for the successful isopropylidene protection of the vicinal diol lies in its stereochemistry: for stearic reasons, the addition should be syn, resulting in cis products. A possible molecular scaffold that fulfills the above requirements is α‐D‐glucopyranose and α‐D‐allopyranose, a pair of C3 epimers, since they both have 5 OH groups, all with the correct stereochemistry. Several sugar derivatives have been synthesized, but the present report focuses on the following five. They were all subjected to a quantitative analysis of their hydrolysis, starting with the reference molecule 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐benzoyl‐α‐D‐glucofuranose (Glc‐Bz or 1.). The C3 pair of the 3‐tosyl derivatives, namely 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐toluenesulfonyl‐α‐D‐glucofuranose (Glc‐Tos or 2.) and 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐toluenesulfonyl‐α‐D‐allofuranose (All‐Tos or 3.) were synthesized and analyzed. It turns out that the target C3 epimers of interest for a better understanding of the anomalous hydrolysis envisaged above are the 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐(N‐imidazole‐1‐sufonyl)‐α‐D‐glucofuranose (Glc‐Imz or 4. ) and 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐(N‐imidazole‐1‐sufonyl)‐α‐D‐allofuranose (All‐Imz or 5.) molecules (Figure 2).[ ¹³ , ¹⁴ ]

Figure 2.

The 5 different sugar derivatives are used for the quantitative analysis of hydrolysis, namely the molecular structure of 1. 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐benzoyl‐α‐D‐glucofuranose (Glc‐Bz); 2. 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐toluenesulfonyl‐α‐D‐glucofuranose (Glc‐Tos); 3. 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐toluenesulfonyl‐α‐D‐allofuranose (All‐Tos); 4. 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐(N‐imidazole‐1‐sufonyl)‐α‐D‐glucofuranose (Glc‐Imz); 5. 1,2;5,6‐Di‐O‐isopropylidene‐3‐O‐(N‐imidazole‐1‐sufonyl)‐α‐D‐allofuranose (All‐Imz).

During the acid‐catalyzed hydrolysis of the isopropylidene group, the following general observations were made.

1. The opening of the 1,2‐OH‐containing rings always proceeded promptly thus the built‐in reference reaction proceeds as expected based on the literature data.

2. Both the 3‐O‐tosyl and the 3‐O‐benzoyl derivatives show no signs of abnormality: both the 1,2‐ and the 5,6‐O‐isopropylidane groups hydrolyze normally as described in the textbooks. The C3‐O‐benzoyl derivative of the gluco stereoisomer was synthesized to evaluate and quantify the electron withdrawing effect of the sulphonyl group.

3. The unexpected interferences occur only between the isopropylidene and imidazole groups, but for both C3 stereoisomers. Therefore, this study focuses on the cause of the anomaly observed for 1,2;5,6‐di‐O‐isopropylidene‐3‐O‐(N‐imidazole‐1‐sulfonyl)‐α‐D‐glucofuranose (Glc‐Imz) and its allo‐epimer (All‐Imz). ^[15] A better understanding of the neighboring group effect between the imidazole and isopropylidene rings is of key importance, especially since the imidazole ring (e. g., in the case of the His of serine proteases) is known to catalyze and not inhibit similar reactions.

We found that under mild acidic conditions, e. g., AcOH/H₂O (8/2), the cleavage of the 5,6‐O‐isopropylidene group is thus inhibited if complemented by the imidazole ring, the rate constants of such normal and abnormal hydrolyses are compared. We hoped that by comparing the reactivity of these diastereomeric derivatives, we can decipher the influence of the configuration and/or the functional group on their hydrolysis rate.

Kinetic Measurements

Reactions were monitored by HPLC for approximately 250 hours. For the imidazole derivatives, the monitoring period was extended by several additional days due to the extremely slow hydrolysis. During this extended period, side products were observed, specifically in the case of the imidazole derivatives. These minor side products result from the hydrolysis of the 1,2‐O‐isopropylidene group, which occurs at a significantly slower rate—approximately 80 times slower than the primary reaction. The ring opening can be observed through the splitting of the anomeric proton doublet and the splitting of the imidazole C2−H proton signals in 1H NMR spectra.

We determined the reaction rate (molecule 1,2 and 3) based on the consumption of the reactants and found that the hydrolysis reaction follows first‐order kinetics. This conclusion was corroborated by analyzing the integrals of the HPLC chromatograms.

Kinetic measurements indicate that molecules 1, 2, and 3 produced the vicinal diol as the main product (Figure 3) in a relatively short time, being 20 times faster than the hindered reactions of molecules 4 and 5. The hydrolysis of molecules 2 and 3 occurs slightly slower, as shown by their rate constants, due to the close proximity of the sulfonyl moiety, which is a stronger electron‐withdrawing group compared to the carbonyl moiety of molecule 1 (Figure 3A, Table 1).

Figure 3.

A) The relative concentration of the substrates changes over time, with hydrolysis completing within the first 24 to 48 hours for most substrates, except for substrates 4 and 5. B) The relative concentration of the product was monitored and tracked using NMR measurements, for molecules 4 and 5. Raw experimental data were fitted by equation $y\sim \left(a_0-a_1\right)\mathbf{e}\mathbf{x}\mathbf{p}\left(-k{x}_n\right)+a_n$ to obtain rate constants using the Desmos program. ^[15]

Table 1.

Comparison table showing the rate constants and Gibbs free energy for the hydrolysis reaction of molecules 1–5. The Gibbs free energy (ΔG) was calculated from the hydrolysis rate constants. Bold numbers indicate that the reaction equilibrium favors the substrates, thus inhibiting the reaction. All calculations were performed at T=298.15 K. $\Delta G=-RTlnK$ .

Molecules	k1	k‐1	Kexp	K	a1	ΔG (KJ/mol)
Glc‐Bz	0.0766	0.0041	0.0806	18.7414	0.0507	−7.2650
Glc‐Tos	0.0291	0.0026	0.0317	11.0255	0.0832	−5.9499
All‐Tos	0.0069	0.0003	0.0072	24.0869	0.0399	−7.8870
Glc‐Imz	0.0113	0.0374	0.0487 0.0105*	0.3023	0.7678	2.9652
All‐Imz	0.0068	0.0881	0.0950 0.1039*	0.0773	0.9282	6.3459

*obtained using the same calculation but based on NMR measurements, closely align with those calculated from the HPLC data.

In contrast, the hydrolysis of both the gluco and allo imidazolsulfonyl derivatives (4 and 5) is significantly inhibited. The difference between molecules 4 and 5, and molecules 2 and 3, is solely the imidazole ring, as both sets of molecules contain the same sulfonyl moiety. Therefore, the decreased reaction rate for molecules 4 and 5 (0.3 and 0.07, respectively) can be attributed to the imidazole moiety. Nonetheless, the reaction rate remains directly proportional to the concentration of a specific variable, as represented by the equation v=k[A].

Since the HPLC integrals for the molecules Glc‐Imz (4) and All‐Imz (5) were corrupted by the presence of minor side products, we sought to measure the reaction equilibrium using NMR. This approach allows us to observe the formation of side products directly.

We monitored the relative concentration of the acetone signal (2.2 ppm), a product of hydrolysis, against the anomeric proton signals, which were used as an internal standard. The NMR measurements were conducted over at least 24 hours to ensure accurate tracking of the reaction progress. (Figure 3B) The results, as summarized in Table 1, demonstrate that the experimentally determined equilibrium constant (K _exp) values from NMR closely align with those calculated from the HPLC data However, it should be noted that these results may also be influenced by the kinetic isotope effect.

Additionally, Table 1 shows that these reactions have a positive Gibbs free energy (molecule 4: ΔG=2.9652, molecule 5: ΔG=6.3459), indicating that the reaction equilibrium favors the substrates (Figure 3).

Initially, the reaction was puzzling because the imidazole group inhibited the acid‐catalyzed hydrolysis reaction. Nature provides numerous examples, where an imidazole moiety serves as a catalyst in proton‐based reactions, acting as a proton transfer moiety,[ ¹⁶ , ¹⁷ ] as illustrated by histidine within the catalytic center of chymotrypsin, trypsin, elastase, trombone etc.[ ¹⁸ , ¹⁹ , ²⁰ ] The two N‐atoms (π, τ) of the imidazole ring allow histidine to switch between protonated and unprotonated states. Therefore, histidine plays a role as a proton transfer moiety in catalyzing acid‐base reactions (Figure 4).

Figure 4.

In the catalytic triad of the trypsin enzyme, the imidazole ring of the His57 side chain functions as a proton transfer unit within serine proteases. In the case of N‐imidazol‐1‐sulfonyl fragments, the imidazole ring serves as a basic moiety in molecules 4 (Glc‐Imz) and 5 (All‐Imz).

Because the imidazole ring is covalently bonded to the sulfonyl moiety via the nitrogen atom, the other nitrogen atom of the imidazole sulfonyl group remains unprotonated, acting as a base and readily protonated even under mildly acidic conditions (Figure 4). The unprotonated imidazole ring exhibits varying acidity levels among its C−H atoms (C2, C4, C5), with the hydrogen atom at C2 being the most acidic. Ab initio computations have shown that the hydrogen bond energy within the imidazole C2‐H… OH₂ water complex is calculated to be ΔE=−3.01 kcal/mol. Upon protonation, the positive charge is distributed equally among the five atoms of the imidazole ring. Therefore, it is reasonable to assume that in our case this moderate hydrogen bond can be enhanced upon protonation of the imidazole ring.

Working Hypothesis

Initially, we hypothesized that the reaction was inhibited because the acidic medium protonates the basic nitrogen on the imidazole ring, potentially leading to interactions with the 5,6‐O‐dioxolane ring. To investigate this hypothesis, we conducted additional experiments including crystallography, NMR studies and computational simulations.

Interaction in the Solid Phase: X‐Ray Data

During the synthesis of molecule 4, an interesting byproduct was observed, namely (3,3’‐O‐bis (1,2;5,6‐di‐O‐isopropylidene α‐D‐glucofuranose)‐sulfate) (molecule 6 or Glc2‐S for short), which intriguingly forms a disaccharide where two glucopyranose molecules are connected by a sulfate group. This byproduct has not been previously described in the literature. Following purification of molecule 4 (Glc‐Imz) using silica gel column chromatography, both the main product and the byproduct were successfully crystallized from the eluent (EtOAc – 3.5 : 1) (Figure 5, Figure S1‐5, Table S1).

Figure 5.

Showing the superposition of the imidazole ring between two states in crystallographic data for molecule 4 (Glc‐Imz).

The crystallographic structure of molecule 4 (Glc‐Imz) reveals that the imidazole rings are close to the 5,6‐O‐dioxolane ring even in their unprotonated state. The “fuzziness” observed in the X‐ray structure suggests that the imidazole ring exists in a superposition of two statistically comparable states, where the imidazole ring center is only 3.8 Å away from the O5 of the 5,6‐O‐dioxolane ring and 4 Å from the nearest methyl group of the same ring. Despite these frozen states captured by crystallography, it is important to note that these interactions are dynamic and can vary in solution.

Currently, structural data are available only for the gluco form (Glc‐Imz). Therefore, we shifted our focus from the solid state to liquid media to investigate these interactions using NMR spectroscopy both epimer Glc‐Imz and All‐Imz.

Interaction in the Solution Phase: NMR Data

To further elucidate the interactions of the unprotonated form of Glc‐Imz in solution, we conducted NMR measurements in acetone, and acetic acid. The Nuclear Overhauser Effect (NOE) data revealed cross‐peaks between the three hydrogens of the imidazole ring and one of the methyl groups of the 5,6‐dioxolane ring. Based on the NOE data obtained in acidic media, it was observed that not only one methyl group, but both methyl groups are near the imidazole ring. This observation suggests that the two rings adopt a closer spatial arrangement under acidic conditions. A comparison of these NOE measurements clearly indicates the presence of an attractive force between the two adjacent rings in acidic media (Figure 6A, B).

Figure 6.

A) NOE cross peaks between the imidazole ring protons (δ=8.2 ppm (s, 1H), 7.65 ppm, 7.21 ppm and one of the methyl group protons (δ=1.20 ppm (s, 3H‐iPr) of the 5,6‐O‐isopropylidene in deuterated acetone (acetone – d₆). B) NOE cross peaks between the imidazole ring protons (δ=8.4 ppm (s, 1H), 7.67 ppm, 7.31 ppm and two of the methyl group protons (δ=1.18 ppm (s, 3H‐iPr) and 1.29 (s, 3H‐iPr) highlighted in orange) of the 5,6‐O‐isopropylidene in deuterated Acetic acid (AcOD‐ d₄/D₂O: 8/2).

Additionally, the hydrogens C2−H and C5−H of the imidazole ring exhibited cross peaks with C6‐2H and C5−H of the glucofuranose ring (Figure S5). This data shows that the two rings are indeed close even when the molecule is solvated.

To investigate our hypothesis regarding the inhibitory role of the protonated imidazole ring in hydrolysis, we performed 1H NMR experiments. Like the unprotonated state, the NOE data showed cross peaks between the hydrogens of the imidazole ring and one of the methyl groups of the 5,6‐dioxolane ring. However, notably, the cross peaks between the C2−H atom of the imidazole ring and the C(3)−H of the glucofuranose ring were smaller than in acetone. This reduction in cross peaks could be caused by isotope exchange with the deuterated solvent, or it may suggest that the imidazole ring has shifted or flipped to a more energetically favorable position (Figure 6, Figure S6).

It is reasonable to consider that if ring flipping can occur in the solid state and is frozen in the crystal structure, it can also be observed during NMR measurements in solvents. However, when the imidazole ring is protonated, one of these superpositions is favored. Furthermore, if the C2−H protons orient toward the lone pair of the C(5)‐O oxygen, the C(4)−H and C5−H hydrogens must point outward from the structure, thus not forming a cross peak with the methyl group of the 5,6‐O‐isopropylidene. Nevertheless, since the cross peaks of all the hydrogens of the imidazole ring with one of the methyl groups remain in the protonated state, there may a stabilizing effect. These observations highlight the significant influence of the proximity of these two rings on their structural and electronic properties.

Additionally, the same cross peaks were found between the imidazole ring and the methyl group of the 5,6‐O‐isopropylidene for the molecule 5, the allo derivative (All‐Imz) too. Figure S7,8.

Lone Pair↔π Stacking

A. Jain and co‐workers have studied the properties of the protonated imidazole ring and found that the positively charged imidazole ring can be attracted by water molecules, due to the partial negative charge of the O‐atom. Such interactions play a crucial role in stabilizing the molecular structure.,[ ²¹ , ²² ] An alternative binding mechanism between the imidazole ring and the dioxolane ring was explored. Such an interaction can led to a unique lp↔π (lp: lone pair) stacking arrangement, between the lone pair electron of oxygen and the positively charged imidazole ring. If such stacking occurs, one of the methyl groups of the 5,6‐dioxolane ring would be equally close to all the protons of the imidazole ring. The current NOE data could be further supported this scenario, showing a cross peak between all three imidazole protons and one methyl group of the dioxolane ring (Figure 6.). It is important to note that the same NOE cross peaks are observed in the case of protonated and unprotonated derivative too.

Interestingly, the C‐2 proton of the imidazole ring has sufficient acidity to undergo exchange with a deuterium isotope. Over time, the integral of the assigned peak δH(C2)=8.41 ppm gradually decreases, in agreement with literature data (Figure 7). ^[23]

Figure 7.

The decreasing peak height/integral indicates isotope exchange between the C2−H of the imidazole ring (δ=8.41 ppm) and the solvent (AcOD‐ d₄/D₂O: 8/2) after approximately 1 day of hydrolysis. The relative integral of the C2−H hydrogen decreased to 47 % of its original value. The ¹H resonance of TMS was used as the internal standard, which remained unchanged. The blue peaks were scanned at the beginning of the reaction, while the red peaks were scanned after 1 day from the same sample (Glc‐Imz in acidic solvent).

Interaction in Gas Phase Computational Approach – “Quantum Sufficit”

Starting from the crystallographic structures, we conducted a conformer search to identify the most stable unprotonated structure of Glc‐Imz as a gas phase model for our research. For this conformer search, we utilized the CREST software, which is sufficiently accurate and fast, thus reducing calculation time. ^[24] Figure S9 Subsequently, we protonated the imidazole τ nitrogen and further optimized at the Gaussian g16 software on B3LYP/6‐311G++ (d, p) basis level, as CREST is unable to handle charged molecules. Figure 8 The results indicated that in the unprotonated structure, the distance between the imidazole ring and the 5,6‐dioxolane ring decreased. However, upon protonation of the imidazole ring in Glc‐Imz⁺, the imidazole ring flipped and formed a hydrogen bond with the C (5)‐O of the 5,6‐dioxolane ring Figure 9.

Figure 8.

The most stable unprotonated conformer (population ~47 %) for Glc‐Imz, (green) and the most stable protonated conformer for molecule Glc‐Imz⁺ (magenta). The distances between the C2 carbon of imidazole ring and the O5 atom of 5,6 dioxolane ring are 3.02 Å.

Figure 9.

Alignment of structures from crystallographic data (blue), the unprotonated most stable conformer (green), and the most stable conformer of the protonated molecule Glc‐Imz⁺ (magenta).

Aligning the crystallographic and calculated structures for both protonated and unprotonated forms revealed that the imidazole ring remained stationary in the unprotonated Glc‐Imz. In contrast, in the protonated form (Glc‐Imz⁺), the imidazole ring flipped to form a hydrogen bond. This hydrogen bond should also be detectable in NMR data, where the C5−H and C2−H of the imidazole ring should distance themselves from the methyl group of the 5,6‐dioxolane ring. Considering that the imidazole ring can form a hydrogen bond with C5−H as well, it is plausible that in about half of the population, the hydrogen bond forms between the C5‐O of the 5,6‐ dioxolane ring and the C2−H and C5−H protons of the imidazole ring. Figure 9

From Figure 9, in the unprotonated state, the imidazole ring appears to exhibit an lp↔π interaction with the dioxolane ring. However, this interaction is very weak or merely superficial. Upon protonation of the nitrogen, this weak interaction readily transitions into a hydrogen bond. The conformer, where the two rings are similarly close, represents approximately 32 % of the population of the unprotonated allo conformer (All‐Imz). For the protonated allo conformers (All‐Imz+1: 31 %, All‐Imz+2: 61 %), refer to Supplementary Figure S10.

We conducted a structural analysis of the Glc‐Tos molecule to investigate the presence of (lp)↔π or other interactions in its unprotonated state. Our findings suggest that, similar to the most stable unprotonated conformer of Glc‐Imz, the aromatic ring of Glc‐Tos may engage in a possible (lp)↔π interaction with the lone pair of the oxygen atom in the 5,6‐dioxolane ring. This conformer could account for up to 34 % of the population, as illustrated in Supplementary Figure S9.

It should be noted that all calculations were performed in vacuo. Consequently, both the population and the structural preferences may vary in the presence of solvation forces, which could perturb the correlation between structures derived from spectroscopic data and those predicted in silico.

Importantly, both lone pair (lp)↔π interactions are observed in the unprotonated states of the Glc‐Imz and Glc‐Tos molecules. Since the aromatic chain of the Glc‐Tos molecule cannot be protonated under these conditions, we propose that the primary cause of the hindered reaction in the case of Glc‐Imz and All‐Imz arises from the hydrogen bonding interaction between the imidazole ring and the dioxolane ring. This intramolecular (CH⋯O) H‐bond is strong enough to stabilize the interaction but too weak to protonate the oxygen atom.

Additionally, Natural Bond Orbital (NBO) analysis of the protonated Glc‐Imz molecule revealed an overlap between the lone pair of the oxygen atom and the C(4)−H bond (Figure 10).

Figure 10.

Overlapping NBO orbitals between the C(4)−H bond of the imidazole ring and the lone pair of the oxygen atom in the dioxolane ring.

To evaluate our hypothesis that this hydrogen bond‐mediated inhibition is limited to weak acids, we attempted the hydrolysis under stronger acidic conditions. Using 1.5 N HCl/MeOH, the reaction was completed within a few hours, confirming that the inhibition occurs only in the presence of weak acids.

In summary, the attractive forces between the positively charged imidazole ring and the surrounding oxygen atoms (C5−O−iPr) are essential for establishing a stable molecular arrangement. These interactions significantly contribute to the hindered hydrolysis of the 5,6‐dioxolane ring in sugar derivatives bearing the 3‐O−N‐imidazole‐1‐sulfonyl group.

Conclusions

Our study has uncovered an unprecedented intramolecular inhibition mechanism in the acid‐catalyzed hydrolysis of acetone‐protected furanose molecules. Structural analyses revealed the critical role of a hydrogen bond (H‐bond) interaction between the imidazole and dioxolane rings, supported by kinetic measurements and NMR NOE data. These findings demonstrate that the inhibition arises primarily from the protonated imidazole ring. This behavior is particularly intriguing, as the imidazole ring, typically known for its catalytic activity in reactions, instead acts as an inhibitor in this context.

Additionally, isotope exchange observed in the imidazole ring during the reaction provides valuable insights into its dynamic role within the system. We also report two novel crystal structures and describe a unique bis‐sugar derivative, offering further structural insights.

These findings significantly advance our understanding of the reaction mechanism, emphasizing how specific ring orientations and steric effects can inhibit an otherwise straightforward reaction, even when involving a typically catalytic moiety such as the imidazole ring.

Experimental

Deuterated acetic acid ‐d₄ was purchased from Cambridge Isotope Laboratories, and D₂O was obtained from Eurisotop. All kinetic measurements were performed using High‐Performance Liquid Chromatography (HPLC). The analysis was conducted on an Agilent 1260 Infinity II system with a Diode Array Detector (DAD). The separation was conducted using a Poroshell 120 EC−C18 column (4.6×150 mm) at a column temperature of 40 °C. The eluent used for the analysis was water and acetonitrile. H‐NMR spectra were measured with the Bruker Avance 700 MHz instrument. Additionally, Mass Spectrometry (MS) spectra were obtained using a Bruker Esquire 3000+ tandem quadrupole mass spectrometer equipped with an electrospray ion source.

For molecule 6 (Glc2‐S) and 1 (Glc‐Bz) exact mass measurements were performed on a high‐resolution Q‐Exactive Focus hybrid quadrupole‐orbitrap mass spectrometer (Thermo Fisher Scientific, Bremen, Germany) equipped with heated electrospray ionization source. Samples were dissolved in acetonitrile‐water 1 : 1 (V/V) solvent mixture. Flow injection analysis was performed using a 50 μL/min eluent flow. Under the applied conditions, the compounds form sodiated molecules, [M+Na]⁺ in positive ionization ESI.

General procedure to hydrolysis: Compounds (1–5), each at approximately 0.12 mmol, were dissolved in 1 mL of 80 % acetic acid solution. At this concentration, the acetic acid is present in a ~100‐fold molar excess relative to the substrate. All reactions were conducted at 19.0±0.5 °C.

Glc‐Bz: 1,2;5,6‐Di‐O‐Isopropylidene‐3‐O‐benzoyl‐ α‐D‐glucofuranose

1,2 : 5,6‐Di‐O‐isopropylidene‐α‐D‐glucofuranose (2 g; 7.6 mmol) was dissolved in dry pyridine (100 mL), cooled at 0 °C and Benzoyl‐chloride (1.47 g; 7.6 mmol) was added to the solution. The suspension was stirred for 1 hour. Then the mixture was poured into HCl: H₂O/1 : 4 solution. The precipitated yellowish crystals were filtrated and dissolved with DCM. The water phase was extracted three times with chloroform. The precipitate was filtered and washed with cold water to get the yellowish‐white crystalline product (2.2 g; 81 %, purity 93 %). R_f=0.92 (EtOAc:Hexane/3 : 1); HRMS (ESI‐orbitrap) m/z: [M+Na]⁺ Calcd. for C19H24O7Na 387.1420; Found 387.1413 (−0.3 ppm). ¹H NMR (700 MHz, AcOD‐d₄/D₂O): δ=8.02 (d, 2H, J=7.7), 7.63 (t, 1H, J=7.63), 7.50 (t, 2H, J=7.63), 6.02 (d, 1H, J=3.43), 5.47 (d, 1H, J=2.45), 4.71 (d, 1H, J=3.5), 4.45 (m, 1H, J=6.51), 4.39 (d,d J=7.49, J=2.59), 4.14 (m, 1H, J=7.36), 4.10 (dd, 1H, J=8.54, J=5.11) 1.53, 1.41, 1.32, 1.27 (4 s, 12H, CH₃), ¹³C NMR (700 MHz, AcOD‐d4/D₂O): δ 25.7, 26.2, 26.7, 26.8 (4×iPr‐CH₃), 67.4 (C‐6), 73.5 (C‐5), 77.5 (C‐4), 80.6 (C‐3), 84.1 (C‐2), 106 (C‐1), 110.4, 113.3 (2×iPr‐C quart.), 129.5 (C‐p), 130.4 (C‐m), 134.5 (C‐o), 166.4 (C‐c) ppm. ^[25] Figure S11,12

Glc‐Tos: 1,2;5,6‐Di‐O‐Isopropylidene‐3‐O‐Toluenesulfonyl‐α‐D‐Glucofuranose

1,2 : 5,6‐Di‐O‐isopropylidene‐α‐D‐glucofuranose (6 g; 23 mmol) was dissolved in dry pyridine (30 mL) and p‐toluenesulfonyl chloride (8.9; 47 mmol) was added to the solution while stirring at room temperature. The suspension was stirred for 3 days. The reaction mixture was poured into ice‐water, the precipitate was filtered, and thoroughly washed with cold water. The white product was recrystallized from ethanol (30 mL, 95 %, purity 94 %) to give white crystalline product (7.5 g; 63 %); R_f=0.57 (EtOAc : Hexane 1 : 1). ESI‐MS: 415.2 [M+H]⁺, calc.:414.1 m/z.; ¹H NMR (700 MHz, AcOD‐d₄/D₂O): δ=7.88 (d, 2H, J=8.12), 7.43 (d, 2H, J=8.05), 5.76 (d, 1H, J=3.15), 4.66 (m, 2H, J=5.42), 4.14 (m, 2H, J=4.48), 3.9 (t, 1H, J=7.49), 2.44 (s, 3H) 1.49, 1.29, 1.27, 1.24 (4 s, 12H, CH₃); ¹³C NMR (700 MHz, AcOD‐ d₄/D₂O): δ 21.6 (Tos‐CH3), 24.8, 26.1, 26.4, 26.5 (4×iPr‐CH3), 67.1 (C‐6), 72.6 (C‐5), 80.4 (C‐4), 83 (C‐3), 84 (C‐2), 105.8 (C‐1), 110.1, 113.6 (2×iPr‐C quart.), 129.1 (C‐m), 130.9 (C‐o), 133 (C‐i), 146.8 (C‐p) ppm. ^[26] Figure S13,14.

All‐Tos: 1,2;5,6‐Di‐O‐Isopropylidene‐3‐O‐Toluenesulfonyl‐α‐D‐Allofuranose

From 1,2 : 5,6‐di‐O‐isopropylidene‐α‐D‐allofuranose (10.0 g; 38.0 mmol) the reaction was carried out with the same method as the formation of Glc‐Tos. After recrystallization of the crude product from 90 % ethanol, the product was obtained as white crystals. (6.3 g; 63 %, purity 91 %). Rf=0.64 (EtOAc : Hexane 1 : 1). ¹H NMR (700 MHz, AcOD‐ d₄/D₂O): δ=7.88 (d, 2H, J=8.04), 7.46 (d, 2H, J=7.98), 5.79 (d, 1H, J=3.13), 4.67 (m, 2H, J=6.78), 4.14 (m, 2H, J=5.2), 3.9 (t, 1H, J=7.41), 2.45 (s, 3H) 1.25,1.38, 1.29, 1.49 (4 s, 12H, CH₃); ¹³C NMR (700 MHz, AcOD‐ d₄/D₂O): δ 21.6 (Tos‐CH3), 24.9, 26.2, 26.5, 26.7 (4×iPr‐CH3), 65.6 (C‐6), 75.5 (C‐5), 77.8 (C‐4), 78 (C‐3), 78.8 (C‐2), 104.8 (C‐1), 110.7, 114.4 (2×iPr‐C quart.), 129.1 (C‐m), 130.8 (C‐o), 133.5 (C‐i), 146.8 (C‐p) ppm. ^[27] Figure S15,16.

Glc‐Imz: 1,2;5,6‐Di‐O‐Isopropylidene‐3‐O‐(N‐imidazole‐1‐Sufonyl)‐α‐D‐Glucofuranose

1,2 : 5,6‐Di‐O‐isopropylidene‐α‐D‐glucofuranose (5.85 g; 23.0 mmol) was dissolved in DMF (100 mL), cooled at 0 °C and under N₂‐atmosphere NaH (1.35 g; 57.0 mmol) was added to the solution. The suspension was stirred for 30 min and cooled to ~ −40 °C. After N′, N‐sulfonyl diimidazole (6.71 g; 33.5 mmol) in DMF (60 mL) was dropped to the reaction mixture and stirred for 30 min again at −40 °C. MeOH was added (0.8 mL) to the solution and stirred for 30 min at −40 °C. The solution was poured into ice‐water (200 mL). The precipitate was filtered and washed with cold water to get the white crystalline product (1.2 g; 80 %, purity 97 %). R_f=0.76 (EtOAc:Hexane/1 : 2); ESI‐MS: 391.1 [M+H]⁺, calc.: 390.4 m/z. ¹H NMR (700 MHz, AcOD‐ d₄/D₂O): δ=8.41 (s, 1H), 7.68 (m, 1H, J=3.01), 7.3194 (m, 1H, J=0.84), 6.09 (d, 1H, J=3.71), 5.09 (d, 1H, J=2.81), 4.90 (d, 1H, J=3.73), 4.15 (dd, 1H, J=8.24, J=2.68) 4.09 (m, 2H), 3.94 (m, 1H) 1.50, 1.35, 1.26, 1.20 (4 s, 12H, CH₃), ¹³C NMR (700 MHz, AcOD‐ d₄/D₂O): δ 24.6, 26.1, 26.4, 26.6 (4×iPr‐CH3), 67.5 (C‐6), 72.5 (C‐5), 80.3 (C‐4), 83.5 (C‐3), 87.6 (C‐2), 106 (C‐1), 110.8, 114.2 (2×iPr‐Cquart), 120.3, 130.1, 138.5 (3×C‐imz) ppm. (During the synthesis of Compound 4, the side product 3,3’‐O‐bis(1,2;5,6‐Di‐O‐isopropylidene‐α‐D‐glucofuranose)‐sulfate (Compound 6) was formed. It was isolated by column chromatography and its structure verified by X‐ray crystallography. ^[14] Figure S17,18.

All‐Imz: 1,2;5,6‐Di‐O‐Isopropylidene‐3‐O‐(N‐imidazole‐1‐sufonyl)‐α‐D‐Allofuranose

1,2 : 5,6‐Di‐O‐isopropylidene‐α‐D‐allofuranose (0.48 g, 1.8 mmol) was dissolved in DMF (9 mL), cooled at 0 °C and under N2‐atmosphere NaH (0.06 g; 2.7 mmol) was added to the solution. The suspension was stirred for 30 min and cooled to ~−40 °C. After N, N′‐sulfonyldiimidazole (0.55 g; 2.7 mmol) in DMF (6 mL) was dropped to the reaction mixture and stirred for 30 min again at −40 °C. MeOH was added (0.4 mL) to the solution and stirred for 30 min at −40 °C. The solution was poured into ice‐water (50 mL). The water was extracted with Et₂O (3×50 mL). The organic phase was collected and dried on MgSO₄. Then the solvent evaporated, and white crystals remained. (1.2 g; 80 %, purity 99 %). R_f=0.61 (EtOAc); ESI‐MS: 391.2 [M+H]⁺, calc.:390.4 m/z. ¹H NMR (700 MHz, AcOD‐ d₄/D₂O): δ=8.34 (s, 1H), 7.59 (s, 1H,) 7.26 (s, 1H), 5.80 (d, 1H, J=3.64), 4.93 (q, 1H, J=4.22), 4.67 (t, 1H, J=4.23), 4.13 (q, 1H, J=5.83), 4.1 (t, 1H, J=6.93), 4.05 (t, 1H, J=7.66), 3.83 (q, 1H, J=4.62), 1.48, 1.35, 1.26, 1.26 (4 s, 12H, CH₃), ¹³C NMR (700 MHz, AcOD‐ d₄/D₂O): δ 27, 28.6, 28.6, 28,7 (4×iPr‐CH3), 68.7 (C‐6), 77.9 (C‐5), 80.3 (C‐4), 80.4 (C‐3), 84.7 (C‐2), 107.4 (C‐1), 113.3, 117.2 (2×iPr‐Cquart), 122.3, 132.4, 138.5 (3×C‐imz) ppm. ^[14] Figure S19,20.

Glc2‐S: 3,3’‐O‐bis (1,2;5,6‐Di‐O‐Isopropylidene α‐D‐Glucofuranose)‐Sulfate

HRMS (ESI‐orbitrap) m/z: [M+Na]⁺ Calcd. for C24H38O14SNa 605.1880; Found 605.1869 (−0.9 ppm).

Crystallographic Studies on Molecule 4 (Glc‐Imz) and 6 (Glc2‐S)

Transparent prism shaped crystals of compound Glc‐Imz and Glc2‐S were obtained from EtOAc/Hexane: 3/2 solution. X‐ray diffraction data from a single crystal were collected at room temperature on a microfocus sealed tube SuperNova diffractometer (Agilent Technologies, Inc.) equipped with an Eos CCD detector using monochromated copper K_α radiation (λ=1.5418 Å). Data reduction was carried out using the CrysAlisPro 1.171.42.58a software. The structures of Glc‐Imz and Glc2‐S were solved by intrinsic methods using SHELXT, ^[28] using the Olex2 v1.3.0 respectively. The structures were refined by full‐matrix least‐squares techniques (SHELXL‐2018/3) on F ² . Hydrogen atoms were generated upon geometric evidence and were refined in the riding positions. The imidazole group shows orientational disorder, indicated by irregular atomic displacement parameters and bond lengths upon refinement in either orientation. It was refined as a mixture of the two orientations with constrained atomic displacement parameters for the alternate atom‐pairs and rigid bond restraints applied. Olex2 and Mercury were used for molecular graphics and structure analysis. Crystallographic data and selected geometric parameters are compiled in Supplementary Tables 2–3, the structures and crystal packing are shown in Supplementary Figures 2, 3, 4, 5. Validation was carried out using CheckCIF/PLATON. Structures of Glc‐Imz and Glc2‐S were deposited with the Cambridge Crystallographic Data Centre. Deposition Numbers 2335760 (for Glc‐Imz) and 2335762 (for Glc2‐S) contain the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service.

Computational Studies for Molecules of Glc‐Imz and All‐Imz

We utilized the CREST (Conformer‐Rotamer Ensemble Sampling Tool) program to identify low‐energy conformers of the studied molecules. For Glc‐Imz, 278 conformers were initially found, but we focused on the 128 conformers with energies within 3 kcal/mol of the lowest energy. From this refined set, we clustered the five main conformers. Similarly, for All‐Imz, 209 conformers were identified, and we concentrated on the 118 conformers within the 3 kcal/mol energy range, clustering the three main conformers.

For the protonated Glc‐Imz⁺, All‐Imz⁺1, and All‐Imz⁺2, we selected the lowest energy conformer, which accounted for 47 %, 31 %, and 61 % of the population, respectively. For All‐Imz, we chose to optimize the two most populated conformers: All‐Imz1, where the interaction is likely with the 5,6‐dioxolane ring, and All‐Imz2, where interaction is more likely with the 1,2‐dioxolane ring. These conformers underwent further protonation at the imidazole ring and were subsequently optimized using the Gaussian 16 software at the B3LYP/6‐311++G (d, p) level of theory, as shown in Figure 11.

Figure 11.

Low‐energy conformers of Glc‐Imz and All‐Imz were identified using CREST, focusing on those within 3 kcal/mol of the lowest energy. The main conformers were clustered, and the lowest energy conformer of protonated Glc‐Imz⁺ (47 %), All‐Imz⁺1 (31 %) and All‐Imz⁺2 (61 %) was optimized using Gaussian 16.

We conducted a similar conformational search for the Glc‐Tos molecule; however, the identified conformers were not subjected to further optimization.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Szaniszló S., Csizmadia I. G., Jákli I., Farkas V., Láng A., Sulyok-Eiler M., Harmat V., Pintér I., Perczel A., Chem. Eur. J. 2025, 31, e202403319. 10.1002/chem.202403319

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.