You are on page 1of 2

NEWS & VIEWS RESEARCH

builds on these concepts. The key idea comes from the fact that, by introducing an asymmetry in the physical geometry of the diode, asymmetry can be created in the devices currentvoltage behaviour such that the same voltage will yield a different amount of current for two opposite directions of the magnetic field. Take, for example, the case of a planar device that has two vertically separated p- and By harnessing the way charge carriers move in a magnetic field, computing n-type semiconductor layers lying in a magblocks based on semiconductor junctions have been made that are reconfigurable netic field whose direction is completely in and can be interconnected to perform complex logic functions. See Letter p .72 the same plane as the device. Because of this in-plane field, charge carriers will be deflected more in one direction in the planar device S AY E E F S A L A H U D D I N is dubbed a p-type semiconductor. By than in the other. But if an out-of-plane field is combining n-type and p-type, many dif- applied in the same planar device, carriers will onventional computers rely on the ferent pn junctions can be made. These be deflected in all directions equally. For the controlled motion of charge in tran- junctions are the basic building blocks of all in-plane field, the asymmetry in carrier deflecsistors, the basic building blocks of semiconductor devices. tion dictates that if the avalanche diode is in integrated circuits. It has been pointed out, One such device is an avalanche diode, in the on state for a particular direction of the however, that if one could devise electronic which highly energetic carriers are injected field, it will be off when the field is reversed. devices based on the spin of particles a beyond a specific electrical voltage from one Therefore, two avalanche diodes under two quantum-mechanical property responsible side of the junction to the other. Owing to their opposite directions of the applied field can profor magnetism completely new computing high kinetic energy, these carriers can gener- vide a complementary pair that can be used, technologies might emerge that could out- ate additional carriers by breaking chemical almost like conventional transistors, to build perform their conventional analogues. Such bonds as they collide with atoms near the junc- up computing blocks. spintronic devices have attracted much tion. In turn, the additional carriers produce One interesting aspect of Joo and colleagues attention among researchers worldwide 1. more carriers in a chain reaction that leads to work is the exploitation of mirror symmeOn page72 of this issue, Joo etal.2 describe an avalanche of carriers (Fig.1). Avalanche try. Because oppositely charged carriers are a computing scheme that combines the phys- diodes are commercially available and are deflected in opposite directions in the magics of charge motion in a magnetic field with routinely used to provide voltage reference netic field, the behaviour of a device that has conventional electronic devices in a way that and electrical-surge protection in integrated a top n/bottom p-layer configuration will be promises to provide a significant step towards circuits. Now, if an avalanche diode is placed exactly opposite to a top p/bottom n one. Thus, a spintronic computing technology*. in a magnetic field, it is conceivable that, owing the requirement to reverse the direction of the Joo and colleagues scheme hinges on the to the Lorentz force, the energetic carriers will magnetic field to create the complementary physics of the Lorentz force on charged parti- be deflected and will therefore need a larger pair can be replaced by simply interchanging cles. When a charged particle moves in a mag- voltage to reach the energy threshold above the ordering of n and p layers. This strategy netic field, it experiences a force that deflects which they can multiply and produce a carrier provides a significant advantage in terms of it in a specific direction. The direction of this avalanche. As a result, the current in the diode flexibility in designing complicated and interforce depends on the sign of the charge, the will decrease35. connected computing blocks by leveraging direction of the magnetic field and the direcJoo and colleagues computing scheme existing expertise on semiconductor processtion of motion. It can be reversed ing. Also, the fact that the direction by changing the sign of any of these of the magnetic field does not have three quantities. So, for example, if to be reversed is highly advantageous Carrier carriers move from left to right in a for the operation of the device at low magnetic field, the force experienced power and for overall scaling up of the by the negatively charged carriers will technology. Carrier act in the opposite direction to that However, improvement is needed de ection experienced by the positively charged before this computing scheme can be carriers. viable for practical applications. Even Avalanche This property can be exploited in the need to generate a unidirectional a semiconductor, in which mobile magnetic field in an integrated microcharges that are predominantly negachip could consume significant energy. Conduction band tive or positive can be created using One possibility, which the authors a technique called doping. Doping suggest, would be to develop techconsists of adding a relatively nologies in which the required field is Valence band small number of atoms of a difproduced by thin-film permaferent material from that of the nent magnets rather than by passhost semiconductor to increase Figure 1 | Bending avalanche carriers. Charge carriers going ing a current through a wire. its negative (electrons) or positive through an avalanche multiplication process in a pn semiconductor However, integrating the materi(holes) charges. A semiconductor junction can be deflected by an applied magnetic field. Here, als needed for this purpose with the junction is depicted through its energy-band diagram, which that has mostly mobile electrons consists of the valence and conduction bands separated by an energy conventional semiconductors is said to be of the n-type category, gap. The deflection leads to a reduction in the electrical current poses a considerable challenge. whereas one that has mainly holes running through the junction. Joo etal.2 have exploited this effect to An attractive aspect of this techdevelop basic computing blocks that can be interconnected to perform nology is the reconfigurability of *This article and the paper under discussion2 were published online on 30January 2013. complex logic operations. the computing blocks such that the
S O L I D -STAT E P H YS I C S

A new spin on spintronics

7 F E B R UA RY 2 0 1 3 | VO L 4 9 4 | N AT U R E | 4 3

2013 Macmillan Publishers Limited. All rights reserved

RESEARCH NEWS & VIEWS


same block can perform different logic functions depending on an external signal. But this can be achieved only if the magnetic field is reversed. The authors suggest using an effect called spin-transfer torque6,7 to attain this reversal. However, such an effect can work only on small magnets, and so this approach may not be scalable. Additionally, the use of energetic carriers in the avalanche diode requires a large amount of energy to be supplied to the diode from external circuits, which is then translated into the kinetic energy of the carriers. This may prove to be a bottleneck for low-power device operation. Nevertheless, none of these difficulties seems to be of a fundamental nature, and greater insight into the capabilities of this technology
NE UROSCIENCE

may lead to completely new ways of doing computation. Sayeef Salahuddin is in the Department of Electrical Engineering and Computer Sciences, University of California, Berkeley, Berkeley, California 94720, USA. e-mail: sayeef@berkeley.edu
1. Insight: Spintronics Nature Mater. 11, 367416 (2012); available at go.nature.com/sarlr9 2. Joo, S. et al. Nature 494, 7276 (2013). 3. Delmo, M. P., Yamamoto, S., Kasai, S., Ono, T. & Kobayashi, K. Nature 457, 11121115 (2009). 4. Wan, C., Zhang, X., Gao, X., Wang, J. & Tan, X. Nature 477, 304307 (2011). 5. Lee, J. et al. Appl. Phys. Lett. 97, 253505 (2010). 6. Slonczewski, J. C. J. Magn. Magn. Mater. 159, L1L7 (1996). 7. Berger, L. Phys. Rev. B 54, 93539358 (1996).

Salty sensations
Salt is important in health and disease, yet how mammals sense it is not completely clear. Evidence in worms suggests that TMC proteins, which are implicated in human hearing, are salt receptors involved in taste. See Letter p .95
B E R T R A N D C O S T E & A R D E M PATA P O U T I A N

alt is one of the oldest and most common food seasonings. And, together with sweet, bitter, sour and the savoury taste umami, it forms the five modalities of taste1. Salt is essential for survival, but its intake has to be tightly controlled because overconsumption impairs ion balance and regulation of blood pressure. Indeed, depending on its concentration, salt can be attractive or repulsive in many vertebrates and invertebrates. Little is known about how organisms sense the main component of salt, sodium chloride. In mice, epithelial sodium channels are required for the creatures to exhibit attractive responses to the taste associated with low salt concentrations, but not for them to show aversion to high salt concentrations2. It is exciting, therefore, that on page95 of this issue, Chatzigeorgiou etal.3 report a promising candidate for a salt sensor*. The authors explored the function of transmembrane channel-like protein-1 (TMC-1) in the nematode worm Caenorhabditis elegans. The TMC family of multipass transmembrane proteins has eight known members in humans and two (TMC-1 and TMC-2) in C. elegans. TMC-1 mutations cause deafness in humans and mice, and in the absence of both TMC-1 and TMC-2 the hair cells of the mouse auditory system cannot show mechanosensory responses4. However, it is not clear whether TMCs are bona fide ion channels and/or a component of the elusive channel
*This article and the paper under discussion3 were published online on 30 January 2013.

complex of the hair-cell mechanotransducer. Chatzigeorgiou et al. report that TMC-1 in C. elegans is expressed in a few sensory neurons, including the ASH neurons, which are important for avoidance of noxious stimuli such as nose touch (the worms reverse their direction when they bump into an object with their nose), hyperosmolarity, heavy metals, acids and high salt concentrations5. Notably, TMC-1 is detected in the sensory cilia of ASH neurons, the site of sensory transduction in these cells consistent with a role for TMC-1 in sensory detection. The researchers behavioural studies These results following deletion establish that the of the tmc-1 gene TMC-1 protein demonstrated the contributes relevance of TMC-1 specifically to in sensing salt, but sodium sensing. not other repulsive chemical or mechanical stimuli such as glycerol, copper and nose touch. They also showed that TMC-1-mediated aversion to salt is specific to sodium ions. In addition, the inability of tmc-1-mutant worms to avoid sodium chloride (NaCl) was overcome by inducing the expression of exogenous TMC-1 in ASH neurons. These results establish that TMC-1 contributes specifically to sodium sensing in ASH neurons. But is TMC-1 sufficient to account for salt sensitivity? To address this question, Chatzigeorgiou et al. induced TMC-1 expression specifically in salt-insensitive ASK neurons and found that this protein confers salt
2013 Macmillan Publishers Limited. All rights reserved

sensitivity in vivo. Remarkably, expression of C. elegans TMC-1 in four distinct mammalian cell lines induced sodium-activated conductance. The threshold of activation was about 140 millimolar NaCl, which is consistent with the NaCl concentrations that induce aversive responses in worms, mice6 and flies7 . TMC-1-induced cationic currents occur mainly in response to sodium and are insensitive to amiloride a blocker of the epithelial sodium channels1. Intriguingly, low NaCl concentrations (10150 mM) are attractive to mice, and this attraction is blocked by amiloride. However, mice exposed to higher concentrations of NaCl show aversive responses that are insensitive to amiloride2. In light of Chatzigeorgiou and colleagues data, it seems that TMC-1 constitutes an amiloride-insensitive ion-channel component that, in worms, is activated by high sodium concentrations. The role of TMC proteins in the mammalian taste system remains to be explored. Can the present data, obtained by genetic studies in worms, be reconciled with the proposal that TMC-1 and TMC-2 are part of the mechanotransduction apparatus in mammalian auditory hair cells4? Chatzigeorgiou and co-workers findings suggest that TMC-1 is not involved in worm mechanotransduction. As for TMC-2, the authors show that tmc-2 mutations do not exacerbate the effects of tmc-1 mutations on the response to sodium, which suggests that TMC-2 is not required for salt detection. Nonetheless, as tmc-2 transcripts are enriched in mechanoreceptors8, the role of TMC-2 in worm mechanotransduction as well as other sensory systems should be investigated. By demonstrating that TMC-1 encodes a putative ion channel that is activated by extracellular sodium ions, Chatzigeorgiou et al. highlight another role for TMC proteins in sensory transduction. Given that the introduction of worm TMC-1 into human cells can induce sodium-activated currents, similarly cloning and characterizing mammalian TMC proteins could provide further insight into their function. Follow-up studies should confirm whether TMC proteins are pore-forming subunits of ion channels or whether they are modulators of as yet unidentified ion channels, as well as address their physiological role in various species. Bertrand Coste is in the Ion Channels and Sensory Transduction Group, CRN2M, CNRSAix-Marseille University, Marseille 13015, France. Ardem Patapoutian is in the Molecular and Cellular Neuroscience Department, Dorris Neuroscience Center, The Scripps Research Institute, La Jolla, California, and at the Genomics Institute of the Novartis Research Foundation, San Diego, California 92121, USA. e-mails: bertrand.coste@univ-amu.fr; apatapoutian@gnf.org

4 4 | N AT U R E | VO L 4 9 4 | 7 F E B R UA RY 2 0 1 3

You might also like