Suppression of Aggregation Enhances Placement Fidelity of Asymmetrical Right Triangle DNA Origami Authors: Suqing Zhou, Yue Wang, Die He, Nan Cui, Zhongxia Wei, Dekai Ye, Weina Fang, Sisi Jia Published: 2026 ABSTRACT DNA origami placement (DOP) exploits shape complementarity between DNA nanostructures and lithographically patterned binding sites to achieve precise spatial positioning, offering a promising route for integrating bottom-up self-assembly with top-down nanomanufacturing. However, nonspecific aggregation of DNA origami on patterned binding sites limits both single-origami occupancy yield and orientational fidelity. Here, we investigated the aggregation mechanism in an asymmetrical right triangle DNA origami system and found that vertex-specific aggregation is primarily driven by unbalanced origami arm twisting. Atomic force microscopy (AFM) revealed that aggregation predominantly occurs at vertices associated with the hypotenuse arm C, where twisting exposes staple domains that promote interorigami binding. Selective removal of short staple strands at the C arm termini reduced rigidity at the vertex, thereby preventing exposure of the staple domain, suppressing interorigami aggregation, and increasing the monomer yield from 72 ± 1.1% to 87.3 ± 1.5%. In contrast, poly­(T) extension at the same sites yields only minor improvement, indicating that base stacking is not the dominant aggregation mechanism. When applied to directed placement on patterned substrates, the optimized design improved orientation matching yield from 33.1 ± 7.8% to 77.1 ± 4.8%, representing a significant improvement in placement fidelity. These results demonstrate that tuning local structural features provides an effective strategy to suppress aggregation and is essential for achieving high-fidelity placement of asymmetrical DNA origami. Introduction Nanomanufacturing typically relies on top-down processing methods centered around lithography techniques. − However, as device feature sizes continue to shrink toward the nanoscale, traditional lithography faces increasing challenges in resolution, fabrication cost, and process complexity. , To break this bottleneck, integrating bottom-up molecular self-assembly strategies with lithographic manufacturing has emerged as a promising route toward next-generation nanomanufacturing. − This strategy not only holds promise for partially or even completely replacing conventional lithography processes, thereby reducing production time and energy costs, but also offers a new avenue for constructing complex nanostructures. Among various self-assembly strategies, DNA origami technology has garnered significant attention due to its exceptional programmability and molecular recognition capabilities. − This technique involves folding a long single-strand DNA scaffold with hundreds of short staple strands to construct well-defined two-dimensional or three-dimensional DNA nanostructures with controllable dimensions. , Benefiting from the high specificity of Watson–Crick base pairing, DNA origami exhibits remarkable structural predictability and accuracy, − making DNA origami an ideal template for organizing functional components such as metallic nanoparticles, fluorophores, and proteins. To achieve controlled integration of DNA origami structures with functional devices, researchers have developed DNA origami placement (DOP) techniques. In DOP, the match between the overall shape of an origami and lithographically patterned binding sites is used to control its in-plane rotation θ. The advantage of DOP lies in its ability to orient thousands of origamis with high yield and fidelity: approximately 95% of sites are occupied by a single origami. , However, most current studies have focused on symmetric origami structures such as equilateral triangles, which can attach to their binding sites in one of multiple equivalent orientations and therefore cannot achieve absolute orientation control. − For applications requiring absolute orientation, such as nanoelectronic or optical devices with fully asymmetrical (C1-symmetric) structures, uniquely oriented placement must be achieved using structurally asymmetrical DNA origami. Yet even with asymmetrical designs, successful placement remains challenging. In one study using right triangles, only 40% of origami bound with the desired alignment, which is insufficient for reliable absolute orientation. This low yield has been attributed to a strongly kinetically trapped regime, in which origami enter the state space at random upon collision with a binding site and proceed directly to a local energy maximum. Furthermore, we observe that multiple right triangles frequently occupy a single binding site, indicating that aggregation during the placement process presents an additional obstacle to achieving high-fidelity, orientation-controlled arrays. However, it remains unclear whether such aggregation originates from preassembled triangles in solution or from interfacial interactions during the placement process itself. In this work, we employ a previously reported right triangle DNA origami as a model system to investigate whether its aggregation behavior affects the orientation accuracy of DOP. We observed that right triangle DNA origami undergoes vertex-to-vertex aggregation during solution assembly. To mitigate this issue, we strategically removed several staples at the vertex regions to increase structural flexibility, which reduced seam twist and minimized the inner-origami gap, thereby suppressing aggregation caused by cross-linking between origamis. , When the optimized origami was used for DOP, the orientation accuracy improved from 33.1 ± 7.8% (N = 150) to 77.1 ± 4.8% (N = 150), demonstrating that suppressing nonspecific interorigami interactions significantly improves placement fidelity. This strategy provides a new approach for achieving high-fidelity, directed placement of DNA origami arrays and offers broader guidance for the surface site-directed placement of other asymmetrical DNA structures. Results and Discussion Design and Placement of the Right Triangle DNA Origami In this study, we employed the right triangle DNA origami previously reported by Gopinath. As shown in Figure a, this structure is assembled from a long scaffold strand and more than 200 staple strands, forming a right triangle origami with an internal cavity. The two perpendicular arms measure 87 and 152 nm, respectively, giving the origami a pronounced geometric asymmetry. The detailed design of the right triangle DNA origami is shown in Figure S1 and Table S1. For clarity, the long perpendicular arm is denoted as A, the short perpendicular arm as B, and the hypotenuse as C; the corresponding vertices are labeled for the 90° angle vertex, b for the acute vertex adjacent to arm B, and c for the acute vertex adjacent to arm A. Asymmetrical right triangle DNA origami placement. (a) Design of the asymmetrical right triangle DNA origami. A single-stranded M13mp18 scaffold is annealed with staples to form origami triangles. (b) AFM image of the asymmetrical right triangle origami. (c) A step-by-step fabrication procedure of the DNA origami placement directed by the lithographically patterned binding sites. Step 1: a photoresist layer was spin-coated on top of the silicon oxide modified with a trimethylsilyl monolayer as a hydrophobic background. Step 2: define binding sites in the photoresist using EBL and develop the resist. Step 3: the silicon was treated with oxygen plasma for activating binding sites to create silanol groups and then the photoresist was removed. Step 4: DNA origami placement on the patterned binding site of the chip. (d,e) Large-area (d) and higher-magnification (e) AFM image of asymmetrical right triangle DNA origami placement on the patterned substrate. (f) Statistical analysis of correctly aligned and misaligned origami on the patterned substrate. Data are presented as mean ± SD from independently analyzed AFM images, with N = 150 analyzed origami structures. Scale bars: (b) 50 nm, (d) 1 μm, (e) 500 nm. In principle, an asymmetrical structure with C1 symmetry should enable unique orientation on shape-matched binding sites patterned on a substrate (Figure c). However, the experimental results showed that the origami did not exhibit the expected orientational selectivity after directed placement (Figure d). The layouts of the patterned binding sites and the corresponding AFM images after development are shown in Figures S2 and S3. Statistical analysis revealed that approximately 66.9 ± 7.8% (N = 150) of the structures failed to align in the designed orientation, whereas only 33.1 ± 7.8% (N = 150) were placed correctly (Error bars are all standard deviation unless otherwise stated). Further analysis indicated that these misoriented structures mainly originated from origami aggregates, together with a smaller fraction of damaged structures, with aggregation being the dominant factor. According to the study by Gopinath and Rothemund, during placement, origami can reach binding sites not only by direct diffusion from solution but also through an indirect pathway in which they first adsorb onto the hydrophobic background and then undergo two-dimensional diffusion to the target sites. When multiple origamis occupy the same patterned site and form dimers or higher-order aggregates during drying, the single-molecule orientational fidelity is disrupted. These observations highlight the need to identify the factors that govern the directed placement of DNA origami, with particular attention to the origin and mechanism of aggregation. Additional AFM images of right triangle DNA origami placement on patterned substrates are shown in Figure S4. Vertex-To-Vertex Aggregation Behavior and Mechanism To identify the key factors affecting placement quality, it is necessary to distinguish whether aggregation originates from the solution-phase self-assembly process or occurs during surface-directed placement. To address this, we evaluated the monodispersity of DNA origami after assembly in solution. The as-prepared samples were deposited onto mica and characterized by AFM. As shown in Figure a, large-area AFM images revealed a substantial population of dispersed origami structures, while magnified images (Figure b) indicated the coexistence of monomers and aggregates composed of two or more origami. Additional air-phase AFM images are shown in Figure S5. Statistical analysis of the AFM images (Figure c) showed that monomers account for 72.0 ± 1.1% (N = 150) of the structures, whereas aggregates constitute 28.0 ± 1.1% (N = 150). This result indicated that nearly one-third of the origami already exist as aggregates prior to placement. These preformed aggregates can directly bind to patterned sites during directed assembly, thereby reducing single-molecule placement yield and compromising orientational fidelity. Aggregation behavior of right triangle DNA origami and proposed mechanism. (a) Large-area AFM image of assembled DNA origami on a mica surface. (b) Representative AFM image highlighting monomeric and aggregated origami. (c) Statistical distribution of monomeric and aggregated structures in the assembled sample. Data are presented as mean ± SD, with N = 150 analyzed origami structures. Error bars represent standard deviation. (d) Statistical analysis of aggregation configurations (b–b, b–c, and c–c interactions). Percentages were obtained by manual classification of aggregate configurations from AFM images; data are presented as mean ± SD, with N = 50 aggregates (e) High-resolution AFM image of vertex regions, highlighting open and closed vertex configurations. (f) Root-mean-square fluctuation (RMSF) map of the right triangle DNA origami, with a magnified view of the vertex region highlighting design details. (g) Proposed aggregation mechanism: geometric asymmetry induces local structural relaxation at vertices associated with the hypotenuse, thereby promoting vertex-to-vertex interactions between origamis. Scale bars: (a) 500 nm, (b,e) 200 nm. AFM images (Figure b) revealed distinct structural features at different vertices. Both monomeric and aggregated structures exhibit relatively open configurations at vertices b and c, which correspond to the junctions between adjacent arms, whereas the right triangle vertex (a) remains comparatively closed and structurally rigid. Statistical analysis of aggregation configurations (Figure d) showed that aggregation predominantly occurs via vertex-to-vertex interactions, with b–b, b–c, and c–c configurations accounting for 37.5 ± 3.7%, 9.7 ± 2.6%, and 52.8 ± 4.9% (N = 50), respectively. Enlarged AFM images showed that the right-triangle origami exhibited both closed and opened vertex conformations (Figure e). The opened vertices were mainly observed at the acute vertex regions, suggesting that local structural deformation at these sites may facilitate vertex-to-vertex interactions. These results suggest that aggregation is primarily associated with vertices b and c. To further support this interpretation at the structural level, we performed molecular dynamics (MD) simulations on the right triangle DNA origami. As shown in Figure f, the RMSF (root-mean-square fluctuation) reveals enhanced local fluctuation at the acute-vertex region (especially vertices c), indicating that this region is more flexible and more prone to local opening. Such increased flexibility may facilitate the exposure of linker/staple domains, thereby promoting vertex-to-vertex interactions between neighboring origami (Figure g). Based on these observations, we propose a mechanistic explanation for the aggregation behavior. Owing to the geometric asymmetry of the right triangle DNA origami, the local structural rigidity varies across different vertices. The right triangle vertex a is relatively stable, whereas the acute vertices b and c, particularly those associated with the hypotenuse C arm, are more prone to local structural deformation. Such deformation increases the separation between arms, preventing the connecting strands from bridging both arms simultaneously. Instead, these strands may hybridize with the scaffold segments on adjacent origami, forming clear DNA bridges between neighboring origami structures (Figure g) and thereby promoting interorigami aggregation. These findings suggest that reducing structural rigidity at key vertices may facilitate proper interarm connections and mitigate aggregation. Regulation of DNA Origami Aggregation via Vertex Structural Modification To test the proposed mechanism and explore structural strategies for mitigating origami aggregation, we sought to modulate the local structural constraints at vertices associated with the C arm, thereby weakening intervertex interactions. Specifically, two structural variants were designed (Figure a). We use a Δ-based notation to denote localized staple modifications. ΔC refers to deletion of selected short staples at the C-arm-associated terminal region, while ΔAB denotes deletion of corresponding staples in control regions along arms A and B. In the first design (ΔC), a subset of short staple strands at the terminal region of the C arm was removed to reduce local structural tension and promote a more closed vertex configuration. As a control, a second variant (ΔAB) was constructed by removing corresponding short staples near the C-arm termini along arms A and B. The sequences of removed staple strands are listed in Tables S2 and S3. Structural modulation of DNA origami to suppress aggregation. (a) Schematic illustration of vertex modification strategies. Short staple strands at the C arm (red) were removed (ΔC), while deletion at the A and B arms (blue) (ΔAB) was used as a control. (b) Agarose gel electrophoresis analysis of DNA origami structures (M13 scaffold, initial structure, ΔC, and ΔAB). (c) Relative band intensities of monomeric and aggregated structures for different origami designs. Band intensities were quantified using ImageJ and normalized to the corresponding total lane intensity. (d) Relative mobilities of monomeric and aggregated bands derived from gel electrophoresis. Band mobilities were quantified using ImageJ and normalized to the corresponding reference band position. (e) AFM image of ΔC-modified DNA origami on a mica surface. (f) Statistical distribution of monomeric and aggregated structures for the ΔC design. Data are presented as mean ± SD, with N = 150 analyzed origami structures. Error bars represent standard deviation. Scale bars: (e) 200 nm. Both variants were assembled following the standard annealing protocol and characterized by 0.6% agarose gel electrophoresis (Figure b). The relative migration and band intensities were quantified using ImageJ. Compared with the original structure, the ΔC variant exhibits an increase of approximately 14% in monomer band intensity (Figure c), indicating a substantial improvement in monodispersity. In addition, a slight decrease (∼2%) in electrophoretic mobility is observed (Figure d), suggesting a marginal reduction in structural compactness and rigidity of the monomeric origami. In contrast, the ΔAB variant showed only modest changes, with a ∼7% increase in monomer intensity and a ∼3% decrease in aggregate intensity. These results indicated that structural modulation at the C-arm-associated vertices plays a more critical role in suppressing aggregation than modifications at other regions. To further evaluate the effect of the ΔC design, AFM characterization was performed under dry conditions (Figure e), followed by statistical analysis of the observed structures. From the AFM images, we observed that the vertices of the origami structures appeared closed, indicating a reduced distance between the arms. The monomer yield of the investigated origami increased to 87.3 ± 1.5% (N = 150), representing a 15.3 ± 1.5% (N = 150) improvement compared to the original design (Figure f), accompanied by a significant reduction in aggregates. These results provide direct experimental support for the proposed aggregation mechanism: by tuning the local structural constraints at vertices associated with the hypotenuse, vertex-to-vertex interactions can be effectively suppressed, thereby improving monomer yield and overall assembly quality. To evaluate whether the ΔC modification compromises structural integrity, we further analyzed the stability of the ΔC origami under the experimental conditions used in this work. As shown in Figure S6, the ΔC modified origami maintains a high fraction of intact triangular structures, with only a slight increase in defective structures over 24 h. These results indicate that, although the ΔC design locally reduces structural constraints, it retains sufficient structural integrity within the time scale and buffer conditions relevant to the placement process. Probing the Role of Base Stacking via polyT Modification In DNA origami systems, base stacking interactions are often considered a potential driving force for structural aggregation. To evaluate whether the aggregation observed in this study is governed by base stacking, we introduced a flexible poly­(T) extension (five thymine bases) at the 5′ termini of staple strands along the C arm (Figure a). This modification is expected to disrupt terminal base stacking and thereby weaken interorigami interactions. Representative air-phase AFM images of the polyT-modified structures are shown in Figure S7. The modified structures were compared with the unmodified origami and characterized by 0.6% agarose gel electrophoresis (Figure a). Effect of poly­(T) modification on DNA origami aggregation behavior. (a) Schematic illustration of poly­(T) extensions at the termini of C-arm staple strands. Agarose gel electrophoresis of the initial and poly­(T)-modified DNA origami structures. (b) Relative mobilities of monomeric and aggregated species derived from gel electrophoresis. Band mobilities were quantified using ImageJ and normalized to the corresponding reference band position. (c) Representative AFM image of poly­(T)-modified DNA origami on a mica surface. (d) Statistical distribution of monomeric and aggregated structures for the poly­(T)-modified design. Data are presented as mean ± SD, with N = 150 analyzed origami structures. Error bars represent standard deviation. Scale bars: (c) 200 nm. Gel electrophoresis analysis (Figure b) showed that the poly­(T)-modified origami exhibits an increase of approximately 5% in electrophoretic mobility relative to the original structure, suggesting a modest change in structural properties, potentially associated with increased rigidity or reduced interparticle interactions. To further assess aggregation behavior, the poly­(T)-modified origami was characterized by AFM under dry conditions, followed by statistical analysis (Figure d). The fraction of monomeric structures increased to 75.3 ± 6.1% (N = 150), only slightly higher than that of the original design 72.0 ± 1.1%, and significantly lower than that achieved with the ΔC modification 87.3 ± 1.5%. These results suggest that introducing flexible poly­(T) extensions at vertex regions is insufficient to effectively suppress aggregation, indicating that base stacking is unlikely to be the dominant driving force for aggregation in this system. Instead, we attribute the aggregation behavior primarily to deformation arising from local structural constraints at vertices associated with the C arm, rather than terminal base stacking interactions. High-Fidelity Directed Placement of DNA Origami Arrays Building on the improved monodispersity achieved with the ΔC design, we next evaluated its performance in directed placement on patterned substrates, aiming to assess its applicability in device integration. Arrays of shape-matched right triangle binding sites with varying orientations were fabricated on silicon substrates using electron-beam lithography. The orientation of adjacent sites was incremented in 30° steps (Figure a), enabling systematic evaluation of orientational control. Orientation-controlled placement of optimized DNA origami on patterned substrates. (a) Schematic illustration of triangular binding sites with different orientations (−90°, −60°, −30°, 0°, and +30°, +60°, +90°) used to evaluate placement accuracy. (b) Large-area AFM image showed the placement of optimized DNA origami on the patterned substrate. (c) Enlarged AFM image highlighting the orientation of individual origami. (d) Statistical analysis of aligned and misaligned origami on the patterned arrays. Data are presented as mean ± SD from independently analyzed AFM images, with N = 150 analyzed origami structures. Error bars represent standard deviation. Scale bars: (b)­1 μm (c) 200 nm. Large-area AFM images (Figure b) show that the ΔC origami are uniformly distributed across the patterned substrate, with the majority of structures aligned in accordance with the designed orientation of their corresponding binding sites. Magnified images (Figure c) further revealed well-defined triangular contours at individual sites, allowing unambiguous identification of origami orientation and indicating that the ΔC structures maintain structural integrity upon surface binding. To quantitatively assess placement fidelity, the orientations of origami across multiple AFM images were statistically analyzed. As shown in Figure d, the ΔC design achieved an orientation matching yield of 77.1 ± 4.8% (N = 150), representing a substantial improvement over the 33.1 ± 7.8% obtained with the original structure. These results demonstrate that suppressing aggregation through vertex structural modificationand thereby increasing monomer yieldis critical for enhancing the quality of directed placement. Additional AFM and SEM images further confirm the placement results after structural optimization (Figures S8 and S9). Furthermore, we analyzed orientation distribution shown in Figure S10. In this analysis, the horizontal axis Δθ is defined as the difference between the designed orientation and the experimentally measured orientation (Δθ = θ design - θ measured). The results show that the Δθ distribution is centered around 0°, indicating that the ΔC origami maintains good orientation fidelity during placement. Compared with the orientation precision reported for the “small moon” origami ±3.2°, further improvement in alignment accuracy is still possible, likely limited by the intrinsic geometric properties of the right triangle design. Nevertheless, our results demonstrate that fine control over local structural features of DNA origami enables high-fidelity directed placement on patterned substrates, providing a robust foundation for the integration of functional nanodevices. Conclusion In this work, we systematically investigated the factors governing aggregation behavior of right triangle DNA origami and demonstrated that suppressing interorigami aggregation promotes the formation of highly monodisperse and orientationally controlled arrays. Based on AFM characterization and statistical analysis of the original structures, we propose that aggregation is primarily driven by vertex-to-vertex interactions arising from local structural separation induced by geometric deformation. Guided by this mechanism, we introduced a structural modification (ΔC) by selectively removing staple strands at the terminal region of the C arm to modulate local vertex constraints. This design increased the monomer yield to 87.3 ± 1.5%, providing direct experimental support for the proposed mechanism. When applied to directed placement on lithographically patterned substrates, the optimized ΔC origami achieved an orientation matching yield of 77.1 ± 4.8%, substantially outperforming previously reported right triangle DNA origami systems that typically yield of around 40%, enabling the formation of highly uniform and well-aligned origami arrays. These findings indicate that modulating the local structural features of DNA origami is an effective strategy for suppressing nonspecific aggregation interorigami aggregation. The optimized structures not only exhibit markedly improved monodispersity but also enable high-fidelity, orientation-controlled placement on patterned surfaces. This approach offers a pathway toward the precise, directed assembly of DNA origami arrays and provides broader insights for the site-specific surface placement of other asymmetrical DNA nanostructures. Supplementary Material