ARTICLE

https://doi.org/10.1038/s41467-021-21920-y OPEN

Digital logic gates in soft, conductive mechanical

metamaterials
Charles El Helou1, Philip R. Buskohl2, Christopher E. Tabor2 & Ryan L. Harne 1✉

Integrated circuits utilize networked logic gates to compute Boolean logic operations that are
the foundation of modern computation and electronics. With the emergence of flexible
1234567890():,;

electronic materials and devices, an opportunity exists to formulate digital logic from com-
pliant, conductive materials. Here, we introduce a general method of leveraging cellular,
mechanical metamaterials composed of conductive polymers to realize all digital logic gates
and gate assemblies. We establish a method for applying conductive polymer networks to
metamaterial constituents and correlate mechanical buckling modes with network con-
nectivity. With this foundation, each of the conventional logic gates is realized in an
equivalent mechanical metamaterial, leading to soft, conductive matter that thinks about
applied mechanical stress. These findings may advance the growing fields of soft robotics and
smart mechanical matter, and may be leveraged across length scales and physics.

1 Department of Mechanical Engineering, The Pennsylvania State University, University Park, PA, USA. 2 Materials and Manufacturing Directorate, Air Force

Research Laboratory, Wright-Patterson Air Force Base, OH, USA. ✉email: [email protected]

NATURE COMMUNICATIONS | (2021)12:1633 | https://doi.org/10.1038/s41467-021-21920-y | www.nature.com/naturecommunications 1

ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-21920-y

S and transmit sensory and motor impulses31, a means to cultivate

tructurally and materially compliant integrated circuits with
reconfigurable electrical functions are an essential founda- decision-making capability in mechanically-robust compliant
tion for human-machine interfaces, soft robotics, and other materials would provide a significant step towards autonomous
future electronics that will serve medicine, science, engineering, soft matter able to assess and react to mechanical-electrical sti-
and industry. These compliant conductors alleviate concerns of muli in a dynamic environment.
failure traditionally encountered when conventional metallic This report introduces a class of soft, conductive mechanical
conductors are subjected to mechanical stress1–5. Liquid metal is metamaterials with programmable elastic instabilities that func-
often leveraged in compliant integrated circuits for the high tion as electronic logic gates able to perform digital computations
conductivity and reversible self-healing behavior realized by on mechanical stress input combinations. Here, we exploit
liquid metal-based circuit interconnects6–8. Two-phase materials reversible compact-to-deployed state transitions in conductive
that combine conductive microparticles with a polymer substrate mechanical metamaterials through a strategic patterning of
also support electrical function when subjected to large strains, compliant substrate and soft, conductive material networks. We
while moreover being amenable to diverse fabrication demonstrate physically reconfigurable metamaterials that
practices9–12. To design the substrate for reconfiguration of a respond to mechanical stress in discrete modes leading to changes
compliant integrated circuit, programmable mechanical defor- in the compliant conductor network to realize all digital logic
mation is often a candidate for reversible shape change3,4. For functions and logic assemblies.
instance, kirigami13 and origami14 have inspired substrates that
fold to tailor electrical behavior15. Elastic instabilities are also
considered to transition electrical states in response to Results
mechanical16 and thermal17 stresses. A conductive mechanical metamaterial digital switch. The
Reconfiguration of electrical networks is indeed the basis of metamaterials employ a constant cross-section consisting of a
digital logic, which is integral for information processing in fundamental square tiling tessellation with square voids at the
modern computers as well as in the human brain18. The search intersections. Shim et al.32 report that this geometry exhibits a
for logic-based information processing in artificial materials has negative Poisson’s ratio and a fully compacted state wherein the
led to concepts of discrete transmittance outputs resulting from porosity of the compacted material is 0% in the ideal case of
elastic wave logic operations in mechanical metamaterials19 and kinematic reconfiguration. In Fig. 1, we integrate this substrate
from infrared wave logic operations in photonic metamaterials20. architecture with conductive ink patterns that leverage the
Logic functions in soft matter have also been emulated by high deformation of the metamaterial to reconfigure electrical circuits
contrast colorimetric outputs from thermochromic elastomers in discrete modes. Through this report, we build upon such
subjected to input pressure states21. fundamental building block to formulate all digital logic gates and
Recent progress on embodiments of mechanologic provide to establish gate assembly and interconnection methods.
another modality of information processing in soft matter. By the The elastomeric material substrates considered here are
concept of mechanologic, a digital bit is abstracted as a reversible, fabricated by casting liquid urethane rubber (Smooth-On
mechanical, or material configuration22,23. Recently, discrete VytaFlex 60) in 3D-printed molds (FlashForge Creator Pro)
shape reconfigurations are cultivated by elastic beam containing the negative of the sample architecture. The substrates
buckling24,25. In formulations of buckling-based mechanologic, also include surface channels in which the conductive networks
the two statically stable configurations represent the physical are applied, Fig. 1(a). The conductive ink utilized in the channels
abstraction of discrete digital bits. Together, the digitized bits is a composite containing silver (Ag) microflakes (Inframat
associated with mechanical buckling modes facilitate logic Advanced Materials, 47MR-10F) and thermoplastic polyurethane
operations according to the design and assembly of the switch- (TPU) elastomer (BASF Elastollan Soft 35 A) to create conductive
able constituents26. This stratagem has been shown for percolating networks. Copper wire leads that pass through the
mechanically27, chemically28, and humidity29 triggered elastic specimen via internal channels allow for electrical connections
instabilities. Controller signaling for pressurized soft robots has from an external voltage source or readout to the Ag-TPU trace
also been demonstrated through mechanologic-based signal terminals. Complete sample fabrication and characterization
processing, suggesting one means for soft matter autonomy30. details are given in the Supplementary Information.
Yet, current embodiments of logic processing of mechanical stress The metamaterials studied in this report are referred to by the
inputs lack digital electrical outputs, which limits means to unit cell that constructs the constant cross-section architecture. In
communicate with actuation and sensory mechanisms that may Fig. 1(a), the metamaterial is assembled from C2 unit cells to be
require electrical feedback to function26,27. Similar to anatomical characterized more fully in the subsequent section of this report.
realizations of intelligence and information processing to monitor An Ag-TPU trace pattern is applied to the elastomeric substrate.

Fig. 1 Compaction principle of electrical network switching in soft, conductive mechanical metamaterials. a Introduction of a metamaterial composed of
C2 unit cell with conductive Ag-TPU trace. b C2 metamaterial in series with a power source and LED array to illustrate switching functionality.
Uncompressed: Open circuit, LED off. Compacted: Closed circuit, LED on.

2 NATURE COMMUNICATIONS | (2021)12:1633 | https://doi.org/10.1038/s41467-021-21920-y | www.nature.com/naturecommunications

NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-21920-y ARTICLE

Fig. 2 Bit abstraction via open/closed circuit. a C2 unit cell dimensions with periodic boundary conditions. b Simulated mechanical response of C2 unit cell
featuring three regimes of general behavior and shape change. c Schematic of C2 metamaterial with Ag-TPU biaxial switch network. Resistance
measurements R1 and R2 are monitored across two terminal pairs. d Experimental resistance measurement with photos depicting the deformations in the
three compression regimes.

An electrical power supply and LED array are interfaced in series the self-contact strain is εcontact = 0.23. The compact state in Fig. 2
with the silver trace terminals at the sample leftmost and (b) represents the compacted shape of the experimental sample in
rightmost sides, respectively. By applying uniaxial displacement Fig. 1(c) that closes the electrical circuit by self-contact of the Ag-
to the top of the metamaterial, the sample deforms into a fully TPU trace.
compacted solid square Fig. 1(b). As a result, the Ag-TPU To demonstrate the relationship between the mechanical and
network closes through self-contact, passing electrical current electrical behavior by the compaction principle, we design a
through the LED array. Once the mechanical load is released, the biaxially functional electrical switch Ag-TPU trace pattern on the
Ag-TPU network opens and the LED array turns off again. surface of a metamaterial composed of C2 unit cells, Fig. 2(c). The
Continued cycling of the metamaterial repeats the electrical metamaterial is loaded quasi-statically in the vertical direction
switch behavior. The principle of reversible mechanical-electrical between rigid aluminum platens using a load frame (see
switching demonstrated in Fig. 1 is harnessed in this report to Supplementary Information). Voltage divider circuits measure
formulate all logic gate operations and logic assembly methods in the electrical resistances R1 and R2 across the horizontal and
a class of soft, conductive mechanical metamaterials. vertical Ag-TPU terminal pairs, respectively.
To characterize the elastic buckling deformation of the Using the C2 unit cell geometry and the Ag-TPU network on
metamaterial unit cells, we develop a finite element model that the sample, the metamaterial functions as a strain gated switch.
simulates the material response observed in experiments (see For instance, Fig. 2(d) shows that the resistances R1 and R2
Supplementary Information). The C2 unit cell geometry is shown through the Ag-TPU trace pairs exhibit a nonconductive, infinite
in Fig. 2(a). The unit cell consists of two rows and two columns of resistance that represents a 0 digital readout. Once the contact
a continuous square tiling pattern. A hyperelastic Neo–Hookean strain εcontact = 0.23 is reached, the networks close as observed by
material model is employed on the 2D cross-section, considering the sudden drops in R1 and R2 to mean resistances near 3 Ω. The
plane strain assumptions and accounting for nonlinear deforma- closed electrical connections therefore represent a 1 digital
tions under quasi-static, uniaxial compressive displacement in the readout. A supporting kinematic model of the C2 unit cell
vertical direction uy . Periodic boundary condition displacements confirms self-contact is induced near applied strain of εcontact =
in x and y are applied on the nodes highlighted with the dashed 0.201, Supplementary Fig. 2(a) (see Supplementary Information).
blue and red lines in Fig. 2(a). The similarity between the sudden changes in resistances R1 and
The mechanical response for the C2 unit cell illustrates three R2 illustrates that the electrical connections may be achieved
regimes during uniaxial compression, Fig. 2(b). Linear elastic biaxially despite uniaxial mechanical input. These results
response (label S1) occurs for small applied strains εy ¼ uy =h, introduce a technique to exploit transitions between uncom-
such as <2% strain. At a critical strain, namely εcritical = 0.0279 in pressed and compacted states of soft, conductive mechanical
Fig. 2(b), a second regime (label S2) displays large decrease in metamaterials to govern digital electrical signals.
uniaxial stiffness and a rotational behavior of the unit cell. With
further increase in strain, a final transition occurs wherein self- Method for digital logic in conductive mechanical metamaterials.
contact and near-total compaction of the unit cell is achieved We leverage this foundation to create logic gates controlled by
(label S3), leading to increase in uniaxial stiffness. For unit cell C2, mechanical deformation. Metamaterials composed from C2 unit

NATURE COMMUNICATIONS | (2021)12:1633 | https://doi.org/10.1038/s41467-021-21920-y | www.nature.com/naturecommunications 3

ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-21920-y

Fig. 3 Bit abstraction via buckling direction. a D1 unit cell dimensions with boundary conditions. b Modal analysis of D1 unit cell with two deformation
states, mode 1(1) and mode 1(2). c Simulated mechanical response of D1 unit cell featuring three deformation regimes. d Schematic of Buffer gate as direct
electrical switch governed by a single mechanical rotation input, corresponding to the buckling mode. e Schematic and (f) experimental image of the NOT
gate with one mechanical input rotation (green) and one digital output terminal (red) QNOT on the metamaterial composed of D1 unit cells. The powered
input terminal is the cyan node, Vcc . Photos of the NOT gate under uniaxial compression as it exhibits buckling (g) mode 1(1) and (h) mode 1(2), along with
the corresponding digital outputs QNOT (red node).

cells may exhibit two rotations in the lowest order buckling mode, input node is highlighted in cyan color and labeled as Vcc on the
Supplementary Fig. 3. This limited switch behavior is a result of the NOT logic gate material shown in Fig. 3(d). The output node is
distinct buckling modes and the choice of Ag-TPU network. As a highlighted in red color. To provide the digital logic operations,
result, we introduce a second unit cell termed D1 that exploits a we exploit the mechanical buckling modes, considering counter-
discontinuity with laterally adjacent unit cells in the left/right clockwise rotation of the D1 unit cell as a digital input of 1 and
directions, Fig. 3(a). clockwise rotation as a digital input of 0, according to
A modal analysis of the D1 unit cell is carried out to electrical conduction through the D1 unit cell layer. The
characterize the lowest order elastic buckling behavior. The unit fundamental Buffer gate exemplifies this digitizing behavior as
cell top and bottom boundaries are relatively displaced with one seen in Fig. 3(d), where counterclockwise rotational collapse of
side free to laterally displace in a given simulation. As shown in the D1 unit cell metamaterial outputs QBuffer of 1, and QBuffer of 0
Fig. 3(b), the D1 unit cell exhibits a buckling mode at a critical for clockwise rotation. The Buffer metamaterial gate is realized by
strain of 0.0089. Two states of deformation may occur for the three D1 unit cells that are tiled together, keeping the
same critical strain. Mode 1(1) exhibits a clockwise rotation of the discontinuity at the central bulk material block. The Ag-TPU
central bulk material block, while mode 1(2) exhibits a counter- trace applied to the surface realizes a switch that operates in
clockwise rotation. Realization of these modes is dependent upon agreement to Buffer logic in accordance with buckling modes of
the vector of shear and uniaxial compression applied respecting this metamaterial.
the horizontal mirror plane. Uniaxial compression simulations The NOT digital logic gate is an inverter of a digital signal,
are conducted to determine the mechanical properties of the D1 converting a 0 to a 1, and vice versa. The NOT gate is thus an
unit cell. The results in Fig. 3(c) likewise reveal 3 collapse regimes inverted Buffer gate. In this embodiment of soft digital logic,
similar in qualitative trend and interpretation with behavior of output signal inversion is achieved by a horizontal or vertical
the C2 unit cell in Fig. 2(b). The D1 unit cell exhibits a lower mirroring of the conductive network. The metamaterial realiza-
critical strain εcritical = 0.0089 and a greater self-contact strain tion of the NOT logic gate that we explore here is shown in Fig. 3
εcontact = 0.40 compared with the C2 unit cell. A noticeable (e) along with the corresponding logic outputs resulting from
difference is also evident in the compacted states. Namely, the D1 each of the clockwise and counterclockwise collapse rotations.
unit cell compacts into a parallelogram cross-section Fig. 3(c), The output digital signal QNOT is 1 for the conductive Ag-TPU
compared with the square compacted state for the C2 unit cell network in Fig. 3(g) considering the clockwise rotation of the
Fig. 2(b). Consequently, the metamaterial unit cells permit unique central bulk material blocks in the unit cells. Conversely, an
approaches to Ag-TPU trace networking to realize reconfigurable output digital signal QNOT of 0 is read out for the buckling mode
electrical interconnects. with counterclockwise bulk material block rotation, Fig. 3(h). As
Digital logic gates are provided with digital inputs and pass expected, these results are the opposite of the Buffer gate
digital outputs. Our formulation of digital logic intrinsically fundamental switch. Consequently, the digital logic realized by
couples discrete mechanical behavior with electrical signals the discrete, modal mechanical deformation of the metamaterial
passed through the metamaterials. Here, we consider a circuit in Fig. 3(f) is analogous to a digital NOT gate. Moreover, Fig. 3
powering node provided to an input terminal of the unit cell. The (g), (h) show that the collapse behavior and thus digital signaling

4 NATURE COMMUNICATIONS | (2021)12:1633 | https://doi.org/10.1038/s41467-021-21920-y | www.nature.com/naturecommunications

NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-21920-y ARTICLE

Fig. 4 Formulation of digital logic gates in soft, conductive mechanical metamaterials. a Schematic of the D1 based metamaterials. Inputs A and B are the
rotations in the first and second layers, respectively. b Modal analysis illustrating four possible deformation states. Schematics and experimental images of
the Ag-TPU network on the metamaterials for (c) AND, (d) NAND, (e) OR, (f) NOR, (g) XOR, and (h) XNOR logic gates. AND and OR include
corresponding switched circuit schematics. Each gate contains two mechanical inputs A and B (green) and one digital output Q (red). The powered input
terminal is the cyan node, Vcc .

occurs for the same values of applied strain. This indicates that inputs through stress vectors and discrete self-contact states in
exploitation of small perturbations around the bifurcation33, or soft mechanical metamaterials35.
critical strain, could be used for high sensitivity control34 of the The Ag-TPU trace networks required to realize the AND,
combined mechanical and electrical functions of the soft, NAND, OR, NOR, XOR, and XNOR gates in this formulation of
conductive logic gate. soft digital logic are shown in Fig. 4(c)–(h). By using the
fundamental Buffer or NOT logic gate Fig. 3(d) for each
mechanical input layer and by electrically networking the outputs
Formulation of all digital logic gates. We build on this mani- of each layer in series or parallel, the remaining AND, NAND,
festation of soft, conductive matter-based digital logic to realize OR, NOR, XOR, and XNOR logic gates may be realized in a
the remaining 6 logic gates: AND, NAND, OR, NOR, XOR, and systematic way. For the AND and OR gates, the corresponding
XNOR. We use two rows of D1 unit cells each containing 5 unit serial and parallel switched circuit schematics are shown in Fig. 4
cells to realize the metamaterial geometries shown in Fig. 4(a). (c), (e). The relations between the serial (AND) and parallel (OR)
The unit cells share the central row of bulk material blocks, which switch assembly schematics and conductive trace networks are
does not influence the layer-by-layer mechanics. We undertake a apparent considering Fig. 4(c), (e). Based on the induced mode in
modal analysis of the materials to identify low order buckling Fig. 4(b), the switches open or close according to (counter)
modes. Two modes are found for critical strains 0.0087 and clockwise assignment of (1)0 digital bits via the Ag-TPU trace
0.0089, while each mode may be manifest in two deformation networks to result in the appropriate digital output, Q. For
states, Fig. 4(b). These modes are realized by the specific vector instance, the AND gate registers QAND = 1 when both layers
combination of shear and uniaxial stresses applied to the top rotate counterclockwise, or 1. Thus, a clear analogy exists between
metamaterial surface. For instance, mode 1(1) requires shear to the circuitry schematic, the Ag-TPU network topology, and the
the right and then uniaxial compression, while mode 2(1) requires mechanical buckling deformations. The Ag-TPU trace networks
uniaxial compression then shear to the left near the critical strain, shown in Fig. 4 are not unique and can be tailored to permit
and vice versa for the mode 1(2) and mode 2(2). Thus, the logic changing positions of the powered node (highlighted in cyan
inputs are represented by the layer-by-layer rotations, illustrated color) and output node (red color). The reading from the output
in Fig. 4(a) with green labels. Input A of the soft, conductive logic nodes (red color) represent the digital output from the soft,
gate indicates the rotation of the top layer of unit cells, while conductive logic gate.
Input B indicates the rotation of the bottom layer of unit cells. Gates such as the NAND in Fig. 4(d) and OR in Fig. 4(e)
This introduces a unique technique to control mechanical digital contain the same fundamental parallel circuit design, yet are

NATURE COMMUNICATIONS | (2021)12:1633 | https://doi.org/10.1038/s41467-021-21920-y | www.nature.com/naturecommunications 5

ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-21920-y

Fig. 5 Processing all digital logic operations in soft, conductive mechanical metamaterials. Logic gate digital output Boolean response at each of the four
compact states for the (a) AND, (b) NAND, (c) OR, (d) NOR, (e) XOR, and (f) XNOR logic gate. The respective mechanical inputs (green) for each of the
deformation states directly relate to (g) the logic gates truth table inputs.

vertically mirrored according to the horizontal midplane. The a digital output QAND = 1 for mode 1(2) when both mechanical
NAND network topology is obtained by applying De Morgan’s inputs are counterclockwise rotations, and a digital output of
theorem, which states that the NAND gate is equivalent to an QAND = 0 for all other combinations. The Supplementary Video 1
inverted OR gate. As described above related the NOT to the shows the transitions among the digitized, modal inputs, and the
Buffer, the inversion operation in this embodiment of soft digital digital outputs, with an LED array indicator for visualization.
logic gates is a vertical or horizontal mirroring of the conductive
trace network. As a result, De Morgan’s theorem is readily
applied to create the NAND metamaterial gate Fig. 4(d) via the Discussion
base construct of the OR metamaterial gate Fig. 4(e). DeMorgan’s This foundation is extensible to the formulation of logic gate
theorem is also applied to design the NOR gate Fig. 4(f) from the combinations used in integrated circuits. For instance, Supple-
embodiment of the AND gate Fig. 4(c). Finally, the XOR mentary Fig. 9 demonstrates a three-input logic gate combination
metamaterial gate Fig. 4(g) and XNOR gate Fig. 4(h) are obtained of OR-NAND on a three layer metamaterial. To create the OR-
through a serial combination of OR-NAND gates and parallel NAND network topology, a negated OR gate Fig. 4(e) interfaces
combination of AND-NOR gates, respectively. Such design with the NAND gate parallel switch Fig. 4(d). Interestingly, by the
protocol is directly analogous to the formulation of conventional unique D1 unit cell design explored here, metamaterial gates with
switched electrical networks that provide digital logic n layers possess 2n self-contacting buckling modes, in agreement
functionality. with the number of binary input combinations for the n input
To validate the functioning of the logic gates, samples are gate assembly. This differentiating characteristic allows for a
fabricated and compressed using the experimental setup in systematic metamaterial design process to generate arbitrary gate
Supplementary Fig. 1(b). Each of the buckling modes is achieved combinations via strategic conductive trace network topologies
through cyclic loading and unloading. Figure 5 compiles (see Supplementary Information), although the platforms
photographs of deformed modes, resulting Ag-TPU network are not as much assembled from modules as they are designed
connectivity, and associated truth tables. into a monolithic metamaterial circuit. Explorations of
The experimental digital outputs for each buckling mode modularly assembled soft logic gates using our mechanical
behavior agree with the respective truth table entry for the ideal metamaterial modules may take inspiration from mechanical
logic gates, Fig. 5. For instance, the AND gate in Fig. 5(a) provides signal transmission concepts26,27 as well as techniques explored

6 NATURE COMMUNICATIONS | (2021)12:1633 | https://doi.org/10.1038/s41467-021-21920-y | www.nature.com/naturecommunications

NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-21920-y ARTICLE

with pneumatic-mechanical realizations of soft logic30. Although and plate. The uniaxial applied strain is calculated from εy ¼ uy =h where h is the
the logic gates formulated here do not compete with the speed macroscopic height of the metamaterial sample or unit cell.
and component scaling of conventional digital microprocessors, For metamaterial D1, the bottom surface is presumed to slide in the simulations
reported in the main text. To permit lateral sliding of metamaterials in the
the conductive metamaterial-based information processing may experimental characterization, we place a polished, rigid aluminum plate on
provide essential robustness of function and material compat- smooth cylindrical shafts that act as a rolling surface for the bottom of the D1
ibility for future autonomous soft matter that emulate numerous samples, Supplementary Fig. 1(b). Such a setup provides uninhibited lateral
biological systems. displacements ux . The direction of cross-section rotation in experiments may be
governed at critical points by manual methods in order to achieve all buckling
By establishing interfaces between soft mechanical metama- modes, since only the lowest order mode occurs naturally in the absence of
terials and digital logic gating, this work formulates a general additional control measures. Voltage divider circuits are used to measure resistance
method to synthesize discrete mechanical configurations with through the Ag-TPU ink traces with a 1.30 Ohm reference resistor, where the
conductive networks to govern digital electrical outputs. As one electrical resistances R1 and R2 are evaluated across the horizontal and vertical Ag-
manifestation of soft, smart matter36, the mechanical metama- TPU terminal pairs, respectively. Data is collected through an acquisition system
(NI USB 6341 Multifunction DAQ) and analyzed in MATLAB.
terial building blocks exploited in this work may be considered as
building blocks for future soft matter assemblies, such as to Finite element modeling methods. ABAQUS Linear Buckling Perturbation and
realize autonomous soft machines with integrated sensory, Dynamic-Implicit simulations are conducted with 2D plane strain models based on
actuation, and decision-making functionalities37. The concepts the constant material cross section. For the Linear Buckling Perturbation simula-
embodied in this report are also free of length-scale dependence, tions, a subspace eigensolver is used to determine the low-order buckling modes.
by leveraging nominally kinematic rearrangement of cellular For the nonlinear Dynamic-Implicit step, a time period that is proportional to the
0.5 mm.min−1 quasi-static experimental load frame loading rate is used. A
solids32, and suggest one means to process mechanical stress hyperelastic Neo–Hookean material model is employed with Young’s modulus of
inputs without need for periodic sensor observation. Conse- 2.07 MPa and Poisson’s ratio of 0.499 for the samples considered here. Free
quently, physical embodiments of these principles may be quadrilateral elements are used for all geometry meshes with seed sizes propor-
exploited for low-power information processing in fields as tional to the smallest geometric feature with at least 0.25 ratio of element char-
acteristic dimension to substrate feature size.
diverse as DNA origami38, electro-optical filters39, reconfigurable For the C2 unit cell, periodic boundary conditions are imposed on the
antenna40, and more. outermost nodes while the bottom boundaries are fixed in y and free to slide in x.
Shown in Fig. 2(a) in the main text, the left-most and right-most boundary nodes
(outlined in red) are constrained to mirror the displacements of nodes in x and y,
Methods and similarly the top-most and bottom-most boundary nodes (outlined in blue) are
Specimen fabrication methods constrained to mirror displacements only in x. For the D1 unit cell and
Metamaterial substrate fabrication. The elastomeric metamaterial substrates are metamaterial assemblies the boundaries are fixed in y and free to slide in x at the
fabricated by casting liquid urethane rubber (Smooth-On VytaFlex 60) in a two- bottom edge.
part mold. The mold parts are designed in CAD software SOLIDWORKS and 3D- By applying compressive displacements uy at the top boundary of the unit cells
printed (FlashForge Creator Pro) with acrylonitrile butadiene styrene (ABS) fila- and metamaterial assemblies, we acquire the buckling modes as well as the
ment. When assembled together, the two-part mold realizes the negative of the mechanical properties through the Linear Perturbation and Dynamic-Implicit
metamaterial substrate shape. The liquid urethane rubber utilized in this research is simulations, respectively.
a two-part material, A and B parts that are initially mixed in a 1 A:1 B volume ratio
and stirred by hand for 2 min. After the material is poured into the mold, it is cured
Statistics and reproducibility. All experiments and simulations are conducted a
for 24 h. The sample is then carefully demolded and prepared for testing. Meta-
minimum of 6 times to confirm repeatability of trends and quantities.
materials using C2 and D1 unit cells are prepared utilizing this procedure.

Conductive ink fabrication and deposition. The conductive ink utilized in the Data availability
channels is a composite containing 35% (volume %, v%) silver (Ag) microflakes Models, codes, and datasets generated during and/or analyzed during the current study
(Inframat Advanced Materials, 47MR-10F) and 65% (v%) thermoplastic poly- are available from the corresponding author on request. All data is available in the main
urethane (TPU) elastomer (BASF Elastollan Soft 35 A). To begin processing the text or the Supplementary Information.
conductive ink, Ag microflakes are first mixed in a glass vial with sufficient N-
Methyl-2-pyrrolidone (NMP) solvent, and sonicated (Branson M2800 Ultrasonic
Cleaner) for 60 min. TPU is then added to the Ag-NMP mixture, and planetary Received: 7 August 2020; Accepted: 4 February 2021;
mixed (KK 300SS Mazerustar) at 225 ´ g for 2-minute increments. The planetary
mixing process is repeated three times with gentle hand stirring in between to
ensure walls of the vial do not collect excess Ag or TPU particles. After the mixture
is given 48 h for the NMP to evaporate at room temperature, the Ag-TPU ink is
ready for application.
Enamel-coated copper wire (22 gauge) leads are passed through small molded- References
in 1.50 mm diameter channels in the substrate to terminate at the front cross- 1. Harris, K. D., Elias, A. L. & Chung, H.-J. Flexible electronics under strain: a
sections of the metamaterials. Metamaterial photographs in Figs. 1 and 3 of the review of mechanical characterization and durability enhancement strategies.
main text and in the Supplementary Video 1 show more clearly how the copper J. Mater. Sci. 51, 2771–2805 (2016).
wires terminate at the front cross-sections. Enamel is only removed from both 2. Rogers, J. A., Someya, T. & Huang, Y. Materials and mechanics for stretchable
extremities of the copper wire ends to eliminate trace electrical shorting along the electronics. Science 327, 1603–1607 (2010).
wire length. To secure the wire leads in the channels through the substrates, a small 3. Zhu, J., Dexheimer, M. & Cheng, H. Reconfigurable systems for
amount of silicone adhesive (DAP All-Purpose) is deposited by syringe in the multifunctional electronics. npj Flex. Electron. 1, 206–214 (2017).
internal channels through the rear-facing cross-section. Using a 3.0 cc dispensing 4. Nassar, J. M., Rojas, J. P., Hussain, A. M. & Hussain, M. M. From stretchable
syringe, the surface channels on the substrate are filled with Ag-TPU ink and
to reconfigurable inorganic electronics. Extrem. Mech. Lett. 9, 245–268 (2016).
allowed to cure around the copper wire leads for 24 h.
5. Huang, S., Liu, Y., Zhao, Y., Ren, Z. & Guo, C. F. Flexible electronics:
stretchable electrodes and their future. Adv. Funct. Mater. 29, 1805924 (2019).
6. Palleau, E., Reece, S., Desai, S. C., Smith, M. E. & Dickey, M. D. Self‐healing
Experimental characterization methods. The unit cells C2 and D1 and meta-
stretchable wires for reconfigurable circuit wiring and 3D microfluidics. Adv.
materials assembled from the unit cells are examined using distinct experimental
Mater. 25, 1589–1592 (2013).
methods by virtue of the unique mechanical behaviors that motivate distinct
electrical circuit switching methods. The specimens are quasi-statically and verti- 7. Ilyas, N., Cook, A. & Tabor, C. E. Designing liquid metal interfaces to enable
cally displaced uy by a polished, rigid platen from a load frame (ADMET eXpert next generation flexible and reconfigurable electronics. Adv. Mater. Interfaces
4, 1700141 (2017).
5600) at a loading rate of 0.5 mm min−1, shown in Supplementary Fig. 1. A laser
8. Markvicka, E. J., Barlett, M. D., Huang, X. & Majidi, C. An autonomously
displacement sensor (Micro-Epsilon optoNCDT ILD1700-200) is attached to the
electrically self-healing liquid metal–elastomer composite for robust soft-
load frame to measure vertical displacements uy of the platen. For metamaterial C2,
matter robotics and electronics. Nat. Mater. 17, 618–624 (2018).
the bottom specimen surface rests on a polished, fixed aluminum plate, Supple- 9. Valentine, A. D. et al. Hybrid 3D printing of soft electronics. Adv. Mater. 29,
mentary Fig. 1(a). As a result, for C2 the top and bottom surfaces of the specimen
1703817 (2017).
have limited motion in lateral displacements ux due to friction against the platen

NATURE COMMUNICATIONS | (2021)12:1633 | https://doi.org/10.1038/s41467-021-21920-y | www.nature.com/naturecommunications 7

ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-21920-y

10. Sears, N. C., Berrigan, J. D., Buskohl, P. R. & Harne, R. L. Flexible hybrid 35. El-Helou, C. & Harne, R. L. Exploiting functionally graded elastomeric
electronic material systems with programmable strain sensing architectures. materials to program collapse and mechanical properties. Adv. Eng. Mater. 21,
Adv. Funct. Mater. 20, 1800499 (2018). 1900807 (2019).
11. Merilampi, S., Laine-Ma, T. & Ruuskanen, P. The characterization of 36. Pishvar, M. & Harne, R. L. Foundations for soft, smart matter by active
electrically conductive silver ink patterns on flexible substrates. Microelectron. mechanical metamaterials. Adv. Sci. 7, 2001384 (2020).
Reliab. 49, 782–790 (2009). 37. Majidi, C. Soft-matter engineering for soft robotics. Adv. Mater. Technol. 4,
12. Sears, N. C., Berrigan, J. D., Buskohl, P. R. & Harne, R. L. Dynamic response 1800477 (2019).
of flexible hybrid electronic material systems. Composite Struct. 208, 377–384 38. Marras, A. E., Zhou, L., Su, H. J. & Castro, C. E. Programmable motion of
(2019). DNA origami mechanisms. Proc. Natl Acad. Sci. 112, 713–718 (2015).
13. Zhang, Y. et al. A mechanically driven form of Kirigami as a route to 3D 39. Zhang, Y. et al. One-step nanoscale assembly of complex structures via
mesostructures in micro/nanomembranes. Proc. Natl Acad. Sci. 112, harnessing of an elastic instability. Nano Lett. 8, 1192–1196 (2008).
11757–11764 (2015). 40. Kim, H. et al. Responsive, 3D electronics enabled by liquid crystal elastomer
14. Rogers, J., Huang, Y., Schmidt, O. G. & Gracias, D. H. Origami mems and substrates. ACS Appl. Mater. Interfaces 11, 19506–19513 (2019).
nems. MRS Bull. 41, 123–129 (2016).
15. Xu, B., Chen, D. & Hayward, R. C. Mechanically gated electrical switches by
creasing of patterned metal/elastomer bilayer films. Adv. Mater. 26, Acknowledgements
4381–4385 (2014). Funding: This research is supported in part by a U.S. Air Force Research Lab (AFRL)
16. Nick, Z. H., Tabor, C. E. & Harne, R. L. Liquid metal microchannels as digital Summer Faculty Fellowship and in part by the National Science Foundation (NSF)
sensors in mechanical metamaterials. Extrem. Mech. Lett. 40, 100871 (2020). Faculty Early Career Development Award (No. 2054970).
17. Heo, H., Li, S., Bao, H. & Ju, J. A passive thermal switch with kirigami-inspired
mesostructures. Adv. Eng. Mater. 21, 1900225 (2019). Author contributions
18. Xie, K. et al. Brain computation is organized via power-of-two-based C.E.H. and R.L.H. designed the research; C.E.H. performed research; C.E.H., P.R.B., C.E.
permutation logic. Front. Syst. Neurosci. 10, 95 (2016). T., and R.L.H. analyzed the data; and C.E.H., P.R.B., C.E.T., and R.L.H. wrote the paper.
19. Bilal, O. R., Foehr, A. & Daraio, C. Bistable metamaterial for switching and
cascading elastic vibrations. Proc. Natl Acad. Sci. 114, 4603–4606 (2017).
20. Dhasarathan, V., Sahu, S. K., Nguyen, T. K. & Palai, G. Realization of all logic Competing interests
gates using metamaterials based three dimensional photonics structures: a The authors declare no competing interests.
future application of 3D photonics to optical computing. Opt. - Int. J. Light
Electron Opt. 202, 163723 (2020).
21. Jin, Y. et al. Materials tactile logic via innervated soft thermochromic
Additional information
Supplementary information The online version contains supplementary material
elastomers. Nat. Commun. 10, 4187 (2019).
available at https://doi.org/10.1038/s41467-021-21920-y.
22. Reif, J. H., Computational Complexity: Theory, Techniques, and Applications
(Springer, New York, 2012), pp. 1821-1836.
Correspondence and requests for materials should be addressed to R.L.H.
23. Merkle, R. C. Two types of mechanical reversible logic. Nanotechnology 4,
114–131 (1993). Peer review information Nature Communications thanks Andrew Pascall and the other,
24. Harne, R. L. & Wang, K. W., Harnessing Bistable Structural Dynamics: for anonymous, reviewer(s) for their contribution to the peer review of this work. Peer
Vibration Control, Energy Harvesting and Sensing (John Wiley & Sons Ltd., reviewer reports are available.
Chichester, 2017).
25. Kochmann, D. M. & Bertoldi, K. Exploiting microstructural instabilities in Reprints and permission information is available at http://www.nature.com/reprints
solids and structures: from metamaterials to structural transitions. Applied
Mechanics Reviews, 050801 (2017).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
26. Song, Y. et al. Additively manufacturable micro-mechanical logic gate. Nat.
published maps and institutional affiliations.
Commun. 10, 1–6 (2019).
27. Ion, A., Wall, L., Kovacs, R. & Baudisch, P. Digital Mechanical Metamaterials,
presented at Proceedings of the 2017 CHI Conference on Human Factors in
Computing Systems (New York, 2017). Open Access This article is licensed under a Creative Commons
28. Jiang, Y., Korpas, L. M. & Raney, J. R. Bifurcation-based embodied logic and Attribution 4.0 International License, which permits use, sharing,
autonomous actuation. Nat. Commun. 10, 128 (2019). adaptation, distribution and reproduction in any medium or format, as long as you give
29. Treml, B., Gillman, A., Buskohl, P. & Vaia, R. Origami mechanologic. Proc. appropriate credit to the original author(s) and the source, provide a link to the Creative
Natl Acad. Sci. 115, 6916–6921 (2018). Commons license, and indicate if changes were made. The images or other third party
30. Preston, D. J. et al. Digital logic for soft devices. Proc. Natl Acad. Sci. 116, material in this article are included in the article’s Creative Commons license, unless
7750–7759 (2019). indicated otherwise in a credit line to the material. If material is not included in the
31. Tortora, G. J. & Derrickson, B. H. Principles of Anatomy and Physiology (John article’s Creative Commons license and your intended use is not permitted by statutory
Wiley and Sons, Inc., New York, 2006). regulation or exceeds the permitted use, you will need to obtain permission directly from
32. Shim, J. et al. Harnessing instabilities for design of soft reconfigurable auxetic/ the copyright holder. To view a copy of this license, visit http://creativecommons.org/
chiral materials. Soft Matter 9, 8198–8202 (2013). licenses/by/4.0/.
33. Scheffer, M. et al. Anticipating critical transitions. Science 338, 344–348 (2012).
34. McInnes, C. R. & Waters, T. J. Reconfiguring smart structures using phase
space connections. Smart Mater. Struct. 17, 025030 (2008). © The Author(s) 2021

8 NATURE COMMUNICATIONS | (2021)12:1633 | https://doi.org/10.1038/s41467-021-21920-y | www.nature.com/naturecommunications