Capability of Sputtered Micro-patterned NiTi Thick Films

Christoph Bechtold1 • Rodrigo Lima de Miranda1,2 • Eckhard Quandt2

Published online: 4 August 2015

� ASM International 2015

Abstract Today, most NiTi devices are manufactured by

a combination of conventional metal fabrication steps, e.g.,

melting, extrusion, cold working, etc., and are subsequently

structured by high accuracy laser cutting. This combination

has been proven to be very successful; however, there are

several limitations to this fabrication route, e.g., in respect

to the fabrication of more complex device designs, device

miniaturization or the combination of different materials

for the integration of further functionality. These issues

have to be addressed in order to develop new devices and

applications. The fabrication of micro-patterned films using

magnetron sputtering, UV lithography, and wet etching has

great potential to overcome limitations of conventional

device manufacturing. Due to its fabrication characteris-

tics, this method allows the production of devices with

complex designs, high structural accuracy, smooth edge

profile, at layer thicknesses up to 75 lm. The aim of this

study is to present recent developments in the field of NiTi

thin film technology, its advantages and limitations, as well

as new possible applications in the medical and in non-

medical fields. These developments include among others

NiTi scaffold structures covered with NiTi membranes for

their potential use as filters, heart valve components or

aneurysm treatments, as well as micro-actuators for con-

sumable electronics or automotive applications.

Keywords NiTi � Shape memory � Magnetron

sputtering � Lithography � Medical applications �Integration of functionality

Introduction

NiTi is a well-established alloy in the medical industry. It

is used in orthodontic applications, for a variety of medical

instruments or for temporary or permanent vascular

implants [1–4]. Its superelastic properties have numerous

advantages, among them constant force delivery, kink

resistance, and different deployment methods [5]. The

latter allow for the use of smaller catheter diameters to

deliver implants and instruments, which is in particular

important for minimal invasive treatments. The increasing

popularity of vascular disease treatments utilizing stents in

the 1990s and the technological ability to fabricate and

structure high-quality NiTi tubes was a main trigger for the

rise of NiTi in medical applications [6]. Stent markets have

been growing since, and NiTi is still one of the main alloys

in this field.

The conventional fabrication route for NiTi implants

comprises a series of processing steps, including melting,

extrusion, hot or cold drawing, laser cutting, and surface

finishing [7–9]. This mass fabrication route has been pro-

ven to be very successful in providing millions of medical

products with high-quality standards. An ongoing trend in

medical device development is further miniaturization of

the medical device while enhancing its overall function-

ality allowing for less invasive treatments and shorter

healing times. This trend is a challenge for the established

This article is an invited paper selected from presentations at the

International Conference on Shape Memory and Superelastic

Technologies 2014, held May 12–16, 2014, in Pacific Grove,

California, and has been expanded from the original presentation.

& Christoph Bechtold

bechtold@acquandas.com

1 Acquandas GmbH, Kaiserstrasse 2, 24143 Kiel, Germany

2 Inorganic Functional Materials, Christian Albrechts-

Universitat zu Kiel, Kaiserstrasse 2, 24143 Kiel, Germany

123

Shap. Mem. Superelasticity (2015) 1:286–293

DOI 10.1007/s40830-015-0029-9

fabrication route. Some of these issues can be addressed

using an approach where NiTi structures are fabricated by a

sequence of microsystem technology processes, i.e., mag-

netron sputtering, UV lithography, and wet chemical

etching. This process is described in detail by Lima de

Miranda et al. [10]. This fabrication route has gained

increasing interest in recent years, especially since the

maximum thickness of thus prepared structures exceeds

75 lm and is therefore in a suitable range for many med-

ical applications. Magnetron sputtering is a vacuum

deposition technique. Inside a high vacuum chamber, Ar-

ions are accelerated towards a target, in our case a cast-

melted NiTi alloy target, target atoms are released due to

the ion bombardment and deposit on a cold substrate sur-

face. Oxide and carbide inclusions that are present in cast-

melted NiTi alloys with sizes of a few micrometers are not

transferred to the deposited film. Thus, a main source of

mechanical failure is eliminated, since oxide and carbide

inclusions, in particular in combination with voids, act as

crack initiation sites and decrease fatigue life [11]. Also,

the corrosion resistance of the material increases signifi-

cantly and exceeds 1000 mV [12]. The biocompatibility of

sputtered NiTi films has been subject of several studies,

which certify excellent properties [13, 14]. The lateral

homogeneity of the deposited films in terms of thickness

and chemical composition depends on the sputtering sys-

tem and on the target homogeneity. Many different sput-

tering systems exist. Small development systems with

static substrate and static magnetron can be used for pro-

totype fabrication, but the area of homogeneous thickness

and composition on the substrate may be small. Lager

systems with larger targets and magnetrons, moving sub-

strates or moving magnetrons yield better homogeneity. In

Fig. 1a, b, the lateral thickness and compositional homo-

geneity of a binary NiTi film on a 600 substrate fabricated

within a production system are shown. The thickness varies

by ±1 % over an area of 150 mm in diameter (measured

with an inductive digital comparator Mahr Extramess

2000), chemical composition was measured by EDX (EDX

error for each data point approx. 0.5–1.0 at.%). Since this

is a wafer-based technology, it is highly scalable and

reproducible.

NiTi films that are deposited at room temperature are

amorphous and require a subsequent heat treatment for

crystallization [15]. Hence, the microstructure of sputter-

deposited NiTi differs to that of bulk materials which

crystallize from the melt and often experience a high

degree of cold work to reduce the dimension of the semi-

finished product. Microstructural investigations of sput-

tered films have been carried out for a large variety of

binary to quaternary compositions and heat treatments,

e.g., [16–22], which demonstrates the large influence of the

two parameters on microstructural features (grain size,

distribution, type, and size of precipitates), transformation

temperatures and consequently their functional properties.

Using sputtering techniques, alloy optimization by varying

the chemical composition can be achieved in numerous

ways, ranging from different target compositions, changing

sputtering power or target-substrate distances to complex

combinatorial methods where a large area of the compo-

sition spread can be investigated simultaneously by high-

throughput methods [22–24].

Standard UV lithography allows for structuring photo-

sensitive resist with feature sizes in the submicron range

[25]. The pattern of the structured resist is then transferred

to the NiTi film, e.g., using wet chemical etching. NiTi

films with thicknesses of 75 lm have been deposited fol-

lowing the process described in [10]. Since sputtered edges

are not perfectly vertical, the minimum feature size

increases with increasing film thickness, i.e., *1 lm for

Fig. 1 a Thickness homogeneity over a 600 wafer area, nominal thickness 70 lm ± 1 %. b Chemical homogeneity over a 600 wafer area

Shap. Mem. Superelasticity (2015) 1:286–293 287

123

very thin NiTi films, and *10 lm for film with 75 lmthickness. Thus, aspect ratios of up to 7–8 can be achieved.

This is in particular interesting for the fabrication of thin

mesh structures where maximum pore or strut size can be

in the range of a few tens of micrometers only.

The NiTi layer can be released from the substrate using

a sacrificial layer technique. When flat substrates are used,

the released, amorphous film is obviously also flat. The

cylindrical shape of vascular implants can be achieved by a

constraint heat treatment, during which the amorphous film

crystallizes within a quartz tube of the desired diameter.

For many applications, an open design (beginning and end

of circumference are not joint) may be feasible, for others

the joint circumference may be required. Pulsatile fatigue

investigations on stent structures with laser-welded joints

of the circumference showed that the fatigue resistance of

the micro-welded joints is sufficiently high, and fractures

were rather observed at stent struts than the welding tags

[26]. Also, the overall functional behavior of NiTi sheet

metal was found to be mainly unaffected by welding, since

no significant change in the stress–strain transformation

plateau was detected [27]. The necessity of jointing can be

avoided when depositing NiTi directly on cylindrical sub-

strates, e.g., using a hollow cathode setup [28] or a rotating

substrate approach [29]. The latter is economically less

interesting due to the significant decrease in the effective

sputtering rate.

Except for laser-welded joints, structured, sputtered

NiTi films exhibit no heat-affected zones (HAZ), in con-

trast to laser-cut tubes. The microstructure is hence less

affected which in combination with the lack of oxide and

carbide inclusions promises higher fatigue life and

improved mechanical properties. Investigations on the

fatigue life of sputtered diamond-shaped samples reveal

indeed a high fatigue endurance limit. Up to mean strains

of 6 %, maximum alternating strains of ±1.5 % did not

lead to a fracture of the specimen during 3 9 10E6 cycles,

a significantly higher value compared to ±0.5 % maximum

alternating strain for NiTi standard material [26].

Limitations of the NiTi sputtering technology are the

limited film thickness and lateral dimension (the latter

determined by the wafer size used, e.g., 600, 800, etc.,), thenecessity to utilize the usable area effectively in order to be

cost efficient (good if many parts fit into the wafer area),

and the lack of cold work as parameter to influence

microstructure and thus functional properties.

The unique features of NiTi sputtering, however, have

high potential to realize some of the next generation medical

implants or tools, where the current technology cannot pro-

vide, e.g., the freedom of design or the ability to combine and

structure a series of additional materials. The capability of

NiTi sputtering is discussed in the following chapter.

Results and Discussion

We have divided this chapter according to what we have

named NiTi sputter technology platforms, namely 2D, 2.5-

dimensional (2.5D), and multi-layered structures. For each

platform, the fabrication process is adjusted in order to

realize different features of a NiTi-based sample.

The 2D platform is the base platform; it describes the

arbitrary, two-dimensional structuring of a NiTi film with

homogeneous thickness up to 75 lm. The characteristics of

the produced samples have been mentioned before: feature

resolution dependent on film thickness, but below 10 lmfor 75 thick films, aspect ratios up to 7–8, thickness and

compositional homogeneity of ±1 % and ±0.4 at.%,

respectively (depending on the sputtering system and target

homogeneity), highly pure material, excellent fatigue

properties, possibility of shape setting, and lateral dimen-

sion limited to wafer size. Examples are shown in Fig. 2.

Figure 2a shows an 8-lm-thick mesh structure with

8 lm strut width and 75 9 20 lm2 diamond-shaped pore

size. Strut width and pore size can be optimized according

to application requirements, e.g., when used as embolic

filters or as surfaces to engineer tissue growth. Figure 2c, d

shows a multitude of parallel, fine lamellae connected to a

supporting beam; Fig. 2e, f a connector structure; and

Fig. 2g, h generic, shape set stent structures of different

diameter.

2.5D structures offer a limited freedom of design per-

pendicular to the film plane. This can be achieved in two

ways. Either the film thickness is homogeneous, but lens-

shaped cavities are patterned into the structured films, see

Fig. 3.

A different approach is shown in Fig. 4 where a second

NiTi layer was deposited and patterned on top of a struc-

tured NiTi base layer. In this case, a 40-lm-thick ring was

structured on top of a 5-lm-thick layer. Within the prior

given limitations, the dimension of each layer can be varied

freely. Hence, reservoirs can be created, e.g., for drug

eluting systems or for micropump systems. Volume and

shape of the reservoirs can be varied in wide range to adjust

for the specific requirements of the application. Using the

second approach novel components with unique properties

can be created.

Figure 5a–c shows a supporting structure (50 lm thick)

with a fine mesh (5 lm thick, 75 9 20 lm2 diamond-

shaped pore size) between the supporting beams. Despite

its stiffness, these kinds of superelastic structures can be

crimped into a catheter of 3 mm without breaking. Fig-

ure 5e, f shows a micro-patterned strut surface, and

Fig. 5g, h stent-like structures with flaps attached.

The application potential of the 2.5D platform in the

medical field is manifold: micro-patterned surfaces for

288 Shap. Mem. Superelasticity (2015) 1:286–293

123

drug eluting systems, tissue engineering or friction control,

robust filters with scaffolds, artificial heart valves, or flow

diverters for aneurysm treatments can be realized.

The third platform involves the deposition and struc-

turing of other materials on a NiTi structure. An example is

the deposition of a radiopaque material, preferentially a

heavy metal such as Ta, Au, or Pt–Ir [4] on a NiTi struc-

ture. In Fig. 6a, a SEM image of a 50-lm-thick stent-like

structure with 35 lm strut with coated with 10 lm Ta is

shown, the corresponding radiopacity results are shown in

Fig. 6b: areas 1–6 contain samples covered with Ta, areas

7–9 uncovered samples of identical NiTi thickness. Areas

2, 3, 5, 6, 7, and 8 contain stent structures, and areas 1, 4,

and 9 diamond-shaped control specimen.

The NiTi structure can be completely or partly coated.

In the former case, the thickness of the radiopaque material

has to be sufficiently high to increase radiopacity but small

enough not to hinder the stress-induced martensitic trans-

formation of the underlying NiTi. Also, delamination of the

radiopaque layer must be avoided during mechanical

loading. In the second case, the radiopaque layer can be

patterned so that heavily loaded areas are uncoated and

thus free to transform, whereas less loaded areas are coated

(Fig. 6c).

Fig. 2 Examples of 2D structures: a, b fine mesh structures, c, d thin lamellae connected to a supporting beam, e, f connector structures, g,h shape set generic stent structures

Fig. 3 Sputtered NiTi film with

patterned lens-shaped cavities

Shap. Mem. Superelasticity (2015) 1:286–293 289

123

A novel field is the combination of flexible NiTi sub-

strates with structured isolating and conducting layers.

Circuit paths can be deposited on NiTi with a top and

bottom isolating layer. The top isolation might be absent in

certain areas so that electrodes for stimulation, sensing, or

mapping can be realized.

So far we have considered mostly superelastic materials

for medical applications. On the other hand, Ti-rich films

exhibiting the shape memory effect can be used as minia-

turized actuator components (Fig. 7). Due to the high work

output of NiTi small, flat, lightweight actuators can be

realized for applications within the automotive or con-

sumable electronics industry. Since the 2D design of the

actuators is free to choose, optimized force-stroke rela-

tionships and additional design elements to facilitate sys-

tem integration of the actuator component (instead of wire

crimping) can be realized.

The lack of oxide and carbide inclusions in sputtered

films promises good fatigue life also for actuator

applications.

Often Joule heating is used as a simple heatingmechanism

to increase the sample temperature above the austenite finish

temperature. Using the combination with isolating and

conducting layers, local heating elements can be created on

the NiTi surface and the NiTi actuator structure can be

activated in selected areas only, if required.

Fig. 4 Patterned NiTi rings of

40 lm thickness on top of a

patterned NiTi film

Fig. 5 Examples for 2.5D structures: a–d fine, thin mesh structure in between thick supporting beams to enhance crimpability, e, f stent-likestructure with micro-patterned dots on the strut surface, g, h stent-like structure with flaps attached for flow diversion in aneurism treatment

290 Shap. Mem. Superelasticity (2015) 1:286–293

123

Summary and Conclusion

The present paper investigates the capability of magnetron

sputtering, UV lithography, and wet chemical etching to

fabricate different NiTi-based components.

The characteristic features of sputtered micro-patterned

NiTi films (feature resolution, maximum thickness, thick-

ness and compositional homogeneity, certain microstruc-

tural aspects, fatigue performance) have been discussed

briefly. The main focus of this paper, however, is not on the

functional properties or the microstructure of sputtered

films, but the freedom of design that can be realized with

the combination of the above-named microsystem tech-

nology processes. In particular, 2.5D structures offer in our

view novel design opportunities in the medical field for

drug eluting systems, filters, valves, tissue engineering, or

flow diverters. Prototypes of 2D and 2.5D structures can be

fabricated with only few adjustments (different lithography

masks) required. The integration of additional functionality

using radiopaque, electrically conducting or insulating, or

magnetic layers is a promising approach to design and

fabricate next generation medical devices.

Experimental

The underlying process to fabricate NiTi films presented in

this paper is a combination of three microsystem technol-

ogy processes: UV lithography, physical vapor deposition,

and wet etching as described in [10]. In order to obtain

Fig. 6 a Stent strut coated with

Ta, b radiopacity results for

coated and uncoated samples,

c partial strut coating with Ta

Fig. 7 a Plain tensile actuator

design with features to facilitate

system integration, b disk-

shaped actuator

Shap. Mem. Superelasticity (2015) 1:286–293 291

123

more complex geometries such as 2.5D structures or

structures that are combined with other materials, some of

the above process steps are repeated or slightly altered. The

NiTi films are deposited using magnetron sputtering devi-

ces at base pressures below 1 9 10-7 mbar, Ar gas flow,

and deposition rates between 3.8 and 5.6 nm/s, depending

on sputtering system.

The processing sequence (Fig. 8) starts with the depo-

sition of a Cu sacrificial layer and a NiTi seed layer.

Substrates are then coated with positive photoresist

(AZ1518) with 2.3 lm thickness using spin coating fol-

lowed by UV exposure on a Karl Suss mask aligner in soft

contact mode for 2 s, a soft bake on a 105 �C hotplate for

2 min and rehydration for another 2 min. The resist is then

developed using AZ716 MIF solution for 1 min, forming

the desired geometry. Hard bake is performed at 120 �C for

20 min. The third step consists of wet etching the NiTi

seed layer using a HF solution with an etching rate of

10 nm/s. In the fourth step, the sacrificial layer is wet

etched which results in mushroom-like structures due to

undercutting during isotropic etching of the sacrificial

layer. As selective etchant, a standard BASF Selectipur

Chromium Etch (etching rate 13.5 nm/s) is used. After

etching, the photoresist is removed using acetone. To

achieve the required Nitinol film thickness, a thick NiTi

layer is now deposited on top of the seed layer. The

mushroom-like structures allow for the growth of the

structured NiTi film while avoiding any coalescence with

adjacent areas in between the mushroom-like structures. As

a final step, the sacrificial layer is removed, again using

BASF Selectipur Chromium Etch, which results in a free-

standing micro-patterned NiTi structure.

Subsequently, the amorphous samples were crystallized

in a high vacuum chamber in order to avoid oxidation

during the annealing process. Heat treatment was carried

out ex situ by means of a rapid thermal annealing system.

The halogen-lamp driven heating chamber enables typical

heating rates of 50 K/s in a vacuum environment of about

10-6–10-7 mbar. The annealing temperature was held

constant for 10 min at the maximum temperature of

650 �C, which turned out to be sufficient for crystallization

of amorphous NiTi. In a second step, films were annealed

for further 10 min at 450 �C, which is a common procedure

for Ni-rich Nitinol samples to induce the formation of

Ni4Ti3 precipitates in order to adjust phase transformation

temperatures.

Acknowledgments We thank Dr. C. Zamponi for EDX measure-

ments and Dr. Jens Trentmann at the department of Radiology and

Neuroradiology at the UKSH Kiel for radiopacity investigations.

References

1. Thompson SA (2000) An overview of nickel–titanium alloys

used in dentistry. Int Endod J 33:297

2. Stockel D (2000) Nitinol medical devices and implants. Minim

Invasive Ther 9(2):81–88

3. Pelton A, Schroeder V, Mitchell MR, Gong X-Y, Barney M,

Robertson SW (2008) Fatigue and durability of Nitinol stents.

J Mech Behav Biomed Mater I 1(2):153–164

4. Stockel D, Pelton A, Duerig T (2009) Self-expanding Nitinol

stents for the treatment of vascular disease. In: Yoneyama T,

Miyazaki S (eds) Shape memory alloys for biomedical applica-

tions. CRC Woodhead Publishing, Boca Raton, pp 237–256

5. Duerig T, Pelton A, Stockel D (1999) An overview of Nitinol

medical applications. Mater Sci Eng A 273:149–160

Fig. 8 Process steps for the production of freestanding micro-patterned NiTi 2D structures

292 Shap. Mem. Superelasticity (2015) 1:286–293

123

6. Pelton A, Duerig T, Berg B, Hodgson D, Mertmann M, Mitchell

M, Proft J, Wu M, Yang J (2005) Nitinol medical devices. Adv

Mater Proc 163(10):63

7. Hodgson D, Russell S (2000) Nitinol melting, manufacture and

fabrication. Minim Invasive Ther Allied Technol 9(2):61

8. Favier D, Orgeas L, Ferrier D, Poncin P, Liu Y (2001) Influence

of manufacturing methods on the homogeneity and properties of

nitinol tubular stents. J Phys IV 11(PR8):541

9. Poncet P, Adler PH, Carpenter S, Wu MH (2003) Manufacture of

Nitinol tubing. In: ASM Materials & Processes for Medical

Devices Conference, Anaheim

10. Lima de Miranda R, Zamponi C, Quandt E (2013) Micropat-

terned freestanding superelastic TiNi films. Adv Eng Mater

15(1–2):66

11. Rahim M, Frenzel J, Frotscher M, Pfetzing-Micklich J, Steeg-

muller R, Wohlschlogel M, Mughrabi H, Eggeler G (2013)

Impurity levels and fatigue lives of pseudoelastic NiTi shape

memory alloys. Acta Mater 61:3667

12. Wohlschlogel M, Steegmuller R, Schussler A (2014) Nitinol:

tubing versus sputtered foil—microcleanliness and corrosion

behavior. In: SMST Conference 2014, Pacific Grove

13. Kealey CP, Whelan SA, Chun YJ, Soojung CH, Tulloch AW,

Mohanchandra KP, DiCarlo D, Levi DS, Carman GP, Rigberg

DA (2010) In vitro hemocompatibility of thin film nitinol in

stenotic flow conditions. Biomaterials 31(34):8864

14. Habijan T, Lima De Miranda R, Zamponi C, Quandt E, Greulich

C, Schildhauer TA, Koller M (2012) The biocompatibility and

mechanical properties of cylindrical NiTi thin films produced by

magnetron sputtering. Mater Sci Eng C 32:2532

15. Ishida A, Sato M, Takei A, Miyazaki S (1995) Effect of heat

treatment on shape memory behavior of Ti-rich Ti–Ni thin films.

Mater Trans JIM 36(11):1349–1355

16. Miyazaki S, Ishida A (1999) Martensitic transformation and

shape memory behavior in sputter-deposited TiNi-base thin films.

Mater Sci Eng A 273–275:106–133

17. Ishida A, Sato M, Kimura T, Miyazaki S (2000) Stress-strain

curves of sputter-deposited Ti–Ni thin films. Philos Mag A

80(4):967–980

18. Otsuka K, Ren X (2005) Physical metallurgy of Ti–Ni-based

shape memory alloys. Prog Mater Sci 50:511

19. Ishida A, Sato M (2006) Microstructure and shape memory

behaviour of annealed Ti51.5Ni(48.5-x)Cux(x=6.5-20.9) thin films.

Philos Mag Lett 86:13–20

20. Ishida A, Sato M, Gao ZY (2013) Properties and applications of Ti–

Ni–Cu shape-memory-alloy thin films. J Alloys Comp 577(S1):S184

21. Zarnetta R, Takahashi R, Young ML, Savan A, Furuya Y,

Thienhaus S, Maaß B, Rahim M, Frenzel J, Brunken H, Chu YS,

Srivastava V, James RD, Takeuchi I, Eggeler G, Ludwig A

(2010) Identification of quaternary shape memory alloys with

near-zero thermal hysteresis. Adv Funct Mater 20(12):1917

22. Zarnetta R, Buenconsejo PJ, Savan A, Thienhaus S, Ludwig A

(2012) High-throughput study of martensitic transformations in

the complete Ti–Ni–Cu system. Intermetallics 26:98

23. Cui J, Chu Y, Famodu O, Furuya Y, Hattrick-Simpers J, James D,

Ludwig A, Thienhaus S, Wuttig M, Zhang Z, Takeuchi I (2006)

Combinatorial search of thermoelastic shape-memory alloys with

extremely small hysteresis width. Nat Mater 5:287

24. Zarnetta R, Savan A, Thienhaus S, Ludwig A (2007) Combina-

torial study of phase transformation characteristics of a Ti–Ni–Pd

shape memory thin film composition spread in view of

microactuator applications. Appl Surf Sci 254(3):743–748

25. Madou MJ (2002) Fundamentals of microfabrication, 2nd edn.

CRC Press LLC, Boca Raton

26. Siekmeyer G, Schußler A, Lima de Miranda R, Quandt E (2014)

Comparison of the fatigue performance of commercially pro-

duced Nitinol samples versus sputter-deposited Nitinol. JMEPEG

23:2437

27. Maletta C, Falvo A, Furgiuele F, Barbieri G, Brandizzi M (2009)

Fracture behaviour of Nickel–Titanium laser welded joints.

J Mater Eng Perform 18:569

28. Marton D, Boyle CT, Wiseman RW, Banas CE (2010) High

strength vacuum deposited Nitinol alloy films and method of

making same. US Patent 8083908 B2

29. Lima de Miranda R, Zamponi C, Quandt E (2009) Fabrication of

TiNi thin film stents. Smart Mater Struct 18:104010

Shap. Mem. Superelasticity (2015) 1:286–293 293

123