Diastereomeric Configuration Modulates Liquid–Liquid Phase Separation and Catalysis in Minimalist Dipeptide Coacervates

ABSTRACT

Although coacervates formed via liquid–liquid phase separation (LLPS) of small molecules are widely recognized as plausible protocell models relevant to the origin of life, how the stereochemistry of small molecules influences LLPS remains largely unexplored. Here we report a set of minimalist dipeptide stereoisomers composed of l ‐proline ( ^L P) or d ‐proline ( ^D P), and l ‐naphthylalanine ( ^L Nal), or d ‐naphthylalanine ( ^D Nal), in which chirality modulates the propensity for LLPS under identical aqueous conditions. Specifically, ^L P ^D Nal and ^D P ^L Nal independently undergo LLPS to form coacervates that selectively accumulate diverse guest molecules and act as efficient crucibles, markedly accelerating stereoselective reactions. By contrast, the remaining stereoisomers, ^L P ^L Nal and ^D P ^D Nal, preferentially access a competing crystallization pathway. Single‐crystal x‐ray diffraction and all‐atom molecular dynamics simulations reveal that this divergence originates from stereochemistry‐dependent variations in intermolecular hydrogen‐bonding patterns and aromatic stacking. Together, these findings establish diastereomeric configuration‐modulated LLPS in a minimalist molecular system and demonstrate how molecular stereochemistry can directly regulate protocell‐like compartmentalization.

EnA set of stereoisomeric dipeptides composed of proline (P) and naphthylalanine (Nal) was designed to demonstrate that molecular chirality can deterministically modulate liquid–liquid phase separation (LLPS) in a minimalist system. Two stereoisomers ( ^LP^DNal and ^DP^LNal) undergo LLPS to form dynamic coacervates, while the remaining stereoisomers ( ^LP^LNal and ^DP^DNal) preferentially access a crystallize pathway under identical aqueous conditions.

1. Introduction

Life is inherently chiral: from amino acids and nucleic acids to proteins, polysaccharides, and even cells, biological systems exhibit well‐defined stereochemical organization [1, 2, 3, 4]. Despite the central importance of chirality to living systems, the physicochemical origins of living systems’ asymmetry remain a profound open question [5]. Addressing this challenge requires identifying mechanisms by which molecular chirality can influence higher‐order organization and the self‐assembly of cell‐like compartments (protocells) under prebiotic conditions [6, 7, 8, 9].

Among the diverse protocells explored to data—including liposomes [10, 11], fatty‐acid and polymer vesicles [12, 13, 14], colloidosomes [15], and oil‐in‐water droplets [16]—coacervates formed by liquid–liquid phase separation (LLPS) have emerged as particularly compelling models [17, 18, 19]. LLPS generates membrane‐free, molecularly crowded compartments capable of selectively concentrating guest molecules and enhancing chemical reactivity, properties that closely parallel key functions of living cells [20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35]. Recently, simple coacervates assembled from single small‐molecule components have attracted growing interest due to their structural minimalism and prebiotic plausibility [8, 9, 36, 37, 38, 39]. For instance, Spruijt and colleagues designed tetrapeptide derivatives that undergo LLPS to form coacervates [36], while similar oligopeptide derivatives have been developed through the rational tuning of intermolecular interactions governing LLPS [40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51]. In recent breakthroughs, Landfester, Silva and coworkers reported a dipeptide‐based phase‐separating system capable of forming coacervate in water [37]. Yan, Yuan, and colleagues developed a library of simple amino acid derivatives that undergo LLPS under prebiotic conditions [39]. In parallel, non‐peptide synthetic small molecules have also been explored as LLPS‐forming building blocks, including pyrene‐, naphthalene‐, and benzene‐derived systems, thereby expanding the molecular diversity of small‐molecule coacervates beyond peptides [52, 53, 54, 55]. Despite these advances, the extent to which molecular chirality can directly regulate LLPS in such minimal systems has received surprisingly little attention. In this context, Perry et al. and Zhou et al. have reported chirality‐selected coacervation arising from interactions between oppositely charged species [6, 7]. In a recent study, Carlone, Daidone, Iacobucci and coworkers reported a tripeptide system exhibiting enantiopurity‐dependent LLPS, in which the racemic mixture formed a solid precipitate rather than coacervates [9]. However, whether stereochemistry can deterministically modulate the emergence of coacervates in single‐molecular‐component systems remains unknown.

Here we demonstrate that molecular chirality alone can decisively govern LLPS in a minimal dipeptide system relevant to protocell emergence. We designed four stereoisomeric dipeptides by coupling l ‐proline ( ^L P) or d ‐proline ( ^D P) with l ‐naphthylalanine ( ^L Nal) or d ‐naphthylalanine ( ^D Nal). Under identical aqueous conditions, ^L P ^D Nal and ^D P ^L Nal spontaneously undergo LLPS to form stable coacervates, whereas their diastereomers, ^L P ^L Nal or ^D P ^D Nal, preferentially access crystallize pathways. Importantly, only the LLPS‐active stereoisomers generate dynamic, molecule‐enriching compartments that function as efficient crucibles, markedly accelerating stereoselective reactions. Through a combination of single‐crystal x‐ray diffraction and molecular simulations, we show that this chirality‐modulated LLPS behavior arises from subtle differences in intermolecular hydrogen‐bonding patterns and aromatic stacking. Collectively, these findings establish a minimalist molecular framework in which chirality directly modulates LLPS and protocell‐like compartmentalization, offering new insight into how molecular stereochemistry could have shape early prebiotic organization.

2. Results and Discussion

To examine whether molecular chirality alone can act as a decisive control parameter for LLPS, we designed four stereoisomeric dipeptide derivatives by coupling either l ‐proline ( ^L P) or d ‐proline ( ^D P) with l ‐naphthylalanine ( ^L Nal) or d ‐naphthylalanine ( ^D Nal), followed by methyl esterification at the C‐termini. The obtained stereoisomers were denoted as ^L P ^L Nal, ^L P ^D Nal, ^D P ^L Nal, and ^D P ^D Nal, according to the chirality of their amino acid building blocks (Figure 1). These small molecules were dispersed in identical aqueous solution at pH = 7.4 to examine the influence of stereochemistry on their self‐assembly and phase‐separation behavior. Using ^L P ^D Nal as a representative example, introducing 30 mM ^L P ^D Nal into aqueous solution produced a milky suspension containing discrete and spherical microdroplets (Figure 1a). The microdroplets typically measured 5.9 ± 2.3 µm in diameter. Microscopic analysis revealed a clear concentration dependence of droplet size, increasing from 2.6 ± 0.8 µm at 10 mM to 10.6 ± 4.0 µm at 50 mM (Figure S1). As the enantiomer of ^L P ^D Nal, the ^D P ^L Nal displayed essentially identical phase‐separation behavior under the same conditions (Figure 1b). In contrast, introducing ^L P ^L Nal or ^D P ^D Nal (30 mM) into the same aqueous medium immediately yielded white precipitates that rapidly sedimented within minutes (Figure 1c,d). Microscopic examination revealed that these precipitates consisted of transparent, plate‐like crystalline structures, clearly distinct from the liquid, dynamic microdroplets formed by ^L P ^D Nal and ^D P ^L Nal.

FIGURE 1.

(a)–(d) Chemical structures of ^L P ^D Nal, ^D P ^L Nal, ^L P ^L Nal, and ^D P ^D Nal (top panel); optical images of molecular self‐assembly in aqueous solution (pH = 7.4, 3 M NaCl, middle panel); microscopy images showing the resulting self‐assembled structures (bottom panel). Scale bars are 50 µm.

Both ^L P ^D Nal and ^D P ^L Nal spontaneously underwent liquid–liquid phase separation (LLPS) in aqueous solution, yielding micrometer‐sized spherical droplets (Figure 2a). To determine whether these microdroplets are coacervates, we examined their physicochemical properties in detail. The ^L P ^D Nal microdroplets readily fused upon contact and rapidly relaxed into spherical shapes within seconds (Figure 2b), confirming their liquid‐like nature rather than a solid‐like hydrogel, despite the common propensity of oligopeptides to form fibers or hydrogels in water [56, 57, 58, 59]. The liquid‐like behavior was also confirmed by fluorescence recovery after photobleaching (FRAP) experiments (Figure S2). To further distinguish these microdroplets from water‐immiscible oil droplets, the molecular compositions of both the microdroplets and the surrounding medium were quantified by using NMR spectroscopy and UV–vis method. In both ^L P ^D Nal and ^D P ^L Nal systems, the surrounding medium contained more than 99 mol% water, whereas the microdroplets themselves contained approximately 90 mol% water (Figures 2c and S3, and S4). The predominance of water in both the microdroplets and surrounding phase confirms that the resulting microdroplets are coacervates. Notably, once formed, the coacervates remained stable for at least one week of observation (Figure S5), without undergoing liquid‐to‐solid transition, in contrast to the aging behavior commonly observed in peptide‐based LLPS systems.

FIGURE 2.

(a) Confocal laser scanning microscopy (CLSM) images of ^L P ^D Nal coacervates (top panel) and ^D P ^L Nal coacervates (bottom panel). Rhodamine 6G was used as fluorescent probe. Scale bars are 10 µm. (b) Time‐series microscopy images showing the fusion of ^L P ^D Nal coacervates. Scale bars are 5 µm. (c) Plots showing the molecular fraction of the supernatant phase (Super.) and coacervate phase (Coacer.). (d) Optical images illustrating the pH‐response behavior of ^L P ^D Nal coacervates. (e) Turbidity measurements showing the pH‐triggered phase transition of ^L P ^D Nal solutions, with a homogeneous solution at pH < 6.8 and LLPS occurring at pH ≥ 6.8. (f) XRD spectra of ^D P ^D Nal (blue line) and ^L P ^L Nal (red line) precipitates.

In typical coacervate systems, LLPS arises from a delicate balance between intermolecular interactions and solvation, subtle perturbations to this balance can profoundly affect phase behavior [38]. Both ^L P ^D Nal and ^D P ^L Nal exhibit a pK _a value of 8.8 (Figure S6), and their phase behavior is strongly pH‐dependent. ^L P ^D Nal and ^D P ^L Nal spontaneously undergo LLPS at pH ≥ 6.8, where intermolecular interactions are sufficiently strong to counterbalance solvation by water (Figures 2d and S7). Under acidic conditions (pH < 6.8), protonation of the pyrrolidine group enhances solvation and suppressed LLPS (Figure 2d). In contrast, ^L P ^L Nal and ^D P ^D Nal, which share the same pK _a value of 8.8 (Figure S6), display markedly different pH‐responsive behavior. These stereoisomers form white precipitates over a broad pH range from 2.0 to 8.0 (Figure S8). Even under acidic conditions, protonation of the pyrrolidine group is insufficient to overcome their strong intermolecular interactions, resulting in persistent precipitation rather than LLPS or dissolution.

In addition to pH, ionic strength markedly affects self‐assembly and phase separation. Increasing the NaCl concentration promoted LLPS in the ^L P ^D Nal and ^D P ^L Nal systems, as evidenced by enlarged droplet sizes at higher salt concentrations (Figure S9). In contrast, ^L P ^L Nal and ^D P ^D Nal formed solid plate‐like precipitates at NaCl concentration above 1.0 M, but underwent LLPS when the NaCl concentration was reduced below 1.0 M (Figures S10 and S11). This trend is consistent with the classical salting‐out effect, whereby elevated ionic strength enhances intermolecular interactions and suppresses solvation, thereby favoring phase separation or precipitation.

Unlike ^L P ^D Nal and ^D P ^L Nal, which form coacervate microdroplets by LLPS, ^L P ^L Nal and ^D P ^D Nal precipitate as solid plate under identical experimental conditions. These precipitates were characterized by using powder x‐ray diffraction (PXRD) at room temperature (Figure 2f). The PXRD patterns of ^L P ^L Nal and ^D P ^D Nal were identical and indicative of highly ordered molecular packing, confirming the crystalline nature of the ^D P ^D Nal and ^L P ^L Nal assemblies. Although pure ^D P ^D Nal forms crystals and pure ^D P ^L Nal forms coacervates, a 1:1 mixture of ^D P ^L Nal and ^D P ^L Nal yields only coacervates rather than a coexistence of coacervates and crystals. This behavior is primarily attributed to the disruption of the ordered self‐assembly of ^D P ^D Nal by ^D P ^L Nal (Figure S12).

Initial observation revealed that enantiomeric pairs ( ^L P ^D Nal and ^D P ^L Nal, as well as ^L P ^L Nal and ^D P ^D Nal) exhibit identical self‐assembly and phase‐separation behavior, whereas diastereoisomers (e.g., ^L P ^D Nal vs. ^L P ^L Nal) display markedly different outcomes. We therefore sought to elucidate the molecular mechanisms underlying these distinct self‐assembly behaviors. For ^L P ^D Nal and ^D P ^L Nal, LLPS arises from a delicate balance between intermolecular interactions and molecular solvation. To understand why only specific stereoisomers access a LLPS state, we analyzed how stereochemistry modulates intermolecular interaction patterns at the molecular level. We grew single crystals of ^D P ^D Nal and ^L P ^L Nal and analyzed their molecular arrangements using single‐crystal x‐ray diffraction (SC‐XRD) [60]. For the ^D P ^D Nal single crystal, the refined lattice parameters were determined to be a = 5.2549(3) Å, b = 9.5775(5) Å, c = 18.1849(8) Å. The detailed molecular arrangements of ^D P ^D Nal molecules is shown in Figure 3a, and is characterized by extensive intermolecular hydrogen bonding (Figure 3b; cyan lines) and aromatic herringbone stacking interactions between naphthyl groups (Figure 3c; cyan lines). Specifically, multiple hydrogen bonds are formed between the amide carbonyl oxygen and methylene hydrogen atoms of the pyrrolidine ring, between the ester carbonyl oxygen and the amide hydrogen, and between the amide carbonyl oxygen and naphthyl hydrogen donors. The herringbone aromatic interactions exhibit an interplanar angle of 78.4° between adjacent naphthyl rings, with C(π)···C(π) distances of approximately 5.415 Å and 5.150 Å, respectively (Figure 3c). Notably, face‐to‐face π‐π stacking between naphthyl rings is absent, as the corresponding centroid‐to‐centroid distances range from 10.924 Å to 14.219 Å, far exceeding typical π–π interaction distances. These structural features indicate that multiple hydrogen bonds and herringbone aromatic stacking dominate the self‐assembly of ^D P ^D Nal in the solid state. A similar molecular organization and intermolecular interaction pattern were also observed in the ^L P ^L Nal single crystal (Figure 3d,f). In both cases, these strong and highly ordered intermolecular interactions overwhelm molecular solvation, thereby favoring crystallization and precipitation over LLPS.

FIGURE 3.

(a) Molecular organization of ^D P ^D Nal single crystal. The unit cell is highlighted by a blue dashed box. The refined lattice parameters are a = 5.2549(3) Å, b = 9.5775(5) Å, c = 18.1849(8) Å. (b) Intermolecular hydrogen‐bonding network (cyan lines, I to IV) in the ^D P ^D Nal crystal. (c) The herringbone‐type aromatic arrangements of the naphthyl rings (V and VI) in the ^D P ^D Nal crystal. (d) Molecular organization of ^L P ^L Nal single crystal, with the unit cell highlighted by a blue dashed box. The refined lattice parameters are a = 5.2565(2) Å, b = 9.5924(5) Å, c = 18.2010(8) Å. (e) Intermolecular hydrogen‐bonding network (cyan lines, I to IV) in the ^L P ^L Nal crystal. (f) The herringbone‐type aromatic arrangements of the naphthyl rings (V and VI) in the ^L P ^L Nal crystal.

To further elucidate the intermolecular interactions underlying the stereochemistry‐dependent phase behavior, we performed conformational search for dimers of ^D P ^L Nal, ^D P ^D Nal, ^L P ^D Nal, and ^L P ^L Nal using the Molclus program (Supporting Information). These calculations revealed that the symmetric dimers of ^D P ^D Nal and ^L P ^L Nal can form up to four hydrogen bonds, whereas the corresponding ^L P ^D Nal and ^D P ^L Nal dimers support a maximum of three hydrogen bonds (Figure 4a). For example, in the ^D P ^D Nal dimer, a highly symmetric hydrogen‐bonding pattern is observed, involving interactions between the amide carbonyl oxygen and the hydrogen atom on the pyrrolidine ring, as well as between the ester methoxy oxygen and the amide hydrogen atom. In contrast, replacing the ^D Nal motif with ^L Nal disrupts this symmetry, resulting in a distinct hydrogen‐bonding pattern that includes interactions between the amide carbonyl oxygen and the hydrogen atoms of the pyrrolidine ring, the amide carbonyl oxygen and amide hydrogen, amide nitrogen and amide hydrogen. This loss of symmetry reduces the overall number and cooperativity of hydrogen bonds in the ^L P ^D Nal and ^D P ^L Nal dimers. Simulation results indicate that hydrogen bonding plays an important role in molecular self‐assembly. This is further supported by the observation that ^L P ^D Nal coacervates are disrupted upon the addition of urea (Figure S13). Subsequently, we quantified and compared the average number of hydrogen bonds between dipeptide–dipeptide and dipeptide‐water using the Gromacs package with general amber force field. All‐atom molecular dynamics (AAMD) simulations were conducted in a 10 nm cubic box containing 120 dipeptides molecules (200 mM). The simulations were performed at pH 7.4, 8.1, and 8.5, respectively. The results show that ^D P ^D Nal and ^L P ^L Nal systems—corresponding to the crystals—form more intermolecular hydrogen bonds than the ^L P ^D Nal and ^D P ^L Nal systems (Figures 4b and S14). In contrast, the number of hydrogen bonds between dipeptides and water molecules exhibits only minimal differences across all four systems (Figures 4c and S14). In addition, we analyzed π–π stacking interactions between the naphthyl groups in each system. As expected, the extent of π–π stacking is substantially higher in the ^D P ^D Nal and ^L P ^L Nal systems than in the ^L P ^D Nal and ^D P ^L Nal coacervate‐forming systems (Figures 4d and S15–S17), consistent with the notion that stronger and more persistent intermolecular interactions favor crystallization. The combined presence of increased hydrogen bonding and enhanced π–π stacking enables ^D P ^D Nal and ^L P ^L Nal to pack more tightly, resulting in reduced solvent‐accessible surface areas. By contrast, ^L P ^D Nal and ^D P ^L Nal form less compact assemblies with increased surface exposure (Figure S18), consistent with their propensity to undergo LLPS rather than crystallization. These simulations provide direct molecular‐level support for our experimental observations, confirming that chirality‐dependent hydrogen bonding and aromatic stacking interactions decisively determine whether a stereoisomer undergoes LLPS or crystallization.

FIGURE 4.

(a) Representative hydrogen‐bonding pattern in dimers of ^D P ^L Nal, ^D P ^D Nal, ^L P ^D Nal, and ^L P ^L Nal obtained from conformational searches. Hydrogen bonds are highlighted by cyan lines. Selected hydrogen atoms are omitted for clarity. Average number of intermolecular hydrogen bonds between dipeptides (b) and between dipeptides and water (c) obtained from all‐atom molecular dynamics (AAMD) simulations. (d) Average π–π stacking interactions between naphthyl groups in each system. These simulations were performed at pH 7.4 condition.

Having elucidated the stereochemistry‐dependent origin of LLPS versus crystallization, we next compared the physicochemical properties of the ^L P ^D Nal and ^D P ^L Nal coacervates with ^L P ^L Nal and ^D P ^D Nal crystals. Coacervates are widely regarded as plausible protocell models due to their special ability to selectively recruit diverse client molecules and generate concentration gradients between the condensed phase and the surrounding environment without employing a physical membrane. Accordingly, we evaluated the selective recruitment capabilities of ^L P ^D Nal and ^D P ^L Nal coacervates. CLSM images revealed that both ^L P ^D Nal and ^D P ^L Nal coacervates efficiently sequester a range of small fluorescent molecules, including negatively‐charged Fluorescein, positively‐charged Thioflavine T, and water‐insoluble Nile Red (Figures 5a–c and S19). These small fluorophores exhibited high partition coefficients ( K ) of 110, 140, and 3800, respectively. Due to the intrinsically hydrophobic nature of coacervates, hydrophobic Nile Red exhibits extremely high K value, which is consistent with previous reports [37]. In contrast, these fluorophores showed negligible partitioning into ^L P ^L Nal and ^D P ^D Nal crystals, as confirmed by CLSM images (Figures S20 and S21). Beyond small molecules, ^L P ^D Nal coacervates were also capable of accumulating various macromolecular guests. For example, poly(diallyldimethylammonium chloride) (PDDA), and dextran with molecular weights of 10 kDa were efficiently taken up into the coacervates (Figures 5d,e and S19), with K values of 190 and 50, respectively. These results indicate that the dipeptide coacervates have a higher accumulation affinity toward polyelectrolytes than toward neutral polymers. The lower K value of small‐molecular coacervates to dextran was attributed to their highly hydrophilic properties, which is consistent with a series of analogous systems reported recently [38, 54]. Similarly, fluorescently labelled proteins, including Glucose Oxidase (GOx), Lipase and Bovine Serum Albumin (BSA) were all enriched within ^L P ^D Nal coacervates, with K values of 160, 50, and 100, as confirmed by CLSM images and gel electrophoresis (Figures 5f–h and S22). Comparable uptake behavior was observed for ^D P ^L Nal coacervates (Figure S19). In addition, ^L P ^D Nal coacervates exhibited a high capacity for the sequestration of deoxyribonucleic acid (DNA), which was attributed to the electrostatic forces between positively charged ^D P ^L Nal molecules and negatively charged DNA [36]. In stark contrast, CLSM images demonstrated that ^L P ^L Nal and ^D P ^D Nal crystals possess minimal affinity for the same polymers, proteins and nucleic acids (Figures S20, S21, and S23). This divergent accumulation property is also available for nanoparticles, as confirmed by the CLSM image showing in Figures 5j and S19. Collectively, these findings demonstrate that ^L P ^D Nal and ^D P ^L Nal coacervates possess a strong and versatile affinity for a broad spectrum of guest molecules, underscoring their significant potential as protocell models capable of performing cell‐like functions. By contrast, their diastereomeric counterparts, ^L P ^L Nal and ^D P ^D Nal, lack these properties. These results provide direct experimental evidence that chirality can modulate self‐assembly and LLPS behavior, thereby influencing the emergence of coacervate protocells.

FIGURE 5.

CLSM images showing ^L P ^D Nal coacervates incubated with Fluorescein (a), Thioflavine T (b), Nile Red (c), FITC‐labelled Dextran (d), RITC‐labelled PDDA (e), FITC‐labelled glucose oxidase (f), RITC‐labelled lipase (g), RITC‐labelled BSA (h), FAM‐labelled DNA (i), Fluorophore‐labelled polystyrene (PS) nanoparticles (j). Scale bars are 10 µm.

The stereochemistry‐dependent self‐assembly and phase‐separating behavior of these molecules gives rise to fundamentally distinct physicochemical properties. Specifically, ^D P ^L Nal and ^L P ^D Nal spontaneously undergo LLPS to form coacervates, which create molecularly‐crowded and highly dynamic compartments that function as efficient reaction crucibles. Importantly, these coacervates are capable of selectively accumulating a wide range of guest molecules, as demonstrated in Figure 5. In contrast, ^D P ^D Nal and ^L P ^L Nal assemble into highly ordered crystalline solids that effectively excludes client molecules. These pronounced physicochemical differences translate directly into divergent reactivity in chemical transformations. To demonstrate this effect, we selected classical aldol reactions between cyclohexanone and benzaldehyde as a model system and performed the reaction in the presence of each stereoisomer (Figure 6a). All reactions were conducted at pH 7.4 and room temperature for 6 h. Product yields and diastereoselectivity were quantitatively determined by NMR spectroscopy and high performance liquid chromatography (HPLC). As shown in Figure 6b, in the presence of ^D P ^L Nal (30 mM), the reaction reached up to 95% conversion within 6 h, with high diastereoselectivity (88:12, anti:syn) and an enantiomeric excess (ee) of 80.7%. These values are comparable to those reported in a recent study using a proline‐containing coacervate system for aldol catalysis, which achieved 79% yield and a 78% ee [8]. A comparable conversion and diastereoselectivity were observed in the ^L P ^D Nal coacervates‐catalyzed system. By contrast, reactions carried out in the presence of ^D P ^D Nal and ^L P ^L Nal crystals exhibited markedly lower product conversions under otherwise identical conditions, despite only minimal changes in diastereoselectivity (Figures 6c and S24). The stark contrast in reaction yields underscores the critical role of coacervate formation in accelerating aldol reactions. The enhanced reaction rates are attributed primarily to the improved solubility and elevated local concentrations of both cyclohexanone and benzaldehyde within the ^D P ^L Nal and ^L P ^D Nal coacervates (Figure S25). To further evaluate the contribution of the molecularly crowded coacervate environment, we performed the same aldol reaction using free l ‐Proline ( ^L P), l ‐prolinamide ( ^L PNH₂ ), and l ‐proline methyl ester ( ^L POMe) as the catalyst under identical conditions. Notably, negligible amount of product was detected (Figures 6c,d and S26–S28). This unexpected outcome can be attributed to the poor solubility of the substrates and the higher activation energy barrier in aqueous solution in the absence of a co‐solvent (for example, DMSO) [61, 62], highlighting that enhanced reactivity arises not from intrinsic catalytic superiority, but from the compartmentalized, solubilizing environment created by the ^D P ^L Nal and ^L P ^D Nal coacervates. Together, these results further confirm the essential role of coacervate environments in promoting chemical reactions.

FIGURE 6.

(a) Schematic illustration of the aldol reaction between cyclohexanone and benzaldehyde. (b) ¹H NMR spectra of products obtained in the presence of ^D P ^L Nal coacervates, ^L P ^D Nal coacervates, ^D P ^D Nal crystals, ^L P ^L Nal crystals, and ^L P ( l ‐Proline). Plots showing the 6 h kinetics (c) and yields (d) of aldol reaction. (e) Product yield for aldol reactions between cyclohexanone and various benzaldehyde derivatives. Purple values indicate yields obtained in the presence of ^D P ^L Nal coacervates, while blue values indicate yields obtained in the presence of ^D P ^D Nal crystals. Error bars in (d) represent standard deviations (n = 3).

Encouraged by these findings, we then examined the generality of the coacervate‐accelerated aldol reaction using a broader range of aldehyde substrates. A total of eight substrates were investigated, including benzaldehyde, 4‐bromobenzaldehyde, 4‐nitrobenzaldehyde, p‐tolualdehyde, 4‐biphenylcarbaldehyde, 2‐naphthaldehyde, 4‐formylbenzoate, and 1‐pyrenecarboxaldehyde. In all cases, reactions conducted in the ^D P ^L Nal coacervates exhibited significantly higher yields than those performed in the corresponding ^D P ^D Nal crystalline system (Figures 6e and S29–S35). For instance, a 93% yield was obtained for 4‐bromobenzaldehyde in the presence of ^D P ^L Nal, whereas the same reaction afforded only a 32% yield in the ^D P ^D Nal system. Specifically, for substrates bearing electron‐withdrawing groups such as nitro, ester or bromo, the reactions afforded yields above 90%, except for 4‐nitrobenzaldehyde, which gave a 76% yield due to its poor solubility. When p‐tolualdehyde, 4‐biphenylcarbaldehyde, 2‐naphthaldehyde and 1‐pyrenecarboxaldehyde were used as substrates, the electron‐donating effect increased the electron density of the aldehyde group, thereby reducing its electrophilicity and lowering the reactivity. As a result, the yields ranged from 58% to 64%, with the exception of 1‐pyrenecarboxaldehyde, whose bulky steric hindrance further reduced the yield to 46%. These results clearly demonstrate that molecular stereochemistry can modulate chemical reactivity indirectly by dictating self‐assembly pathways and LLPS behavior.

This stereochemistry‐dependent divergence in phase behavior and catalytic performance provides a compelling link between molecular chirality, compartmentalization, and reaction efficiency. By showing that subtle differences in stereochemistry can determine whether a system forms reactive, protocell‐like coacervates or inert crystalline solids, this work highlights a plausible physicochemical mechanism by which chirality could have influenced the emergence of functional chemical compartments on the prebiotic Earth.

3. Conclusion

In summary, this work demonstrates that molecular chirality can function as a decisive physicochemical parameter modulating liquid–liquid phase separation in a minimalist small‐molecule system. By systematically varying stereochemistry within a set of closely related dipeptide derivatives, we show that access to a coacervate state is not merely modulated, but deterministically permitted or prohibited by stereochemical configuration. This finding establishes chirality as a direct regulator of LLPS, extending its influence beyond molecular recognition and asymmetric catalysis to the level of micro‐scale organization. Mechanistically, our combined structural and computational analyses reveal that stereochemistry‐dependent molecular interactions decisively govern the phase‐separating behavior. Importantly, only those stereoisomers capable of sustaining fluid, exchangeable intermolecular contacts give rise to stable coacervate compartments. These LLPS‐derived assemblies exhibit hallmark protocell functions, including selective molecular accumulation and enhanced reaction kinetics, underscoring the functional consequences of chirality‐dependent phase behavior.

Biomolecules such as polysaccharides, nucleic acids, and proteins are predominantly homochiral. In this work, we demonstrate that heterochiral molecules preferentially form coacervate‐based protocells, whereas their homochiral counterparts tend to crystallize. Our aim is not to explain the origin of biomolecular chirality, but rather to show that, in a prebiotic context, the chirality of small molecules can direct the emergence of protocell‐like compartments. In this regard, the present study establishes chirality‐modulated liquid–liquid phase separation in a minimalist molecular system and demonstrates how the stereochemistry of small molecules can directly regulate coacervate compartmentalization. In a very rencet prepeint [63], Dhiman and co‐authors demonstrated the chirality‐dependent coacervation of tripeptide derivatives and how the resulting coacervates exhibit different accumulation behavior toward a chiral dye (L,L‐Ala‐Ala‐Cy5) and an achiral dye (Thioflavin T). These two work together provide new insight into how molecular stereochemistry may have become intertwined with compartmentalization during the early stages of chemical evolution. We anticipate that chirality‐controlled LLPS represents a general yet underexplored principle for designing synthetic protocells and for understanding the origins of biological organization.

Author Contributions

Shuai Peng: methodology, software, formal analysis, data curation, visualization, investigation. Xiaokun Zhang: methodology, software. Xin‐Li Shi: methodology. Meng Yu: formal analysis, supervision, funding acquisition, writing – review and editing. Shoupeng Cao: formal analysis, supervision, funding acquisition, writing – review and editing. Ning Gao: funding acquisition, writing – original draft, writing – review and editing, conceptualization, formal analysis, visualization, validation, project administration, resources, supervision.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supportng File 1: anie72824‐sup‐0001‐SuppMat.docx.

Supportng File 2: anie72824‐sup‐0001‐DataFile.zip.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.