TiNi as a nano-actuator: experimental verification of excitation by electron-beam heating
A. D. Johnson, M. Fanucchi, V. Gupta, V. Martynov, V.Galhotra 
TiNi Alloy Company Inc., 1621 Neptune Drive, San Leandro CA 94577

K. Clements

Innovation On Demand Inc.


High work output per unit volume makes TiNi actuators attractive as a means of powering nano-robotic movement.  Such devices are being developed for manipulation of structures at near the molecular scale.  In these very small devices (one micron scale), one means of delivery of energy is by electron beams.
Movement of mechanical structures a few microns in extent has been demonstrated in a scanning electron microscope.  Results of these and subsequent experiments will be described, with a description of potential structures for fabricating moving a microscopic x-y stage.

Miniaturization of mechanical devices is evolving toward nanometer scale, requiring handling and assembly of objects as small as a few nanometers. Manipulation of samples and specimens smaller than a few microns in size demands a technology that at present does not exist. Assemblers are needed that can grip collections of molecules, releasing them from their present location, lifting, rotating, and forcefully placing them in a new environment.

Existing micropositioners do not provide the requisite flexibility of motion for assembly tasks that are contemplated.

Forceful shape memory alloy actuators can be scaled to micron size.  These devices are thermally powered and so require a source of heat energy: this heat may be supplied by conduction, joule heating, infrared light, or other means.

We have started to explore use of a scanning electron microscope beam to provide heat energy to energize thermal actuators.  Prototype actuators are fabricated by sputter deposition of titanium-nickel thin film, photolithographic patterning, and chemical milling.  A scanning electron beam is positioned to produce local heating, and to observe the resulting motion. 


In 1982 K. Eric Drexler introduced the idea of assemblers of molecular size in his book “Engines of Creation”.[1]  Nanotechnology is the subject of at least one international conference, and one commercial venture has been organized and funded to invest research and planning in this technology. [2]  Although commercial realization of nanotechnology may be years away, there is strong indication that research following the human genome project, and particularly the study of protein structure and function, will require tools to manipulate components of the cell.  Development of these tools is a demanding, exciting, and challenging research subject.
Atomic force microscopy can be used to move individual atoms but not to grip larger objects with enough force to hold against local forces.  In recent investigations of the properties of carbon nanotubes, piezoelectric stepper motors have been used to manipulate structures orders of magnitude smaller than the drivers.[3] 
Manipulation of objects this small would be improved if the end-effectors were not much larger than the objects they control.  In analogy to the shoulder-wrist-finger arrangement of the human hand, gross positioning should be managed by actuators of macroscopic size, and fine control by end-effectors of much smaller size.
The force of actuation should be produced as close as possible to the point of application.  This implies that manipulation of sub-micron size objects requires micron-size actuators. Conventional actuators (electromagnetic, piezoelectric) do not scale well to micron size. A promising form of actuation is heat-actuated devices, particularly shape memory actuators. [4]  Photolithography provides means of fabricating devices of sub-micron size.  Miniature shape memory alloy (SMA) actuators rely on joule heating to cause the phase change.  In the sub-micron range it is difficult to make electrical connection, especially on devices that move.  To solve this problem, we  have chosen to actuate sub-micron scale shape memory alloy devices by electron-beam excitation. 

Current devices are fabricated as small as a few hundred microns using conventional microlithography.  Shrinking this technology to sub-micron dimensions raises at least two questions: (i) Will the shape memory property be preserved when the dimensions are as small or smaller than the crystal domains?  And (ii) How can such small objects (sub-micron) be selectively heated to produce actuation?  This research effort was undertaken to provide preliminary proof-of-concept answers.


Verification of Phase Change

At this conference one year ago, the question was raised: how small can shape memory thin film be shown to demonstrate the shape memory property.[5]  A partial answer to this question was obtained: a film approximately 100 nm (about 200 atomic layers) thick was shown to undergo a phase transformation as indicated by a change in resistivity.  This result is shown in Fig. 1. The change in slope and hysteresis loop are typical for TiNi shape memory alloy.

Figure 1:  Resistivity versus temperature for TiNi film sputter-deposited on silicon oxide.  Film thickness is approximately 100 nm.

Verification of shape recovery

A scanning electron microscope e-beam may be used to provide the energy, and beam steering can bring a spot of energy a fraction of a micron in diameter to bear on a sample.  How much power must be delivered to the sample, and can the beam provide this much energy in a short enough time to effect a shape change?
The estimated energy required to actuate a TiNi specimen 4 x 10 x 100 microns by heating it from the room temperature to the transition point is about 1.3 x 10 –5 joules (DT = ~80 C, DH = ~25 j/gm, Cp = ~0.3 J/gmoC, density = 6.4 gm/ cm3). The power available from the electron beam is 2x10-3 watt (for accelerating voltage ~20 KV the beam current ~10-7 A) and the estimated heating time is ~6x10-3 sec.
Demonstration of shape recovery requires that the specimen be pre-strained (stretched, compressed, or bent) while it is in its low-temperature state and then heated above the phase transformation temperature. The sample used is a fragment of TiNi film 4 micrometers thick deposited on silicon oxide, patterned with fenestrations about 40 microns in diameter, and removed from the substrate. This film was further etched to diminish the width and thickness of the web elements.  The resulting web was torn apart, producing small protrusions about 1-2 micron wide and 20 microns long.  Some of these were bent during tearing, others were deformed manually using a micromanipulator.  One such structure is shown in Figure 2.

Figure2.  Scanning electron beam image of TiNi film with fenestrations.  The wider web structures are about 20 microns wide.

The specimen was placed on a heated pedestal in an ISI60 scanning electron microscope.  The pedestal was equipped with a heater and thermocouple so that the base temperature could be controlled and measured. This fixture is shown in Figure3.  Fluke instruments were used to record temperature and current through the heater, and an IRF 640 field-effect transistor with a variable gate voltage was used to control the current through the heater to vary the temperature of the substrate.

Figure 3:   Heated sample holder for SEM.

The chamber was evacuated and the beam was started.  A picture was obtained at 1.5 kx magnification.  The sample holder was heated with resistance heater to a temperature above ambient and below the transition temperature of the TiNi.  This was to enable the electron beam to bring the temperature through a relatively small temperature change to effect the phase transformation.  The beam was centered approximately on the bent portion of the microbeam.

The SEM beam aperture was opened to impart the maximum current to the specimen, using spot mode, and current in the beam was increased. Typical current used was 70 to 100 nano-amperes measured with a Kiethley picoammeter connected between the sample and ground potential.  This current was applied to the sample for between 2 and 10 seconds.  After exposure, the beam current was reduced and further pictures taken.

The results are shown in Figure 4 (a through c).  The first picture shows the sample previous to heating.  Images b,c show the progressive actuation as successive parts of the device were heated by electron beam. Approximately 30 degrees of recovery was achieved.  The lever is about 2 microns in diameter and 20 microns long.

                        a)                                             b)                                             c)

Figure 4:Sequence of images from the SEM showing the results of heating the TiNi specimen by SEM electron beam.  Partial actuation was achieved as evidenced by motion of the tip of the lever.


The micro-cantilever moving about 30 degrees from the original position was observed in the experiment .  This was not due to thermal expansion as it did not reverse when the temperature was reduced. Thus actuation of a micro-scale device by scanning electron beam has been demonstrated, showing that the e-beam can provide enough energy to cause the phase transformation under controlled conditions.
It should be possible to construct a platform having x-y motion by placing pairs of opposed bending cantilevers in each direction,so that partial actuation of one cantilever pushes the platform while pre-straining the opposing cantilever. Larger-scale translational motion can be achieved with multiple actuators operating in sequence against a ratchet.


[1]        K. Eric Drexler, Engines of Creation: the coming era of nanotechnology, Anchor Doubleday, 1986
K.Eric Drexler, Nanosystems, Molecular Machinery, Manufacturing, and Computation, John Wiley & Sons, Inc., 1992
[2]        James R. von Ehr, Zyvex “the first nanotechnology development company”, http://www.zyvex.com
[3]        MF Yu, M. J. Dyer, G.D. Skidmore, H.W. Rohrs, XK Lu, K.D. Ausman, J.R. Von Ehr,R..S. Ruoff, 3 Dimensional Manipulation of Carbon Nanotubes under a Scanning Electron Microscope, Sixth Foresight Conference 1998.

[4]        A. D. Johnson, “Vacuum-Deposited TiNi Shape memory Film: Characterization and Applications in Micro-Devices,” J. Micromech. Microeng. 1(1991) 34-41.

P. Krulevitch, A.P. Lee, P.B. Ramsey, J.C. Trevino, J. Hamilton, M.A. Northrup, “Thin film Shape Memory Alloy Microactuators, J. MEMS, vol. 5, No. 4, December 1996. 

(They show that SMA has the highest work output per unit volume of any actuating technology.)

[5]        Deepak Srivastava, NASA Ames research Center, Moffett Field, 650 604 3468 deepak@nasa.gov; private communication to Vikas Galhotra at TiNi Alloy Company.