The twain finally meet: Nanowires and nanotubes combined to form intracellular bioelectronic probes

Feb 03, 2014 by Stuart Mason Dambrot feature
Schematics and SEM images of the ultrasmall BIT-FET. (A) Schematic illustration of an intracellular bioelectronic probe. (Left) General scheme of a probe for intracellular electrophysiology recording. (Right) A magnified view of the tip of a sub-10-nm bioelectronic probe and its related size to single ion channel. (B) Schematic structure of the ultrasmall BIT-FET. Green, yellow, blue, and gray colors represent SiO2 layer, metal contact, SiNW, and silicon nitride substrate, respectively. (C) SEM images of the ultrasmall BITFET at different fabrication steps. A GeNW branch was first grown on top of SiNW (I), followed by a subsequent H2O2 etching of top part of GeNW to shrink its diameter down to sub-10-nm regime (II). A final view of an ultrasmall BIT-FET with nanotube ID ∼8 nm, and SiO2 wall thickness ∼10 nm is presented in (III). Inset of III is the closeup of the tip of the ultrasmall SiO2 nanotube. White dashed lines in II and III indicate the point below which the GeNW and SiO2 is protected by photoresist during H2O2 and BHF etching, respectively. All scale bars: 100 nm. Credit: Copyright © PNAS, doi:10.1073/pnas.1323389111

( —Miniaturized bioelectronic probes stand to transform biology and medicine by allowing measurement of intracellular components in vivo. Recently, scientists at Harvard University and Peking University designed, fabricated and demonstrated bioelectronic probes as small as 5 nanometers using a unique three-dimension nanowire-nanotube heterostructure. (A heterostructure combines multiple heterojunctions – interfaces between two layers or regions of dissimilar crystalline semiconductor – in a single device.) Through experimental measurements and numerical simulations, the researchers showed that these devices have sufficient time resolution to record the fastest electrical signals in neurons and other cells, with integration into larger chip arrays potentially providing ultra-high-resolution mapping of activity in neural networks and other biocellular systems.

Prof. Xiaojie Duan discussed the paper that she, Graduate Researcher Tian-Ming Fu, Prof. Charles M. Lieber and their co-authors published in Proceedings of the National Academy of Sciences. She first points out that nanotube probes and their heterojunction with silicon nanowire field-effect transistors (SiNW FETs) become mechanically less stable as diameter is reduced. "When the nanotube gets smaller and smaller," Duan tells, "it gets easier to break the nanotube at the junction area with the SiNW. In the application of using the for intracellular bioelectronic detection, there will be various forces, such as the capillary force from the liquid, as well as interaction between the probe and the . These forces may break the probe if we have a weak junction between it and the SiNW."

Another issue is that electrical sensitivity is also reduced as nanotube diameter decreases, because the nanotube inner diameter (ID) defines the effective device gate area. "In the recording of intracellular transmembrane potential using our probe," Duan explains, "cytosol fills the nanotube and acts as the gate electrode for the underlying SiNW FET." Cytosol (also termed intracellular fluid or cytoplasmic matrix) is the liquid found inside cells, excluding organelles and other cytoplasmic components. "The cytosol potential change modulates the carrier density of the SiNW FET, thereby changing its conductance," Duan continues. "This is how our probe works for bioelectronics recording." The contact area between the cytosol and the SiNW – defined by the inner diameter of the nanotube – determines conductance modulation effectiveness. In other words, if the nanotube inner diameter is too small, the SiNW FET gate area will be too small as well.

Intracellular resting membrane potential recording. (A) Schematics (Upper) and differential interference contrast optical microscopy images (Lower) of an HL-1 cell manipulated by a glass micropipette to approach (I), contact (II), penetrate (III), and retract (IV) from a phospholipid-modified ultrasmall BITFET probe. The red arrow indicates the position of the ultrasmall nanotube tip. Because pure SiO2 nanotube is optically transparent, the GeNW template of this device was not etched for imaging. Scale bar: 2 μm. (B) Representative electrical recording results from a ∼10 nm ID ultrasmall BIT-FET device; in this case, the GeNW was etched to yield the ultrasmall SiO2 nanotube. Down- and up-pointing green arrows mark the beginning of cell penetration and withdrawal, respectively. The upper and lower horizontal dashed lines indicate the extracellular and intracellular potentials. Quasi-static water-gate measurements made before/after cell measurements show <2% change in the device conductance and sensitivity. Credit: Copyright © PNAS, doi:10.1073/pnas.1323389111

The researchers were also presented with the fact that high-frequency dynamic response may degrade with decreasing nanotube inner diameter due to increasing solution resistance in the nanotube. "When the nanotube inner diameter decreases," Duan explains, "the resistance of the solution inside the nanotube will increase, because of the decrease of the solution conductor cross-section." In addition, the speed at which the underlying SiNW FET can respond to a signal is determined by the product RC of the capacitance and the resistance of the solution conductor inside the nanotube – so if the nanotube is getting smaller, the resistance is getting bigger. This means that the SiNW FET will need more time to respond – and if the signal change is too fast, the probe will not be able to reliably record it. "That is what we mean by 'high-frequency dynamic response will degrade with decreasing nanotube inner diameter,' Duan adds, "However, we found that for our probe, even if we decrease the nanotube inner diameter to as small as 5 nm, the probe is still able to record a 3 kHz signal faithfully – a bandwidth sufficient in most cases for recording neural and cardiac signals."

The researchers were also confronted by the need to use active semiconductor nanowire field-effect transistor detectors to overcome the limitations of probe size reduction. "The word active' is compared to the 'passive' nature of recording with metal electrodes," Duan points out. "For metal electrode recording, a portion of the transmembrane potential Vm will be dropped or lost at the electrode/electrolyte interface, so the signal recorded will be smaller than the real transmembrane potential." When the size of the metal electrode is reduced, the impedance value at the electrode/electrolyte will increase – and at some point, this impedance will get so large that the recorded signal will be obscured by noise. However, for the FET recording, the cytosol potential change is reflected by the semiconductor channel conductance change, which is independent of the probe/electrolyte interface impedance. "Since decrease in probe size will therefore not affect the signal amplitude," Duan adds, "using the FET to sense potential is a very effective way to overcome the limitations of probe-size reduction."

Schematics of the fabrication flow for the ultrasmall BIT-FET. (A) SiNWs (blue) are dispersed on substrate (solid gray). (B) S/D contacts are defined by EBL followed by thermal evaporation. (C) Au nanodots are defined on SiNWs between S/D using EBL and thermal evaporation. (D) GeNWs (red) are grown on top of the SiNWs through nanocluster-catalyzed CVD process. (E) A thin layer of photoresist (transparent gray) is spin coated on the chip to protect the lower GeNW part. (F) The resulting H2O2-etched GeNWs following photoresist liftoff. Only the GeNW above the photoresist in E is thinned by etching in H2O2. (G) SiO2 is conformally deposited over the entire chip by ALD. (H) A thin layer of photoresist (transparent gray) is spin coated to protect the lower region of chip. (I) The resulting BHF etched structures following liftoff. The region of SiO2 above the photoresist layer in H is etched to ca.10-nm thickness. (J) Photoresist with thickness smaller than the GeNW heights is deposited. (K) The resulting structure following BHF etching of SiO2, which exposes the tips of the GeNWs. Isotopic BHF etching yields a small taper with thinner SiO2 at the topmost part of the structure. (L) The GeNW is removed by H2O2 etching to form an ultrasmall nanotube connected to the bottom SiNW FET. Copyright © PNAS, doi:10.1073/pnas.1323389111

Another critical challenge was synthetically integrating and nanowires. "The nanotube is made of silicon oxide and the nanowire is made of silicon," Du an notes. "For cell recording, the nanotube needs to be built vertically onto the nanowires – meaning that a three-dimensional heterostructure is needed." While the heterostructure could be fabricated in various ways, not all of them provided the required controllability at the probe scale. Therefore, the researchers grew a germanium nanowire (GeNW) on top of the silicon nanowires, with the GeNW acting as a template for the nanotube. "Depositing SiO2 on the GeNW/SiNW heterostructure, then selectively removing the core GeNW, resulted in the desired nanotube/SiNW structure," notes Duan. "Using this method, the nanotube inner diameter can be controlled by the GeNW diameter, the outer diameter can be controlled by the SiO2 thickness, and the nanotube length can be defined by the GeNW growth time. This gives us complete and easily implemented control over the probe's dimensions."

Finally, the scientists had to investigate and model the bandwidth effect of phospholipid coatings, which are important for intracellular recording. "We use phospholipid coating to assist the nanotube probe to penetrate the cell membrane," Duan notes. "Since the bandwidth of our probe is important for the probe to be able to record fast neural or cardiac signal, we need to make sure this phospholipid modification will not overly affect the bandwidth." Because the phospholipid layer will decrease the cross section of the solution conductor inside the nanotube, it impacts bandwidth in two ways by changing the capacitance and resistance of the solution inside the nanotube. (The phospholipid-modified probe bandwidth is easily determined by applying a fast artificial signal to the solution, and then recording how the conductance of the device changes over time. This provides the time the device needs to respond to this signal and thus the device's bandwidth.) "Therefore," Duan explains, "if the nanotube is large, this thin phospholipid layer will not cause too much of a difference. However, for our probe, the sub-10 nm nanotube size is almost the same scale as the lipid layer – so we have to carefully examine how it affects probe bandwidth, both experimentally and theoretically."

The scientists addressed these myriad challenges in designing, fabricating, and demonstrating the probe with three key innovations. "The first is the use of FET as potential sensing element, which in principle enables us to overcome the size limit on the probe." Duan explains. "However, FETs have conventionally existed in a linear geometry with connections that preclude access to the inside of cells." The second innovation was the solution to this problem – namely, the design of a vertical SiO2 nanotube on top of the nanoscale FET, which allowed them to introduce cytosol into the cell without having to insert the FET channel inside the cell, which would be more invasive. The third key is the design of a relatively large nanotube base with a much sharper nanotube tip, resulting in a sub-10 nm probe without sacrificing its mechanical strength and electrical sensitivity.

Electron microscopy characterization of the ultrasmall BIT-FET. (A) Representative SEM (Zeiss Ultra Plus field-emission SEM) images of intermediate fabrication steps of the ultrasmall BIT-FET. (Left) Device after 30-nm ALD coating of SiO2. (Right) Device after first step of selective BHF etching of the upper80% portion of the SiO2 to ca. 10 nm (Fig. S1 H and I). White dashed lines in I and II indicate the point below which the SiO2 is protected by photoresist during BHF etching. Scale bars: 200 nm. (B) False-colored transmission electron microscopy (JEOL 2100 TEM) image of an ultrasmall nanotube. This tube was fabricated following the same procedure as described in SI Text, and deposited onto lacey carbon grids (Ted Pella) from ethanol suspension. It has a tip ID ∼7 nm and bottom ID ∼80 nm. False color is used here to distinguish the SiO2 nanotube from background amorphous carbon. Scale bar: 50 nm. Credit: Copyright © PNAS, doi:10.1073/pnas.1323389111

An important aspect of the study was ensuring that the bioelectronic devices had sufficient time resolution to record the fastest electrical signals in neurons and other cells. "The signals in neural system are normally in the millisecond scale," Duan points out. "That means that to reliably record these signals, the recording device needs to have a bandwidth measure in kilohertz." While the probe's bandwidth decreases with the decrease of nanotube diameter, the scientists found that even for probes with inner diameters as small as 5 nm, the bandwidth is still around 3 kHz (a time resolution ~ 0.3 ms) in physiological solution. "This means that our probes have sufficient time resolution to record the fastest electrical signals in neurons and other cells," Duan adds.

Moreover, Duan points out, the scientists found that measuring the cell transmembrane resting potential with these ultrasmall bioelectronic devices demonstrates the capability for intracellular electrophysiology studies. "When we measured the transmembrane resting potential of HL-1 cell with our new probes, we found that with the phospholipid modification, the nanotube can easily and reliably penetrate the cell membrane, allowing the FET to record the intracellular transmembrane potential at full amplitude." After retracting the nanotube from the cell, the recorded potential can immediately revert to the extracellular potential. "Reliable cell membrane penetration and stable recording of intracellular transmembrane potential prove the capability of our probes for intracellular electrophysiology studies," notes Duan.

Sensitivity of different device structures. (A, B) Schematics of the ultrasmall BIT-FET without and with Ge overcoating on the SiNW, respectively. I and II correspond to the BIT-FET devices before and after Ge core etching. III show schematically typical conductance (G) vs. water-gate (Vwg) measurements from these distinct structures. Credit: Copyright © PNAS, doi:10.1073/pnas.1323389111

Moving forward, says Duan, the researchers' are planning to scale up their work to integrate the probes into high-density, large-scale array for large-scale mapping of neural activities; use the probes to record neural signals from small subcellular structures/organelles; and investigate other applications in which the probes will provide substantially greater spatial resolution and minimal invasiveness than other techniques.

In addition, the scientists might consider developing other innovations. "For example," Duan illustrates, "a major challenge in using our ultra-small probes for recording from small subcellular structures is to accurately position them with respect to the subcellular structures of interest. We're looking at either labeling our probe with fluorescence dye – or other biocompatible materials – to mark the nanotube at high resolution, or using specific targeting in which the probe's biochemical surface groups define the specific cell location being studied."

Duan sees other areas of research that might benefit from their study, including:

  • Neuroscience: Since recording of intracellular action potential and other low frequency transmembrane potential signal is important to study the neural network.
  • Medicine: By studying how the drugs affect the recorded intracellular action potentials or other transmembrane potentials, the probes can provide an attractive technique for parallel or high-throughput screening of drugs targeting the ion channels.
  • Biophysics: The process of cell membrane penetration by the phospholipid-modified nanotube can serve as a useful platform for studying the cell membrane's fusion-related biological processes.

Explore further: Carbon nanotube harpoon catches individual brain-cell signals

More information: Sub-10-nm intracellular bioelectronic probes from nanowire–nanotube heterostructures, Proceedings of the National Academy of Sciences Published online before print on January 13, 2014, doi:10.1073/pnas.1323389111

Related Stories

Recommended for you

First direct observations of excitons in motion achieved

Apr 16, 2014

A quasiparticle called an exciton—responsible for the transfer of energy within devices such as solar cells, LEDs, and semiconductor circuits—has been understood theoretically for decades. But exciton ...

User comments : 2

Adjust slider to filter visible comments by rank

Display comments: newest first

1 / 5 (1) Feb 04, 2014
Beware that a gang of Harvard "researchers" has already stole in Nature journals and, with further support of the MIT's ones, in ASC Nano Lett both the ideas and money of taxpayers. There are numerous swindlers from David H. Koch Inst. for Integrative Cancer Research and Dept of Chemical Engineering, also with Dept of Chemistry and Chem. Biology and School of Eng and Applied Science of Harvard University at & .
Their plagiaristic compilation titled Macroporous nanowire nanoelectronic scaffolds for synthetic tissues (DOI: 10.1038/NMAT3404) and Outside Looking In: Nanotube Transistor Intracellular Sensors ( was funded by NIH Director's Pioneer Award (1DP1OD003900) and a McKnight Foundation Technol Innovations in Neurosc Award, also a Biotechnology Research Endowment from the Dept. of Anesthesiology at Children's Hospital Boston and NIH grants GM073626, DE01-3023/6516.
not rated yet Feb 14, 2014
@rsklyar, you have no credibility because I clicked on the first link and it took me to a page where in the first paragraph, the word "peaces" is used when it is supposed to be "pieces". If you can't even get that right, I don't see how you can get much else right. The authors of the work that you link to are obviously sloppy and/or uneducated, and as such, have no credibility. Good day.

More news stories

Wiring up carbon-based electronics

Carbon-based nanostructures such as nanotubes, graphene sheets, and nanoribbons are unique building blocks showing versatile nanomechanical and nanoelectronic properties. These materials which are ordered ...

Tiny power plants hold promise for nuclear energy

Small underground nuclear power plants that could be cheaper to build than their behemoth counterparts may herald the future for an energy industry under intense scrutiny since the Fukushima disaster, the ...