Transient electronics represents an emerging technology whose defining feature is an ability to dissolve, disintegrate or otherwise physically disappear in a controlled manner. Envisioned applications include resorbable/degradable biomedical implants, hardware-secure memory devices, and zero-impact environmental sensors. 2D materials may have essential roles in these systems due to their unique mechanical, thermal, electrical, and optical properties. Here, we study the bioabsorption of CVD-grown monolayer MoS2, including long-term cytotoxicity and immunological biocompatibility evaluations in biofluids and tissues of live animal models. The results show that MoS2 undergoes hydrolysis slowly in aqueous solutions without adverse biological effects. We also present a class of MoS2-based bioabsorbable and multi-functional sensor for intracranial monitoring of pressure, temperature, strain, and motion in animal models. Such technology offers specific, clinically relevant roles in diagnostic/therapeutic functions during recovery from traumatic brain injury. Our findings support the broader use of 2D materials in transient electronics and qualitatively expand the design options in other areas.
Transient electronics is a class of technology defined by its ability to physically disappear or disintegrate in a controlled manner and/or at a specified time. Systems of this type can be designed to dissolve in water or biofluids, in a biocompatible and environmentally benign fashion, after a stable operating period1,2. The potential applications include medical diagnostic and therapeutic platforms that eliminate, through bioabsorption, any adverse long-term effects; environmental sensors that avoid the need for collection and recovery; and consumer devices that disintegrate with minimal processing costs and associated waste streams1,2. Recent reports describe bio/eco resorbable devices for use in transient photonics3, micromotors4, microsupercapacitors5, batteries6, triboelectric nanogenerators7, and diagnostics of electrical activity on the cerebral cortex8.
A critical challenge is in the development of constituent materials that are biocompatible and biodegradable9, yet also offer electronic grade properties as conductors, insulators, and semiconductors1,2. Recent work in this direction establishes some attractive options. Dissolvable metals include magnesium (Mg), iron (Fe), zinc (Zn), molybdenum (Mo), and tungsten (W)10. Dielectric polymers include poly(vinyl alcohol), polyvinylpyrrolidone, poly(lactic-co-glycolic acid) (PLGA), polylactic acid, and polycaprolactone, as well as biopolymers such as cellulose and silk. Such materials offer appealing attributes as substrates and encapsulation layers11. Inorganic dielectrics such as silicon dioxide (SiO2), silicon nitride (Si3N4), and magnesium oxide (MgO) can be used as interlayers and gate insulators. Choices for semiconductors include thin films or nanostructures of inorganic and organic materials, such as silicon (Si), germanium (Ge), Si-Ge, zinc oxide (ZnO), and 5,5′-bis-(7-dodecyl-9H-fluoren-2-yl)-2,2′-bithiophene (DDFTTF) and perylene diimide12,13.
The most sophisticated transient technologies use silicon nanomembranes (Si NMs), due to an advantageous combination of transport properties and nanoscale thicknesses, the latter of which allows their complete dissolution in water or biofluids due to hydrolysis over timescales that are relevant for degradation in the environment or the body14. A disadvantage is that the dissolution rates depend strongly on local chemistry, and can be strongly affected by the pH, the concentration, and type of ions and other factors. As a result, relatively thick Si NMs (>150 nm) are typically used to ensure operating times in the range of a few weeks15. Additional limitations of Si NMs follow from their brittle mechanical properties, with tensile failure strains of less than 1%, and their optical absorption throughout the visible range16. Therefore, developing transient semiconductors that simultaneously minimize, to a fundamental level, the total material content, and maximize the mechanical robustness, the electrical performance characteristics and the optical transparency represents an important direction for research in this emerging field.
Recent interest in two-dimensional (2D) transition-metal dichalcogenides derives from a set of unique electrical, optical, thermal, and mechanical properties that makes them attractive candidates as active materials in compact and lightweight integrated electronic systems17,18. Among these materials, molybdenum disulfide (MoS2) is one of the most important and extensively studied. In general, 2D MoS2 has several advantages19. First, the 2D confinement of electrons, especially in the case of monolayer MoS2, imparts ideal properties as a channel material for high-performance electronic or optoelectronic devices20. Second, its strong in-plane covalent bonding and atomic layer thickness yield excellent mechanical strength (breaking strain >2.2%), flexibility, and optical transparency; this collection of properties is important for applications in transparent, bendable devices21. High-throughput synthesis methods, such as those based on chemical vapor deposition (CVD) and metal-organic CVD, allow growth of large-scale MoS2 atomic layers and establishing pathways for their integration into practical systems22,23,24,25. Relevant properties will further be affected by its grain size due to the different defect densities associated with grain boundary and point defect17,23. Although many of these features are potentially crucial in transient electronics, little is known about the processes of biodegradation, in vivo and in vitro cytotoxicity, and immune responses associated with large-scale monolayer MoS2. In addition, preliminary work with exfoliated MoS2 nanosheets are not readily useful for this type of electronic devices and systems26,27,28,29.
Herein, we report the dissolution characteristics and behavior of both isolated crystals and continuous films of CVD-grown monolayer MoS2 in phosphate buffered saline (PBS) solutions at various temperatures and with different pH levels. Reducing the grain size of monolayer MoS2 and increasing the density of intrinsic defects increase the dissolution rate in PBS solution, thereby providing a means to control the lifetime. Long-term cytotoxicity and immune monitoring tests on as-grown monolayer MoS2, on ultrasonically dispersed MoS2 flakes, and on MoS2 dissolved in PBS solutions provide a complete set of information relevant to use in implantable systems. To demonstrate some possibilities in high-performance devices, we fabricated MoS2-based bioabsorbable, multifunctional sensors with capabilities for measuring pressure, temperature, strain, and acceleration. Experiments in live animal models illustrate the function of such sensors deployed for measuring temperature in the intracranial space in mice. The results suggest that introduction of 2D semiconducting materials into transient electronic systems may lead to significant opportunities in the development of flexible, stretchable, conformal, and transparent devices and systems with minimal materials load on the environment and/or surrounding biology. Such transient devices may find roles in temporary biomedical implants and resorbable environmental monitors.
Dissolution characteristics and behavior of monolayer MoS2
To reveal the chemistry and chemical kinetics of dissolution of monolayer MoS2 in PBS solution, we first investigated isolated crystals with a small number of grain boundaries (GBs). The combination of highly crystalline structures and well-delineated grains are beneficial for the analyses presented subsequently. Monolayer MoS2 crystals with sharp triangular corners and grain size of 5–25 μm were synthesized on SiO2/Si substrates using atmospheric pressure chemical vapor deposition (APCVD) (Supplementary Fig. 1a), as described in the Methods Section. The characteristic Raman peaks are at 382 cm−1 (in-plane E2g mode) and 402 cm−1 (out-of-plane A1g mode) (Supplementary Fig. 1b). Such a small difference in frequency (20 cm−1) between the two Raman peaks is consistent with the monolayer nature of the crystals30. The photoluminescence (PL) spectra support this conclusion, as a sharp excitonic A1 peak appears at 1.82 eV, indicating a high degree of crystallinity (Supplementary Fig. 1c)31. Because the SiO2/Si substrate is chemically unstable in PBS solution8, the dissolution tests used MoS2 crystals transferred onto a sapphire substrate coated with a thin layer of a photodefinable epoxy (SU-8, Microchem Corp.) with smooth surface (root mean square (RMS) roughness ~0.258 nm) to enable high-resolution atomic force microscopy (AFM).
The dissolution tests involved monolayer MoS2 crystals with the shapes of arrowheads. Second harmonic generation (SHG) nonlinear optical analysis reveals MoS2 grains with different orientations and the distribution of GBs32, as shown in Fig. 1a. Differences in SH intensity highlight three constituent domains, with GBs in between. Regions of lattice mismatch, defects and dislocations appear at the GBs, including S single (VS1) and double vacancies (VS2), and Mo dangling bonds (Fig. 1b)33,34. For the defect-free areas, the kinetic energy barrier is 1.59 eV, and the chemical stability is high at ambient conditions. A VS1 or VS2 defect on the MoS2 surface reduces the kinetic energy barrier to <0.8 eV, thereby creating a local area with increased reactivity, commonly passivated with oxygen35,36. Upon immersion in PBS solution, the defect-rich GBs have increased rates of reaction with water and ions compared to pristine surfaces.
These inferences are supported by accelerated dissolution tests of MoS2 crystals in PBS solution (pH 7.4) at 75 °C for 8 days, as determined by morphological and structural measurements every 4 days. Figure 1c–e shows optical micrographs (OM), Raman A1g intensity maps and AFM images of MoS2 crystals throughout this period. The results indicate clearly that the GB regions dissolve first and serve as the starting points of chemical reactions that extend slowly toward the crystalline regions (Fig. 1c–e, Supplementary Fig. 2). Moreover, continuous reductions in the intensities of the Raman and PL peaks in the GB regions are consistent with this overall conclusion (Fig. 1f–g). The magnified AFM images provide further support (Fig. 1h). Specifically, after 4 days, a gap with a width of ~80 nm appeared, which widened to ~180 nm after 8 days.
As is well known, MoS2 monolayer has two phases: trigonal prismatic (labeled as 2H) and octahedral (labeled as 1T). The 2H-MoS2 is relatively stable, but semiconductor with poor conductivity. The 1T-MoS2 is metastable at room temperature, but metallic material with good conductivity7. Given the bandgap at 1.81 eV in Fig. 1g, we can infer that the CVD-grown monolayer MoS2 crystals are in 2H phase with semiconducting property33. Due to the electronic structure of 2H-MoS2, the chemical bonds between Mo and S atoms are robust in an aqueous medium but can be broken by the presence of strong reducing agents, such as alkali metals (Li, Na or K)37,38,39. The existence of Na ions leads to lattice distortions of MoS2 and formation of Na2S via conversion of pristine 2H-MoS2 to 1T-NaMoS2 and finally to Na2S (see the reaction equation 1 and 2 in Supplementary Note 1)39,40. Therefore, an increase in the concentration of Na and K ions will accelerate the degradation processes. Moreover, MoS2 will be oxidized by the O2 in the PBS solution to yield MoO42−, which itself can readily dissolve (see the reaction equation 3 in Supplementary Note 1). Based on the empirical kinetic equation 4 in Supplementary Note 1, the increment in pH value caused by an increase in the concentration of OH− ions can further accelerate the reaction rate28.
Because of these combined chemical effects, after immersion for 21–35 days, the dissolved GB regions widen to 1–2 μm and localized holes begin to appear in the central areas of the crystals, likely due to reactions that initiate at point defects (VS1 or VS2) (Supplementary Fig. 3a–b). Complete dissolution occurs within 49 days (~0.016 nm day−1) (Supplementary Fig. 3c). The overall conclusion is that dissolution of monolayer MoS2 crystals in PBS occurs as a defect-induced etching process41. Due to the reduced kinetic energy barrier, the MoS2 GBs dissolve first, followed by point defects, and then regions that extend along the horizontal direction perpendicular to the broken edges, eventually merging and leading to complete dissolution. The rate of this dissolution across a single crystalline domain of MoS2 (from edge to center) is 10–20 nm day−1 in PBS solution (pH 7.4) at 75 °C. For polycrystalline monolayer MoS2, the above result suggests that the total dissolution time can be decreased with smaller grain size, thereby providing an approach for tuning the lifetime of bioabsorbable electronic devices. For this reason, investigations reported here focus on large-scale, continuous MoS2 films with polycrystalline texture and small grain size.
Low-pressure chemical vapor deposition (LPCVD) provides a scalable means for synthesizing such films (for more information on this process, please refer to the Methods Section). With a seed layer, large-scale, continuous films of MoS2 can be grown on SiO2/Si substrates (Supplementary Fig. 4a). However, such MoS2 films are typically polycrystalline, thus GBs are easily formed because of the merger of crystalline domains with different orientations. The GBs significantly affect the mechanical, optical, thermal, and electrical properties due to the associated defect distribution34. For instance, monolayer MoS2 with smaller grain size exhibits lower mobility, resulting from the extra scattering for electron transport but more uniform electrical properties23. Thus, it is feasible to tailor the properties of 2D MoS2 and its devices via defect engineering. We studied the large-scale MoS2 continuous film with an average grain size of ~200 nm, which possesses proper properties for target devices (Supplementary Fig. 4b). Compared with the isolated crystals, Raman spectrum of such continuous film also shows two characteristic peaks corresponding to E2g and A1g modes, and the frequency difference between these two modes is ~20 cm−1 (Supplementary Fig. 4c), indicating that the film is monolayer indeed24. Besides, the film demonstrates a sharp A1 peak at 1.87 eV (Supplementary Fig. 4d), suggesting a relatively high quality but small grain size that are promising for the bioabsorbable application26. X-ray photoemission spectroscopy (XPS) spectra show an S2s peak at 226.7 eV and Mo 3d peaks at 229.5 and 232.6 eV, corresponding to the 3d5/2 and 3d3/2 doublets, respectively (Supplementary Fig. 4e). Peaks at 162.4 and 163.6 eV arise from the S 2p3/2 and 2p1/2 orbitals, respectively (Supplementary Fig. 4f)24. The stoichiometric ratio of Mo to S estimated from the respective integrated peak areas of the XPS spectra is 1:1.95, consistent with the presence of the S vacancies in the film34.
Figure 2a–h shows the results of studies of dissolution of MoS2 monolayer films in PBS solution (pH 7.4) at 60 °C for 0–12 days. These figures include images obtained through OM, Raman A1g intensity mapping, AFM and the related Raman, PL, transmittance, and XPS spectra collected every 4 days. As time passes, the continuity of the MoS2 film decreases gradually and then vanishes after 12 days (~0.067 nm day−1), which is four times faster than the one of the isolated crystals (Fig. 2a). The continuous loss of Mo and S atoms during the dissolution process leads to expected changes in the intensity maps of the Raman A1g peak (Fig. 2b). The AFM images in Fig. 2c show that the film dissolves in three steps: first, the GBs dissolve, thereby isolating the grains; second, the edges of the grains continue to react, leading to a reduction in their sizes; and third, the MoS2 grains dissolve entirely. It is obvious that the total dissolution time of polycrystalline monolayer MoS2 in PBS solution is fundamentally determined by its grain size. Smaller grains need a shorter time to be etched from the edge to the center with the rate of 10-20 nm day−1. Besides, the surface roughness increases from 0.36 to 0.57 nm after 8 days, due to steps and edges that appear near the GBs. After 12 days, the RMS roughness rapidly decreases to 0.32 nm, comparable to that of the underlying layer of SU-8 (0.26 nm), suggesting full dissolution (Supplementary Fig. 5). The Raman and PL spectra are consistent with these observations; characteristic peaks for MoS2 disappear after 12 days (Fig. 2d–e). It should be noted that there is not obvious Mo-O related Raman peak located between 200–300 cm−1 (Supplementary Fig. 6a).
As the area of exposed SU-8 increases, the light transmittance and absorption increase and decrease, respectively (Fig. 2f). The concentration of Mo in PBS solution before and after monolayer MoS2 dissolution was measured by using inductively coupled plasma mass spectrometry (ICP-MS) (Supplementary Fig. 6b). Before the dissolution, Mo concentration in original PBS solution was 2.736 ppb. After one and five pieces of MoS2 monolayer samples (1 × 1 cm) dissolved in PBS solutions, the related Mo concentrations increased from 4.919 to 14.303 ppb (5 times), which means that the Mo atoms transfer from the SU-8 substrate to the PBS solution. Besides, the continued dissolution of the MoS2 samples increases the Mo concentration. We further evaluated the elemental composition and chemical state of the dissolved MoS2 film via XPS (Fig. 2g–h). The results show that the peak positions remain constant and their amplitudes decrease, indicating no changes in either the bandgap or the doping level or type42. Besides, Mo-O bonding peaks were also not observed at ~233 and ~236 eV in the XPS spectra, regardless of the dissolution time, because the intermediate, ionic molybdate tetroxide created from the oxidation of Mo spreads quickly into the PBS solution28. Other relevant peaks, such as P 2p (133.3 eV) and Na 1s (1071.7 eV), increase slightly, as a possible consequence of solute adsorption (Supplementary Fig. 6c-e). In other words, the dissolution products remain in the solution rather than on the substrate.
A series of dissolution tests performed in PBS solutions at different temperatures and pH levels reveal the effects of these parameters on the rates (Fig. 2i). Optical images of a sample of MoS2 under such different conditions appear in Supplementary Fig. 7. At pH 7.4, complete dissolution requires 3, 4, 12, and 72 days at temperatures of 85, 75, 60, and 40 °C, respectively. As the pH increases to 12 through the addition of NaOH solution (0.1 mol L−1), the dissolution time decreases to 2, 3, 5, and 56 days under the same temperatures. At 40 °C, the dissolution rates at these two pH levels are ~0.011 and ~0.014 nm day−1, which are much slower than that of Si NMs (ranging from a few nanometers to hundreds of nanometers day−1)14. Note that the degradation process slows as the temperature of PBS solution (pH 7.4) reduces to 37 °C (~99 days) or even 30 °C (~200 days), as expected in Supplementary Fig. 8. The activation energies (~105 meV at pH 7.4 and ~147 meV at pH 12) are extracted by fitting the temperature-dependent dissolving rate as shown in the Fig. 2i.14 Additional details about the fitting are provided in Supplementary Fig. 9. It should be noted that the addition of NaOH solution increases the concentration of both Na+ and OH− ions. Therefore, for this system, the observed acceleration in dissolution with increasing pH results from effects of both Na+ and OH−, per the reaction equations 2 and 4 as shown in Supplementary Note 1.
Cytotoxicity and immune monitoring tests of monolayer MoS2
The biocompatibility of the starting materials and the products of dissolution are critically important for applications in implantable electronics. Recent studies report results on the toxicity of synthetic MoS2 using assays based on salivary gland cells, A549, and HEK293f cells28,43,44,45. Here, we explore the cytotoxicity of MoS2 both before and after their dissolution, using starting materials as continuous films, isolated flakes, and as pre-dissolved in PBS solution.
Figure 3a shows fluorescence images (top) and apoptotic/necrotic cell death profiles analyzed by flow cytometry (bottom) of PC12 pheochromocytoma neuronal origin cells upon 24 h culture with none, HDPE, ZDEC, and MoS2 continuous films. While PC12 cells cultured on HDPE and MoS2 film did not show any detectible changes in morphology (top) or AnnexinV/7-AAD (7-Amino-Actinomycin D) staining (AnnexinV−/7-AAD− live, AnnexinV+/7-AAD− early apoptotic, AnnexinV+/7-AAD+ late apoptotic, and AnnexinV−/7-AAD+ necrotic cell populations), those cells cultured on ZDEC demonstrated less than 5% viable cells (AnnexinV−/7-AAD− live population). Similarly, HUVEC cells cultured on ZDEC film showed significantly high percentages of dying cells (AnnexinV+/7-AAD+ late apoptotic), while those on HDPE or MoS2 did not show any changes in viability (Fig. 3b). The results from three independent experiments were plotted as bar graphs in Fig. 3c.
We next examined the long-term biocompatibility of MoS2 films in vitro. To this end, we chose L-929 cells, mouse fibroblast cells, as they are suitable for short-term and long-term tissue biocompatibility tests. Cell growth and proliferation cultured on MoS2 atomic layers were analyzed by fluorescence imaging (Fig. 4a). Cells adhere tightly to the MoS2 and spread well during the culture period. After 24 days, most cells remain viable and occasionally form clusters, consistent with no adverse effect on cell proliferation in vitro (Supplementary Fig. 10). Analysis of cell growth in culture media with various concentrations of MoS2 flakes and end products of MoS2 dissolution in PBS solution yield additional insights (Fig. 4b-d, Supplementary Fig. 11). Fluorescence images show no obvious changes in terms of cell growth for both the MoS2 flakes and solution with dissolved MoS2, as in the results of negative control samples (e.g., HDPE). Cells exposed to PU-ZDEC, however, show damaged cell morphology, as expected for the positive control (Supplementary Fig. 12 and 13)46. Staining cells with Annexin V and 7-AAD at 4, 12, and 24 days following their culture on a MoS2 layer, reveal that the percentage of live (Annexin V−7-AAD−), early apoptotic (Annexin V+7-AAD−) and late apoptotic (Annexin V+7-AAD+) cells are comparable to those cultured without MoS2 (Supplementary Fig. 14). Furthermore, a cell counting Kit-8 (CCK-8) assay, as summarized in Fig. 4e, shows no adverse effects on the cell viability, similar to the finding for HDPE. By contrast, considerably lower cell viability (~36%) was obtained with cells exposed to PU-ZDEC, as expected for this positive control for toxicity.
Implanting samples with MoS2 layers on PET substrates subcutaneously into BALB/c mice and monitoring them for four weeks in a specific pathogen-free facility defines the long-term cytotoxicity and biocompatibility in a vital animal model. Based on the rate of dissolution summarized in Fig. 2, the implanted MoS2 will gradually dissolve into the interstitial fluid of mice after 4 weeks (Fig. 4f). The body weights of all MoS2-implanted mice increase in parallel to those of the sham control groups (Fig. 4g). Figure 4h and Supplementary Fig. 15 show results of immunohistochemical staining of tissues with MoS2 implants. No obvious tissue damage appears following implantation up to 28 days, indicating that MoS2 does not cause any serious immunologic or inflammatory reactions46. Furthermore, lymphocyte distributions in the peripheral blood of MoS2 implanted mice are comparable to those in sham control groups (Fig. 4i); the percentages of CD4+T, CD8+T, B, and NK cells, which could potentially initiate immune responses against implanted MoS2, remain unchanged throughout four weeks of implantation. These results demonstrate the safety, biocompatibility, and non-toxicity of transient MoS2 bioelectronics, suggesting their suitability for long-term use in the human body.
Bioabsorbable and multifunctional sensors based on monolayer MoS2
These findings on the biocompatible and the bioabsorbable nature of MoS2 allow its use in implantable, temporary biomedical sensors7,21,47. Here, we describe a set of MoS2-based bioabsorbable sensors capable of precision measurements of pressure, temperature, strain, and motion. One area of application is the monitoring of pressure or temperature in the intracranial space during healing from a traumatic brain injury, where the sensor functionality is only necessary during the critical risk period for the patient, typically on a timescale of several days. Figure 5a presents an enlarged schematic illustration and an optical image of a representative device. The construction involves purely bioabsorbable materials, including a PLGA membrane (30 μm thick) as a supporting layer on a substrate of nanoporous silicon, with Mo as the electrode. The top and bottom layers of SiO2 (about 100 nm thick)15,48 provide electrical passivation and act as barriers against biofluids. Details on the materials and fabrication strategies appear in the Methods Section.
To yield a mechanism that is responsive to the pressure of a surrounding biofluid, an etched structure of relief on the surface of the substrate (nanoporous silicon, about 60–80 μm thick) forms a square-shaped air cavity with a depth of 50–60 μm after capping with a film of PLGA. The total size of the substrate is 1.5 mm × 2.5 mm × 0.08 mm and the trench is 0.67 mm × 0.8 mm × 0.03 mm. A layer of MoS2 with an interdigitated geometry acts as a piezoresistive sensor near one of the edges of the cavity, where the bending induced strains are maximized. The resistance of this sensor decreases linearly across the full range of pressures related to those in the intracranial space (0–80 mmHg). The slope of the response is 65 kΩ (mm Hg)−1, consistent with the calculated results49, and a gauge factor of ~60 lies within a range of expected values for MoS221. Motion sensors can be constructed with this same platform by addition of a test mass of PLGA (i.e., a single-axis accelerometer, Fig. 5b-d). Furthermore, the device exhibits a temperature-dependent resistance that is linear (Fig. 5c), thereby enabling its use as a temperature sensor with a sensitivity of ~51 kΩ/°C. The acceleration sensor can also detect the motion of the sensor with a cantilevered mass of PLGA (30 μm thick) (Fig. 5e).
In addition to the individual constituent materials, the biocompatibility of the entire system is important to consider49. The layers of SiO2 and Mo and the nanoporous silicon substrate dissolve in PBS solution (37 °C, pH 7.4) at rates of 23, 8, and 10 μm day−1, respectively. PLGA (75 to 25, lactide–glycolide) is known to dissolve in biofluids within 4–5 weeks50. Figure 5f and Supplementary Fig. 16 show a series of optical images of a typical device at various times after immersion in a PBS solution at 70 °C. A bioabsorbable polymer (polyanhydride) serves as an encapsulation layer. After dissolution and water penetration through this layer after a few days, the Mo components dissolve (within 48 h), followed by the SiO2 layers (within 72 h). Based on results described in previous sections, the MoS2 monolayer will completely dissolve within several months, which is somewhat longer than those for the PLGA51. Therefore, after the metal electrode and PLGA disappear, the MoS2 remains but likely mechanically fractures into small flakes that will dissolve over ~2–3 months. The longevity of the monolayer MoS2 in biofluids facilitates the construction of bioabsorbable electronics and biosensors with longer lifetimes than those with Si NMs. The ability to tune the dissolution rate through control of the grain sizes and the defect densities represents an attractive feature in this context.
Sensors implanted into the intracranial space of animal models (mice) allow demonstrations of modes of use that are relevant for monitoring of patients during a recovery period following a severe traumatic brain injury. Measurements of temperature in this context provide an example. Figure 5g–h outlines the surgical process for implantation. Dissolvable surgical glue prevents leakage of cerebrospinal fluid from the intracranial region. Figure 5i summarizes measurements of temperature using this platform, along with comparisons to data obtained with a commercial, non-resorbable device. The intracranial temperature was varied and controlled via heating or cooling blanket placed under the animal. These results show that real-time in vivo monitoring of biological signals by an entirely bioabsorbable sensor correlate well with results from a conventional device.
In summary, the studies described here yield a broad collection of results related to materials, device, and applications aspects of the use of MoS2 as isolated crystals and large-area polycrystalline films in biodegradable electronic systems. The results reveal that dissolution in simulated and actual biofluids begins at the defect-rich areas of the material, such as the GBs and locations of point defects. More than 2 months are needed to completely dissolve a representative polycrystalline MoS2 monolayer (grain size ~200 nm) in PBS solution at 37 °C, where the rates depend on temperature, pH and the type, and concentration of ions in the solution. Further results demonstrate that monolayer MoS2 is a biocompatible semiconductor, based on analyses of long-term tests in cell level assays, and of in vivo immunological evaluations in live animal models, thereby suggesting suitability for use in bioelectronics. At the device level, implantable biomedical sensors based on CVD-grown monolayer MoS2 can monitor intracranial temperature well over a specified period before dissolving completely. Applications in bioabsorbable and biocompatible transient sensors, which are capable of monitoring critical parameters in the intracranial space associated with recovery from traumatic brain injury, demonstrate the type of applications of interest in clinical medicine. These findings pave the way for the integration of 2D materials into transient, water-soluble electronic platforms, and for their use in biologically and environmentally resorbable technologies.
Synthesis of monolayer MoS2 isolated crystals and continuous films
In this study, we prepared isolated crystals and continuous films of monolayer MoS2 on SiO2/Si substrates by using APCVD22 and LPCVD methods23,24, respectively. Before synthesis, the SiO2 (300 nm)/Si substrates were cleaned by ultrasonication in acetone, isopropyl alcohol, and deionized (DI) water for 10 min each. The SiO2/Si substrates were then immersed in a Piranha solution at 70 °C for 2 h, and subsequently, perylene-3,4,9,10-tetracarboxylic acid tetrapotassium salt (5 μM aqueous solution, 2D semiconductor) was spin-coated onto their surfaces as a seeding solution. We synthesized monolayer MoS2 isolated crystals in a single-zone CVD furnace using the APCVD approach22. MoO3 powder (10 mg, ≥99.5%, Sigma-Aldrich) was placed in an alumina boat and loaded into the center of the furnace chamber. The treated SiO2/Si substrate was covered on the boat in an upside-down manner. Another alumina boat containing S powder (1 g, ≥99.5%, Sigma-Aldrich) was loaded 15 cm away from the center of the furnace. The temperature and deposition time was 750 °C and 30 min, respectively, and the heating rate was 50 °C/min. An Ar:H2 (50:7.5 sccm) gas mixture was used as the carrier gas to reduce the atmosphere to promote the reaction. After synthesis, the furnace was cooled naturally to room temperature. We then fabricated monolayer MoS2 continuous films in a three-zone CVD furnace using the LPCVD method23,24. S (1.4 g) and MoO3 (35 mg) powders were placed separately in two small quartz tubes upstream of the furnace, and the SiO2/Si substrate was placed downstream while facing upstream. The temperatures used in zone 1 (S powder), zone 2 (MoO3 powder), and zone 3 (SiO2/Si substrate) were 120, 590, and 700 °C, respectively. During the synthesis process, the horizontal reaction chamber was maintained at a low pressure (0.7 Torr) with 100 sccm Ar as a carrier gas. After 40 min, the furnace was turned off and cooled down slowly to room temperature.
Dissolution test for monolayer MoS2 isolated crystals and continuous films
Note that both SiO2 and Si dissolve in PBS solution. Therefore, before starting the dissolution test, we transferred the as-grown monolayer MoS2 from SiO2/Si to SU-8/sapphire because the latter has much higher stability than the former in PBS. To create the SU-8/sapphire substrate, we cleaned sapphire in the same way as we cleaned SiO2/Si, and then spin-coated it with SU-8 (10% at 3500 rpm, MicroChem). We then soft-baked the substrate for 5 min at 65 °C, treated it with UV light for 2 min, and then hard-baked it for 30 min at 150 °C. The final thickness of the SU-8 layer was ~100 nm. A poly(methyl methacrylate) (PMMA) solution was spin-coated onto MoS2 as a supporting layer, and the SiO2 layer was etched away using a diluted HF solution, followed by rinsing in DI water. The PMMA/MoS2 film was transferred to the SU-8 layer and dried naturally (>8 h). After removing the PMMA in acetone, the monolayer MoS2 crystals and films were successfully transferred on the SU-8/sapphire substrate with excellent adhesion.
To investigate the dissolution process and behavior of the MoS2 monolayer, we performed a series of dissolution tests in 60 mL of 1.0 M PBS (Sigma-Aldrich) on a hot plate with different temperatures, pH levels, and times. The PBS solution was replaced after every 2 days to maintain a same concentration of the solution throughout the test. For monolayer MoS2 crystals, an accelerated dissolution test was carried out in a PBS solution (pH 7.4) at 75 °C for 49 days. The dissolution test for the monolayer MoS2 continuous films was performed in PBS solutions (pH 7.4–12) at 40–75 °C for 72 days. The pH level was controlled by adding a suitable amount of NaOH solution (0.1 mol L−1). At the end of the dissolution tests, the MoS2 samples were removed from the PBS solutions, rinsed with DI water, and blow-dried using N2 gas.
Characterization of monolayer MoS2 before and after the dissolution tests
The morphology of monolayer MoS2 was investigated both before and after the dissolution tests by using an optical microscope (Eclipse LV100ND, Nikon) and an atomic force microscope (NX10, Park system). To reveal and confirm the location of the GB in MoS2 crystals, a dual-mode erbium-doped fiber laser (Insight Deepsee Dual, Spectra-Physics) was combined with a confocal microscope (IX83, Olympus) to create SHG nonlinear optical images of the crystals. Real-time SHG images were observed after the two spatially overlapped beams were directed to a galvanometric x-y directional mirror controlled by a scanning system (Fluoview1000, Olympus). The Raman and PL spectra and related mapping images were observed using a Raman microscopic system that had a laser wavelength of 532 nm and power of 0.4 mW (Uni-RAM2, UniNanoTech). The transmittance spectra of the monolayer MoS2 continuous films were measured by a UV–Vis spectrophotometer (V-650, JASCO). The concentration of Mo in PBS solution before and after monolayer MoS2 dissolution was measured by using ICP-MS (7900, Agilent). In addition, XPS (K-alpha, Thermo VG) was used to investigate the binding energies of Mo and S in the monolayer MoS2 films that were produced using different dissolution times.
Cell culture and direct contact tests
Mouse fibroblasts from a clone of strain L (NCTC clone 929, KCLB-10001; KCLB, Korea) were cultured in a supplemented Eagle’s minimum essential medium (MEM, 10% fetal bovine serum (FBS), L-glutamate, penicillin, and streptomycin), rat adrenal gland PC-12 (CRL-1721; ATCC, USA) were cultured in a supplemented RPMI-1640 Medium essential medium (RPMI-1640, 10% FBS, L-glutamate, penicillin, and streptomycin), and human endothelial HUVEC (PCS-100-013; ATCC, USA), were cultured in a supplemented F-12K Medium essential medium (F-12K, 10% FBS, L-glutamate, heparin, ECGS, penicillin, streptomycin), and incubated at 37 °C in a humidified 5% CO2 atmosphere. Each cell was pre-cultured for 24 h in culture medium, which was supplemented with a 10% FBS in 24-well plates and exposed for 24 h to the samples placed in the center of each well, which meant that they covered 10% of the cell layer surface. The morphological changes indicating cytotoxicity and cell growth characteristics were recorded as photographs by using an inverted microscope (CKX41, Olympus, Tokyo, Japan). A live/dead assay (Invitrogen, Carlsbad, CA) was used to test cell viability after extended on-chip culturing. Continuous monolayer MoS2 film with L-929, PC-12, and HUVEC cells were incubated with a 1 μM acetol methoxy derivative of calcein (calcein AM, green; live) and 2 μM ethidium homodimer (red; dead) for 35 min in a PBS solution. The cells were then rinsed twice with PBS before the samples were immediately imaged by the inverted microscope; green fluorescence indicated that the cells were viable, while red fluorescence meant that the cells were dead. The images were used to count and calculate the densities of cells in the fluorescein isothiocyanate (FITC, green; live) and tetramethylrhodamine (TRITC, red; dead) channels. The ratio of the integrated density in the FITC to TRITC channel defined the cell viability. The viability of the L-929 cells that came in contact with the sample materials and devices was measured by CCK-8 (Dojindo Laboratories, Kumamoto, Japan) according to the manufacturer’s instructions. To summarize, L-929 cells were incubated with the HDPE, ZDEC, MoS2 flakes in DI water/PBS, and dissolved MoS2 in PBS on 24-well plates for 24 h before the addition of 10% CCK-8 solution. Cells were then incubated for an additional 2 h at 37 °C to form water-dissoluble formazan. A 50 μL formazan solution was then collected from each sample and added to one well in a 96-well plate. Three parallel replicates were prepared. The absorbance at 450 nm was determined using a microplate spectrophotometer (iMark, Bio-RAD).
Flow cytometry with apoptosis assays
Anti-mouse CD3, CD4, CD8, CD19, NK1.1, CD14, Gr-1, and F4/80 monoclonal antibodies (mAbs) conjugated with FITC, phycoerythrin, phycoerythrin-cyanine dye, peridinin chlorophyll protein complex with cyanin-5.5, and allophycocyanin were purchased from eBioscience (San Diego, CA, USA). Isolated cells were resuspended in a 100 μL FAC buffer (PBS containing 2% FBS and 0.02% sodium azide) solution and incubated with anti-CD16/CD32 mAbs (2.4G2, Fc block) to block the FcRIII/II receptors. The cells were then incubated for 20 min at 4 °C. Flow cytometry was performed with FACS Canto II (BD Bioscience, San Diego, CA, USA), and the data were analyzed by the FlowJo software (ThreeStar, USA). Fifty thousand lymphocyte populations gated by forward scatter/side scatter were analyzed. Annexin V-FITC Apoptosis detection kit was purchased from BD Biosciences (San Diego, US). The L-929 cells were then analyzed for their expression of Annexin V and 7-AAD to determine the number of viable cells: Annexin V negative and 7-AAD negative (Annexin V−/7-AAD−); cells undergoing apoptosis, Annexin V positive and 7-AAD negative (Annexin V+/7-AAD−); and dead cells or cells that were in the last stage of apoptosis, Annexin positive and 7-AAD positive (Annexin V+/7-AAD+).
In vivo tissue biocompatibility tests
Sterile implant samples, each approximately 3 mm × 10 mm, were prepared aseptically. We used 6-week-old female mice, the mouse groups were sham only (n = 3), 1 week (n = 3), 2 weeks (n = 3), and 4 weeks (n = 3). The mice were implanted with continuous monolayer MoS2 films for four weeks and euthanized prior to the excision of muscle tissues and macroscopic examination of the implant sites. The inserted devices initially adhered to the subcutaneous layers (hypodermis) due to adequate adhesion to the skin layer based on van der Waals forces. A thin transparent layer of fibroblasts was found to have grown over the devices, which helped the inserted devices to stay in place. A microscopic evaluation was then conducted to further determine the tissue response.
The mice were euthanized via CO2 asphyxiation, and the implanted samples and surrounding tissues were excised. The tissues were then fixed in 10% formalin, embedded in paraffin, cut into 4 μm sections, and stained using H&E for histological analysis. The experiments were repeated three times. Statistical significance was determined by one-way analysis of variance followed by Dunnett’s multiple comparison tests. Significance was ascribed at p < 0.05. All analyses were conducted using the Prism software (Graph Pad Prism 5.0).
Fabrication of MoS2-based bioabsorbable sensors
The fabrication began with the spin-coating of polyimide (~1.2 μm, Sigma-Aldrich) and a sacrificial layer of PMMA (~100 nm, MicroChem) on temporary silicon carrier substrates. Electron-beam evaporation was used to develop a layer of SiO2 (thickness 100 nm) that could encapsulate the MoS2 atomic layer and protect it during reactive ion etching (RIE). Mo was also deposited by electron-beam evaporation for metal electrode. Monolayer MoS2 grown by the LPCVD method was transferred and patterned using photolithography and CHF3/O2 plasma etching (35/15 sccm, 100 W, 10 s). The casting and patterning of a top coating of D-PI (~1.2 μm) facilitated the formation of the MoS2 channel and connected it to a place near the neutral mechanical plane. Defining a mesh structure across the multilayer (D-PI/SiO2/D-PI/PMMA) by RIE before immersing it in a buffered oxide etchant exposed the base layer of PMMA and allowed it to be released into acetone for transfer onto a PLGA film (~30 μm thick). The whole device was heated to a temperature near the glass transition temperature of PLGA (55–60 °C; the lactide/glycolide ratio was 75:25) to make it more stable. Last, the top layer of the PI was eliminated by RIE.
Evaluation in animal models
Studies were performed in strict accordance with the recommendations in the “Guide for the Care and Use of Laboratory Animals” of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee of Washington University in St. Louis. Male Lewis rats (n = 3) weighing 275–300 g (Charles River Laboratories) received subcutaneous injections of buprenorphine hydrochloride (0.05 mg/kg, Reckitt Benckiser Healthcare Ltd, USA) for pain management, and of ampicillin (50 mg/kg; Sage Pharmaceuticals, USA) to prevent infection at the implantation site before the surgical process. Animals were anaesthetized with isoflurane gas (4% for introduction and 2% for maintenance) and held in a stereotaxic frame for the duration of the surgical procedure and measurements. Opening a craniectomy and dural, implanting bioresorbable temperature sensor on the cortical surface, sealing the craniectomy with a biodegradable surgical glue (HistoacrlyR, Braun Corporation, Spain), and suturing the skin implanted the fully resorbable biosensing device in intracranial space. Comparison testing with commercial thermistor (DigiKey Electronics, USA) implanted in parallel to bioresorbable sensors demonstrated the functionality of the bioresorbable sensors. In vivo functionality tests of temperature sensors involved three trials using different batches of devices and animals for the reproducibility52,53.
Data supporting the findings of this study are available within the article and its Supplementary Information file, and from the corresponding authors upon reasonable request.
Fu, K. K., Wang, Z., Dai, J., Carter, M. & Hu, L. Transient electronics: materials and devices. Chem. Mater. 28, 3527–3539 (2016).
Tan, M. J. et al. Biodegradable electronics: cornerstone for sustainable electronics and transient applications. J. Mater. Chem. C 4, 5531–5558 (2016).
Camposeo, A. et al. Physically transient photonics: random versus distributed feedback lasing based on nanoimprinted DNA. ACS Nano 8, 10893–10898 (2014).
Chen, C. et al. Transient micromotors that disappear when no longer needed. ACS Nano 10, 10389–10396 (2016).
Lee, G. et al. Fully biodegradable microsupercapacitor for power storage in transient electronics. Adv. Eng. Mater. 7, 1700157 (2017).
Jia, X., Wang, C., Zhao, C., Ge, Y. & Wallace, G. G. Toward biodegradable Mg-air bioelectric batteries composed of silk fibroin-polypyrrole film. Adv. Funct. Mater. 26, 1454–1462 (2016).
Zheng, Q. et al. Biodegradable triboelectric nanogenerator as a life-time designed implantable power source. Sci. Adv. 2, e1501478 (2016).
Yu, K. J. et al. Bioresorbable silicon electronics for transient spatiotemporal mapping of electrical activity from the cerebral cortex. Nat. Mater. 15, 782–791 (2016).
Hwang, S.-W. et al. Materials and fabrication processes for transient and bioresorbable high-performance electronics. Adv. Funct. Mater. 23, 4087–4093 (2013).
Yin, L. et al. Dissolvable metals for transient electronics. Adv. Funct. Mater. 24, 645–658 (2014).
Hwang, S. W. et al. High-performance biodegradable/transient electronics on biodegradable polymers. Adv. Mater. 26, 3905–3911 (2014).
Bettinger, C. J. & Bao, Z. Organic thin-film transistors fabricated on resorbable biomaterial substrates. Adv. Mater. 22, 651–655 (2010).
Wu, H. et al. Biocompatible inorganic fullerene-like molybdenum disulfide nanoparticles produced by pulsed laser ablation in water. ACS Nano 5, 1276–1281 (2011).
Yin, L. et al. Mechanisms for hydrolysis of silicon nanomembranes as used in bioresorbable electronics. Adv. Mater. 27, 1857–1864 (2015).
Kang, S.-K. et al. Dissolution behaviors and applications of silicon oxides and nitrides in transient electronics. Adv. Funct. Mater. 24, 4427–4434 (2014).
Lee, W., Jang, H., Jang, B., Kim, J. H. & Ahn, J. H. Stretchable Si logic devices with graphene interconnects. Small 11, 6272–6277 (2015).
Lin, Z. et al. 2D materials advances: from large scale synthesis and controlled heterostructures to improved characterization techniques, defects and applications. 2D Mater. 3, 042001 (2016).
Tan, C. et al. Recent advances in ultrathin two-dimensional nanomaterials. Chem. Rev. 117, 6225–6331 (2017).
Ganatra, R. & Zhang, Q. Few-layer MoS2: a promising layered semiconductor. ACS Nano 8, 4074–4099 (2014).
Radisavljevic, B., Radenovic, A., Brivio, J., Giacometti, V. & Kis, A. Single-layer MoS2 transistors. Nat. Nanotechnol. 6, 147–150 (2011).
Park, M. et al. MoS2-based tactile sensor for electronic skin applications. Adv. Mater. 28, 2556–2562 (2016).
Lee, Y. H. et al. Synthesis of large-area MoS2 atomic layers with chemical vapor deposition. Adv. Mater. 24, 2320–2325 (2012).
Zhang, J. et al. Scalable growth of high-quality polycrystalline MoS2 monolayers on SiO2 with tunable grain sizes. ACS Nano 8, 6024–6030 (2014).
Chen, X. et al. Lithography-free plasma-induced patterned growth of MoS2 and its heterojunction with graphene. Nanoscale 8, 15181–15188 (2016).
Kang, K. et al. High-mobility three-atom-thick semiconducting films with wafer-scale homogeneity. Nature 520, 656–660 (2015).
Kalantar-zadeh, K. et al. Two-dimensional transition metal dichalcogenides in biosystems. Adv. Funct. Mater. 25, 5086–5099 (2015).
Gan, X., Zhao, H. & Quan, X. Two-dimensional MoS2: a promising building block for biosensors. Biosens. Bioelectron. 89, 56–71 (2017).
Wang, Z. et al. Chemical dissolution pathways of MoS2 nanosheets in biological and environmental media. Environ. Sci. Technol. 50, 7208–7217 (2016).
Kurapati, R. et al. Enzymatic biodegradability of pristine and functionalized transition metal dichalcogenide MoS2 nanosheets. Adv. Funct. Mater. 27, 1605176 (2017).
Park, W. et al. Photoelectron spectroscopic imaging and device applications of large-area patternable single-layer MoS2 synthesized by chemical vapor deposition. ACS Nano 8, 4961–4968 (2014).
Ling, X. et al. Role of the seeding promoter in MoS2 growth by chemical vapor deposition. Nano Lett. 14, 464–472 (2014).
Yin, X. et al. Edge nonlinear optics on a MoS2 atomic monolayer. Science 344, 488–490 (2014).
van der Zande, A. M. et al. Grains and grain boundaries in highly crystalline monolayer molybdenum disulphide. Nat. Mater. 12, 554–561 (2013).
Zhou, W. et al. Intrinsic structural defects in monolayer molybdenum disulfide. Nano Lett. 13, 2615–2622 (2013).
Kc, S., Longo, R. C., Wallace, R. M. & Cho, K. Surface oxidation energetics and kinetics on MoS2 monolayer. J. Appl. Phys. 117, 135301 (2015).
Rong, Y. et al. Controlled preferential oxidation of grain boundaries in monolayer tungsten disulfide for direct optical imaging. ACS Nano 9, 3695–3703 (2015).
Dungey, K. E., Curtis, M. D. & Penner-Hahn, J. E. Behavior of MoS2 intercalation compounds in HDS catalysis. J. Catal. 175, 129–134 (1998).
Wang, L., Xu, Z., Wang, W. & Bai, X. Atomic mechanism of dynamic electrochemical lithiation processes of MoS2 nanosheets. J. Am. Chem. Soc. 136, 6693–6697 (2014).
Wang, X., Shen, X., Wang, Z., Yu, R. & Chen, L. Atomic-scale clarification of structural transition of MoS2 upon sodium intercalation. ACS Nano 8, 11394–11400 (2014).
Zhang, L. et al. In situ TEM observing structural transitions of MoS2 upon sodium insertion and extraction. RSC Adv. 6, 96035–96038 (2016).
Yamamoto, M., Einstein, T. L., Fuhrer, M. S. & Cullen, W. G. Anisotropic etching of atomically thin MoS2. J. Phys. Chem. C 117, 25643–25649 (2013).
Yang, L. et al. Chloride molecular doping technique on 2D materials: WS2 and MoS2. Nano Lett. 14, 6275–6280 (2014).
Irimia-Vladu, M. et al. Biocompatible and biodegradable materials for organic field-effect transistors. Adv. Funct. Mater. 20, 4069–4076 (2010).
Teo, W. Z., Chng, E. L., Sofer, Z. & Pumera, M. Cytotoxicity of exfoliated transition-metal dichalcogenides (MoS2, WS2, and WSe2) is lower than that of graphene and its analogues. Chemistry 20, 9627–9632 (2014).
Appel, J. H. et al. Low cytotoxicity and genotoxicity of two-dimensional MoS2 and WS2. ACS Biomater. Sci. Eng. 2, 361–367 (2016).
Park, G. et al. Immunologic and tissue biocompatibility of flexible/stretchable electronics and optoelectronics. Adv. Healthcare Mater. 3, 515–525 (2014).
Park, J. et al. Electromechanical cardioplasty using a wrapped elasto-conductive epicardial mesh. Sci. Transl. Med. 8, 344ra386 (2016).
Park, K. et al. Stretchable, transparent zinc oxide thin film transistors. Adv. Funct. Mater. 20, 3577–3582 (2010).
Mittal, R. et al. Use of bio-resorbable implants for stabilisation of distal radius fractures: the United Kingdom patients’ perspective. Injury 36, 333–338 (2005).
Wu, L. & Ding, J. In vitro degradation of three-dimensional porous poly(D,L-lactide-co-glycolide) scaffolds for tissue engineering. Biomaterials 25, 5821–5830 (2004).
Hao, J. et al. In vivo long-term biodistribution, excretion, and toxicology of PEGylated transition-metal dichalcogenides MS2 (MMo, W, Ti) Nanosheets. Adv. Sci. 4, 1600160 (2017).
Ross, S. & Sussman, A. Surface oxidation of molybdenum disulfide. J. Phys. Chem. 59, 889–892 (1955).
Gao, J. et al. Aging of transition metal dichalcogenide monolayers. ACS Nano 10, 2628–2635 (2016).
This work was supported by the National Research Foundation of Korea (NRF) funded by the Korean government (MSIT) (NRF-2015R1A3A2066337, NRF-2016M3A9B6948342).
The authors declare no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
About this article
Cite this article
Chen, X., Park, Y.J., Kang, M. et al. CVD-grown monolayer MoS2 in bioabsorbable electronics and biosensors. Nat Commun 9, 1690 (2018). https://doi.org/10.1038/s41467-018-03956-9
Nature Synthesis (2022)
Nano Research (2022)
International Journal of Precision Engineering and Manufacturing-Green Technology (2021)
Analytical and Bioanalytical Chemistry (2021)
Nano Research (2021)