DNA Mold-Based Fabrication Method

• DNA mold-based fabrication is a bottom-up method that uses DNA’s programmed geometry to template inorganic or hybrid nanostructures.
• It integrates techniques like Langmuir–Blodgett film transfer, DNA origami molds for metal growth, and ALD coatings for 3D architectures.
• Process tunability is achieved by controlling experimental parameters such as surface pressure, reagent concentrations, and post-growth annealing.

Searching arXiv for the cited DNA mold-based fabrication papers and closely related work. [Tool call: arxiv_search] Query: "DNA mold-based fabrication silver nanostructures (Hadlich et al., 18 Sep 2025) Helmi Ye molecular electronic technologies (Helmi et al., 17 Mar 2025) Langmuir-Blodgett DNA films (Dai et al., 2010) atomic layer deposition DNA origami crystals (Ermatov et al., 2024)" DNA mold-based fabrication denotes a class of bottom-up nanofabrication strategies in which DNA functions as a geometrically programmed or interfacially organized structural template for the formation of inorganic, hybrid, or device-integrated architectures. In the literature represented here, the term spans fractal DNA films transferred by the Langmuir–Blodgett technique, hollow DNA origami cavities that confine seeded metal growth, and micrometre-sized DNA origami crystals that are conformally coated by atomic layer deposition. Across these implementations, the shared principle is that DNA specifies nanoscale topology first, while transfer, metallization, conformal coating, or electrical contacting converts that topology into a functional material system (Dai et al., 2010, Ermatov et al., 2024, Helmi et al., 17 Mar 2025, Hadlich et al., 18 Sep 2025).

1. Conceptual scope and core implementations

The methodological scope of DNA mold-based fabrication is broader than direct DNA metallization alone. One implementation uses calf-thymus DNA organized at an air–water interface and transferred as nanopatterned films whose morphology is governed by Langmuir–Blodgett surface pressure. A second uses DNA origami molds with an internal cavity and seed-binding sites to direct the growth of continuous metallic nanostructures. A third employs DNA origami crystals as periodic 3D frameworks that are first stabilised, then conformally coated with metal oxides by ALD. A fourth integrates DNA-directed self-assembly with top-down nanofabrication to construct single-molecule transport devices (Dai et al., 2010, Ermatov et al., 2024, Helmi et al., 17 Mar 2025, Hadlich et al., 18 Sep 2025).

Platform	DNA geometry	Fabrication outcome
Langmuir–Blodgett DNA film	Fractal-like nanopatterned DNA thin films on mica	DNA networks and silver metal patterns
DNA origami cavity mold	Hollow cuboid of 64 parallel duplex helices	Continuous silver nanowires
DNA origami crystal	Diamond-lattice crystal of tetrapod monomers	ZnO, TiO$_2$, and IrO$_2$ coated 3D nanoarchitectures
DNA-directed molecular package	Five-layer DNA origami block with molecular adapters	Single-molecule electrical transport devices

A common misconception is to treat these as variants of a single experimental recipe. The record instead shows several distinct fabrication logics: interfacial pattern transfer, cavity-confined seeded growth, conformal infiltration or shell deposition, and hybrid bottom-up/top-down device assembly. This suggests that “DNA mold-based fabrication” is best understood as a methodological family defined by the use of DNA to predefine geometry, rather than by any single chemistry or instrument.

2. Langmuir–Blodgett DNA films as interfacial molds

In the Langmuir–Blodgett implementation, calf-thymus deoxyribonucleic acid was dissolved in deionized water (18 M$\Omega\cdot$cm) to a final concentration of 3 $\mu$g mL$^{-1}$ at pH 6.5 and room temperature ($\approx 20\,^\circ$C). The spreading phase was octadecylamine in chloroform at 1 mM. The subphase was either pure water or the DNA aqueous solution, the temperature was $20\,^\circ$C, and nitrogen flow was used to minimize dust and oxidation. The ODA solution was allowed 20 min solvent evaporation before compression in an ATEMETA LB-105 trough; barrier compression speed was 1 cm min$^{-1}$, and substrate dipping speed was 4 mm min$^{-1}$ (Dai et al., 2010).

The surface-pressure versus molecular-area isotherms showed that on pure water the collapse area was $\approx 21$ $_2$0 per ODA molecule, whereas on the DNA subphase the limiting area expanded by $_2$1 $_2$2, interpreted as evidence of DNA entrapment under the ODA monolayer. Morphology was strongly pressure-dependent. Fractal–dendritic networks were transferred at 10–20 mN m$_2$3, a close-packed network at $_2$4 mN m$_2$5, and overcompressed aggregates or “islands” at $_2$6 mN m$_2$7. AFM images at 15 mN m$_2$8 showed an open, highly ramified fractal with pore sizes up to $_2$9 nm; at 20 mN m$\Omega\cdot$0, a denser fractal with branch widths $\Omega\cdot$1 nm and fiber height $\Omega\cdot$2 nm; at 30 mN m$\Omega\cdot$3, an almost closed network with pores $\Omega\cdot$4 nm; and at 45 mN m$\Omega\cdot$5, overcompressed islands in which fractal character was lost (Dai et al., 2010).

Characterization combined dynamic force microscopy in tapping mode with phase imaging and XPS. AFM used an SPA400 instrument with silicon microcantilevers of spring constant $\Omega\cdot$6 N m$\Omega\cdot$7, in air at room temperature. XPS employed monochromatic Al K$\Omega\cdot$8 radiation at 15 kV/10 mA, take-off angle $\Omega\cdot$9, and pass energies of 80 eV for survey and 40 eV for high-resolution spectra. The reported peaks were C 1s at 284.6 eV, N 1s at 399.7 eV, and P 2p at 133.6 eV, the latter confirming the phosphate backbone of DNA in the composite film (Dai et al., 2010).

The growth mechanism was discussed in terms of diffusion-limited aggregation. The stated sequence was adsorption of DNA onto the subphase/monolayer interface, surface diffusion of DNA under the ODA monolayer, nucleation at “seed” points with irreversible branching, and coalescence into denser networks or islands upon further compression. The associated scaling forms were

$\mu$0

and

$\mu$1

The original work did not explicitly measure $\mu$2, but the use of these expressions locates the morphology within a DLA framework (Dai et al., 2010).

The DNA networks were subsequently metallized using silver acetate solution and hydroquinone solution, each of volume 100 mL; the DNA-LB-coated mica was immersed in the mixed bath for 2 min at room temperature, then rinsed with deionized water and dried under N$\mu$3. The resulting silver wires exactly traced the DNA network, with average wire width $\mu$4 nm, average wire height $\mu$5 nm, and net Ag coating thickness $\mu$6–7 nm. XPS after deposition showed disappearance of P 2p and appearance of Ag $\mu$7 at 368.2 eV and Ag $\mu$8 at 374.2 eV, confirming Ag$\mu$9 formation (Dai et al., 2010).

3. DNA origami cavity molds for continuous silver nanostructures

A more explicitly “mold-based” architecture uses a hollow DNA origami cuboid of 64 parallel duplex helices with outer dimensions $^{-1}$0 nm and inner cavity $^{-1}$1 nm. Four single-stranded anchor extensions project into the cavity center for gold seed binding, and orthogonal overhang patterns on the cuboid ends, denoted interfaces A/B/C/D, mediate chain formation by specific hybridization. Assembly used 10 nM M13-derived p8064 ssDNA scaffold and 100 nM of each staple in folding buffer containing 5 mM Tris·HNO$^{-1}$2, 1 mM EDTA, 11 mM Mg(NO$^{-1}$3)$^{-1}$4, and 5 mM NaNO$^{-1}$5 at pH 8.0, with annealing from $^{-1}$6C for 5 min to $^{-1}$7C over 15 h, followed by PEG precipitation (Hadlich et al., 18 Sep 2025).

Gold seeds were introduced by conjugating 5 nm AuNPs to 5′-thiolated DNA via salt-aging, then hybridizing the DNA–AuNP seeds to mold anchors at $^{-1}$8C $^{-1}$9C at 1 K/17 min in buffer containing 0.5 mM TBE, 11 mM Mg(NO$\approx 20\,^\circ$0)$\approx 20\,^\circ$1, and 350 mM NaNO$\approx 20\,^\circ$2. A three-fold excess of AuNPs per binding site ensured $\approx 20\,^\circ$3 loading. Seed-loaded monomers were PEG-purified and multimerized overnight at room temperature by mixing complementary interfaces in 1:1 stoichiometry, yielding linear chains up to nine units long (Hadlich et al., 18 Sep 2025).

Silver growth required exchange into 0.1 M boric acid, 11 mM Mg(NO$\approx 20\,^\circ$4)$\approx 20\,^\circ$5, pH 8.0. All subsequent steps were carried out under red-light illumination and in the absence of Cl$\approx 20\,^\circ$6 to suppress silver–chloride precipitation and photoinduced nucleation. The theoretical filling ratio was defined as

$\approx 20\,^\circ$7

with $\approx 20\,^\circ$8 Ag atoms per mold. Thus, 50% TFR corresponds to $\approx 20\,^\circ$9 Ag atoms per mold or 80 $20\,^\circ$0M AgNO$20\,^\circ$1 when using 0.25 nM mold. Under optimal conditions, AuNP-loaded molds at 0.1 nM and two seeds each were exposed to 160 $20\,^\circ$2M AgNO$20\,^\circ$3 and 160 $20\,^\circ$4M NH$20\,^\circ$5OH in 1:1 stoichiometry for 10 min; growth was stated to be essentially complete after 2 min, but extended to 10 min for uniformity (Hadlich et al., 18 Sep 2025).

The principal mechanistic complication was the strong interaction of Ag$20\,^\circ$6 with DNA. Ag$20\,^\circ$7 was described as binding electrostatically to the sugar–phosphate backbone and preferentially to N7 of guanine and N3 of cytosine, even forming Ag$20\,^\circ$8-mediated G–Ag–G base pairs. Excess Ag$20\,^\circ$9 could distort the mold and create spontaneous nucleation points. Seedless or off-template growth became significant at higher seed concentrations or excessive Ag$^{-1}$0: at seed concentrations $^{-1}$1 nM, Ag$^{-1}$2/hydroxylamine spontaneously nucleated off-template particles, and Ag$^{-1}$3:HA ratios $^{-1}$4 produced spontaneous satellites. The stated reduction step was

$^{-1}$5

with reduction complete within minutes (Hadlich et al., 18 Sep 2025).

Post-growth thermal annealing on TEM grids was introduced to promote coalescence. Samples were ramped from ambient to target temperature at $^{-1}$6 K/min in static air and held for 30 min at 150, 200, 250, or $^{-1}$7C. At $^{-1}$8C the jagged, grainy surface became smooth and segments fused; above $^{-1}$9C, Ostwald ripening generated necked or broken structures. The optimum was 150 $^{-1}$0C for 30 min. Structural characterization gave diameters $^{-1}$1 nm, mean segment length 488 nm at 0.1 nM seeds and 200% TFR ($^{-1}$2), and 335 nm after 150 $^{-1}$3C anneal ($^{-1}$4). SAED showed fcc Ag peaks (111, 200, 220), no Ag$^{-1}$5O rings, and a slight 200-plane shift indicating surface oxidation; STEM-EDS showed strong Ag L signal and only trace O K at edges. After N$^{-1}$6-plasma removal of DNA and EBL contacting with Cr/Au, all 17 wires were insulating, with $^{-1}$7 pA at 5 V (Hadlich et al., 18 Sep 2025).

This result directly addresses a recurrent misunderstanding in the field: faithful geometric replication and apparent structural continuity do not by themselves establish electronic continuity. The silver nanowires were reported as not oxidized in the abstract, yet the detailed characterization and transport measurements were consistent with persistent grain-boundary or surface-oxide barriers (Hadlich et al., 18 Sep 2025).

4. DNA origami crystals as 3D molds for atomic layer deposition

DNA mold-based fabrication also extends from 1D channels and 2D networks to 3D periodic architectures. In the ALD-based implementation, a single-strand M13-derived scaffold of 8 634 nucleotides was folded by $^{-1}$8 distinct staple strands, each at final 120 nM, into a tetrapod monomer comprising four 35 nm-long arms, each a 24-helix bundle of diameter $^{-1}$9 nm. These monomers assembled into a diamond-lattice, rod-connected cubic crystal with crystal diameter 5–10 $\approx 21$0m, pore diameter $\approx 21$1 nm, and lattice constant $\approx 21$2 nm (Ermatov et al., 2024).

Folding used 1$\approx 21$3 TE (10 mM Tris-HCl, 1 mM EDTA, pH 8.0) plus 20 mM MgCl$\approx 21$4, and crystal growth used 1$\approx 21$5 TE plus 26 mM MgCl$\approx 21$6. The monomer ramp was $\approx 21$7C to $\approx 21$8C over 25 min at 2 min/$\approx 21$9C, hold at $_2$00C for 10 min, then cool to $_2$01C at 0.1 $_2$02C/min. Crystal annealing was $_2$03C $_2$04C at 2.5 h/$_2$05C, followed by incubation at $_2$06C for 8 h, removal of supernatant, and repetition of the $_2$07C $_2$08C ramp (Ermatov et al., 2024).

Because the physical and chemical stability of DNA nanostructures is generally limited to aqueous buffer conditions, a silicification step was used to stabilise the crystals. TMAPS was added to a final concentration of 0.5 mM and shaken for 30 min, TEOS to 0.3 mM and shaken for 1 h, then acetic acid at 0.5 M was added with a temperature ramp from $_2$09C to $_2$10C over 1 h, followed by two washes with H$_2$11O and two with IPA, each with 12 h sedimentation. The simplified reaction was

$_2$12

This produced a conformal, $_2$13–3 nm SiO$_2$14 shell around every helix, increased mechanical rigidity sufficiently for air drying without collapse, and preserved the $_2$15 nm pore size and internal connectivity (Ermatov et al., 2024).

ALD was then used to deposit ZnO, TiO$_2$16, and IrO$_2$17. For ZnO, diethylzinc and H$_2$18O were used at $_2$19C with base pressure $_2$20 Torr and N$_2$21 carrier 150 sccm; the pulse sequence was DEZ $_2$22 s, purge 15 s, H$_2$23O 1 s, purge 14 s, with stop-flow during pulse. For TiO$_2$24, TTIP with H$_2$25O at $_2$26C used the same timing, whereas TDMAT with H$_2$27O was run at $_2$28C or $_2$29C with Ar 420 sccm, TDMAT 0.25 s/25 s purge, and H$_2$30O 0.06 s/25 s purge. For IrO$_2$31, Ir(acac)$_2$32 and O$_2$33 were used at $_2$34C with N$_2$35 40 sccm and O$_2$36 generated from 1% O$_2$37 in N$_2$38 at 500 sccm; each cycle comprised Ir(acac)$_2$39 1.6 s + 5 s static + 10 s purge repeated three times, followed by O$_2$40 4 s + 2 s static + 10 s purge (Ermatov et al., 2024).

Growth per cycle values were $_2$41 $_2$42 for ZnO, $_2$43 $_2$44 for TiO$_2$45, and $_2$46 $_2$47 for IrO$_2$48, with thickness determined by

$_2$49

Representative examples were 22 ZnO cycles giving 4.9 nm, 160 TiO$_2$50 cycles giving 5.1 nm, and 80 IrO$_2$51 cycles giving 3.1 nm. SEM/FIB cross-sections and EDX maps indicated a uniform oxide shell on all interior rods and surfaces, with no gradient from center to edge. Direct ALD on bare DNA crystals was also demonstrated by combining critical point drying with low-temperature ALD: after IPA exchange and CO$_2$52 supercritical drying, TiO$_2$53 deposition at $_2$54C yielded an $_2$55 nm shell while retaining octahedral shape and a hexagonal surface motif with minor distortion (Ermatov et al., 2024).

As a proof-of-concept functional application, IrO$_2$56-coated DNA crystals were used for electrocatalytic water oxidation. On FTO/glass (12 $_2$57/sq) in 0.5 M H$_2$58SO$_2$59, 80 ALD cycles produced a 3 nm IrO$_2$60 shell. At 1.8 V vs RHE in cycle 2, current density was $_2$61 mA/cm$_2$62 for the reference without DNA, $_2$63 mA/cm$_2$64 for 1$_2$65DNA, and $_2$66 mA/cm$_2$67 for 3$_2$68DNA. The overpotential at $_2$69 mA/cm$_2$70 was 1.58 V for the reference, 1.55 V for 1$_2$71, and 1.53 V for 3$_2$72. Current decreased by $_2$73–30% from cycle 2 to 20 due to IrO$_2$74 dissolution and partial crystal detachment, while SEM after OER showed preserved morphology (Ermatov et al., 2024).

5. Device-level integration in molecular electronics

DNA mold-based fabrication has also been coupled to conventional nanofabrication to produce single-molecule electrical transport devices. The relevant origami was a five-layer DNA origami block of $_2$75 nm built on the p7249 M13 ssDNA scaffold at 10 nM with $_2$76 unique staple strands at 200 nM each. Two staples were replaced by porphyrin-conjugated adapters A and B, 280 nM each, to capture a Zn-porphyrin derivative bearing two terminal alkynes, SMe gold-coordination groups, and an internal $_2$77-conjugated core. Folding used 10 mM Tris, pH 8.0, 1 mM EDTA, and 16 mM MgCl$_2$78, with annealing at $_2$79C for 20 min followed by a ramp to $_2$80C over 16 h at 1 $_2$81C per 45 min, then holding at $_2$82C. Purification used three rounds of PEG precipitation at 20 000 rcf for 30 min at $_2$83C (Helmi et al., 17 Mar 2025).

Porphyrin–DNA conjugation used two distinct 19–25 nt oligomers, one 3′-Cy3–azide and one 5′-Cy5–azide. In templated CuAAC, adapters A, B, and a third template strand were annealed at 10 $_2$84M each in 1$_2$85 TBE and 11 mM MgCl$_2$86, followed by addition of Zn-porphyrin at 7$_2$87 molar excess in 15% DMF, then 600$_2$88 sodium ascorbate, 420$_2$89 THPTA, and 60$_2$90 CuSO$_2$91, with 16 h incubation at $_2$92C. Analysis by 20% denaturing PAGE, HPLC, and UV–Vis confirmed the product, with a Soret band at $_2$93 nm, Cy3 at 554 nm, and Cy5 at 648 nm (Helmi et al., 17 Mar 2025).

Gold nanoparticle integration used citrate 60 nm AuNPs functionalized with 5′-thiol–poly-T$_2$94, then mixed with purified origami at a 1:5 origami:AuNP ratio in 0.5$_2$95 TBE/11 mM MgCl$_2$96. The package was heated to $_2$97C and cooled to $_2$98C over 1 h at $_2$99C/3.5 min, then purified by 1% agarose gel. Typical stepwise yields were $\Omega\cdot$00 for origami folding, $\Omega\cdot$01 for porphyrin conjugation, $\Omega\cdot$02 for AuNP docking, and $\Omega\cdot$03 for package purification, giving an overall package yield

$\Omega\cdot$04

The text summarized this as $\Omega\cdot$05 (Helmi et al., 17 Mar 2025).

Subsequent processing transferred the DNA-assembled packages to Si wafers with 100 nm thermal SiO$\Omega\cdot$06, pre-patterned by EBL with large Au markers and an $\Omega\cdot$07 local grid of 2 $\Omega\cdot$08m-spaced 100 $\Omega\cdot$09m crosses. PMMA 950K A6 at 4500 rpm gave a $\Omega\cdot$10 nm resist layer; 2 $\Omega\cdot$11m windows were opened by exposure and development in MIBK:IPA (1:3), followed by drop-casting 5 $\Omega\cdot$12L of purified packages, ambient drying, DI-water rinse, and acetone removal of PMMA. SEM localization identified AuNP dimers relative to grid and markers with $\Omega\cdot$13 nm uncertainty. Contact electrodes were then patterned with a $\Omega\cdot$14 nm gap and formed by thermal evaporation of 5 nm Cr / 120 nm Au. Single-AuNP test devices at 77 K exhibited $\Omega\cdot$15, with linear $\Omega\cdot$16–$\Omega\cdot$17 up to 1 mA (Helmi et al., 17 Mar 2025).

Electrical characterization of the completed devices established the importance of DNA-directed placement for yield. Empty-gap controls at 2 nm gave $\Omega\cdot$18 devices with $\Omega\cdot$19. Zn-porphyrin devices at 2 nm gave $\Omega\cdot$20 devices with a bimodal conductance distribution: a low-$\Omega\cdot$21 mode with $\Omega\cdot$22 and $\Omega\cdot$23, and a high-$\Omega\cdot$24 mode with $\Omega\cdot$25 and $\Omega\cdot$26, corresponding to $\Omega\cdot$27 of devices. The overall device yield for functional molecular bridging was $\Omega\cdot$28, compared to $\Omega\cdot$29 for electromigrated or STM break junctions. The general yield model was stated as

$\Omega\cdot$30

For $\Omega\cdot$31 heterogeneous components, the scaling relation was

$\Omega\cdot$32

with $\Omega\cdot$33 and $\Omega\cdot$34 (Helmi et al., 17 Mar 2025).

6. Tunability, limitations, and interpretation

Across the reported implementations, tunability is achieved by a small set of experimentally direct control variables. In the Langmuir–Blodgett case, transfer surface pressure controls whether the DNA film is open and dendritic, nearly closed, or aggregated into islands. In the cavity-confined silver-growth case, morphology depends on mold concentration, AgNO$\Omega\cdot$35 concentration, Ag$\Omega\cdot$36:reductant stoichiometry, and post-growth annealing. In the ALD case, layer thickness follows $\Omega\cdot$37, so cycle number directly sets shell thickness, while process temperature determines whether direct coating of bare DNA crystals is possible. In the molecular electronics workflow, overall yield is governed by multiplicative stepwise probabilities rather than by any single bottleneck parameter (Dai et al., 2010, Ermatov et al., 2024, Helmi et al., 17 Mar 2025, Hadlich et al., 18 Sep 2025).

Several limitations recur. DNA nanostructures are generally limited to aqueous buffer conditions unless stabilised; the origami crystal work addressed this by silicification and by low-temperature ALD after critical point drying (Ermatov et al., 2024). Silver is particularly challenging because Ag$\Omega\cdot$38 interacts strongly with DNA, which promotes off-template nucleation and structural heterogeneity unless chloride is excluded, illumination is controlled, and reagent concentrations are constrained (Hadlich et al., 18 Sep 2025). Morphological fidelity also does not guarantee electronic functionality: silver wires that exactly traced a DNA network were obtained in the Langmuir–Blodgett study, but conductivity was not the focus there, while the later continuous silver nanowires remained insulating in two-terminal measurements despite successful templating (Dai et al., 2010, Hadlich et al., 18 Sep 2025).

The broader significance of the method family lies in the decoupling of geometry definition from material conversion. DNA establishes cavity, network, or crystal topology with nanometre-scale addressability; metallization, ALD, or top-down contacting then imparts optical, catalytic, or electronic function. A plausible implication is that progress in DNA mold-based fabrication will depend less on the precision of DNA self-assembly—which is already high in these examples—than on controlling interfacial chemistry during conversion to inorganic or device-compatible forms.