Qin Y. Micromanufacturing Engineering and Technology

Подождите немного. Документ загружается.

MICRO-ELECTROPLATING

TECHNOLOGY

Electroplating is the key step in the fabrication of

metallic micro-components and tools such as

masks and molding tools . Established routine

processes are often referred to by simply giving

the name of the plating solution. However, every

user performs the basic process with differences in

electrolyte formulation and operation due to spe-

cific fabrication environments such as the plating

apparatus or specific material properties. Never-

theless, some general aspects in micro-electroplat-

ing will be pointed out next. From the standpoint

of a lithographic pattern, an ideal plating solution

has the following properties:

the resist structures are not changed (no swell-

ing, no thermal loading introduced);

the mechanical stress remains very low (a few

tens of MPa);

the grain structure does not change with height;

small and large areas grow at the same rate.

Some limiting aspects arise from the desire to

plate fairly thick layers in excess of 10 mm. This is

usually referred to as electroforming and many

commercial solutions can already be excluded.

The complexity of LIGA electroplating can best

be described using a sketch of the current-density

vs. voltage relationship (Fig. 13-9) [24].

At the onset of the net current density (1 and 2),

therateislimitedbythekineticsatthesurface.The

growth rate depends exponentially on the voltage,

so this is not a robust working point. Furthermore,

this corresponds to very low growth rates and unac-

ceptably long plating times. In region 4, the current

density is limited by diffusion of some species of

which the concentration at the cathode becomes

zero. To keep the diffusion zone small, both strong

convection and high ion concentrations in the elec-

trolyte are usually employed. Both aspects will raise

the diffusion-limited current density.

From the electroplating point of view, the lith-

ographic pattern is a nuisance. First, it distorts the

flow across the surface (Fig. 13-10(a)). On the one

hand, eddy currents may form and on the other

hand, convection into the structures, particularly

for high aspect ratios, may be completely inhib-

ited. In the latter case, the diffusion length corre-

sponds to the total resist height while, for good

convection, the diffusion length may be as small as

a few micrometers. This means that locally, the

current-density-potential relationship will vary

TABLE 13-3

Deep X-ray Software Packages Available: (a) http://www.ipal.sandia.gov/ip-

details.php?ip=4874; (b) http://www.kit.edu; and (c) tabata@se.ritsumei.ac.jp

Program & Platform Status Possibility

LEX-D

Dos Commercial Source: bending magnet

Optics: mirror, double

mirror, beam-stop

Dose: primary and

secondary

Development: 4 dimensions

DoseSim

Windows Freeware Source: bending magnet

Optics: mirror, double

mirror, beam-stop

Dose: primary

Development: 1 dimension

X3D

LINUX Freeware Source: bending magnet

Optics:

Dose: primary

Development: 4 dimensions

CHAPTER 13 Deep X-Ray Lithography 211

significantly. Second, the pattern distorts the

homogeneity of the electrical field, with stronger

fields corresponding to greater potentials on the

current-density-potential curve. This effect is

sketched in Fig. 13-10(b).

The ar t of elec troplating therefore involves

finding th e right chemic al composition tha t gives

the desired properties of the deposited layers and

mak ing/keeping the s olution stable against

changes associated with the electrochemical reac-

tions during plating. The lithographic pattern

introduces the requirement of low sensitivity to

local variations in potential and local var iations

in mass transport. The use of additives alters the

per formance of the plating process drastically.

Additives may be employed to change the

current-density-potential curve, where a leveler

typically makes the curve near to the operating

poi nt flatter, an inhibitor reduces the current

density at high voltages and a brightener reduce s

the potential for the same current density by

facilitating the nucleation of new grains. The

finer grains usual ly are rela ted to a great er hard-

nes s of the deposit. The electroplati ng procedures

used in the LIGA pr ocess chain are gold plating

for X-ray mask fabrication and nickel plating for

molding tools and metallic micro-components.

Further alloys may be deposited in resist molds

such as Ni-Co, Ni/P or Ni-Fe. Different plating

solution properties (bath temperature; composi-

tion; thickness accuracy; aspect ratio), which

produce a range of hardnesses of between 280

and 680 Vicker s, are listed in Table 13-4.

FIGURE 13-10 (a) Mass-transport phenomena in micro-

electroplating. The lithographic pattern may have a strong

influence on the length of the diffusion zone which, in turn,

will result in locally differing current densities, and (b) inho-

mogeneous field distribution due to the lithographic pat-

tern. Isolated small features focus the field lines and grow

faster than wider structures. The same effect leads to higher

rates near to resist edges.

FIGURE 13-9 General current-density-potential relationship [24].

212 CHAPTER 13 Deep X-Ray Lithography

The parts are typically ‘overplated’ to ensure

complete filling; an electroplating step is not able

to deliver a homogeneous thickness all over the

substrate and a polished surface, which is sometimes

needed. The top surface (electroplating front) is typ-

ically lapped and polished to the final part thickness

and aspect. Other techniques are also available:

1. Ultra-precision milling. This involves cutting

with a milling tool while the workpiece is

rotating. Using natural-diamond cutting

tools, a broad range of materials such as, for

example, non-ferrous met als (NiCo, NiP, Ni,

Au, etc.) or plastics, can be machined with a

surface roughness of below 5 nm R

. Only the

diamond can be sharpened to the required

level of accuracy. Cutting-edge sharpness

and roundness are crucial to the manufactured

quality of the workpiece.

2. Turning. This involves cutting with a rotating

milling tool while the workpiece is fixed.

Actually, only hard Au can be turned using

this technique;

3. Grinding and polishing. The polishing can be

effected electrolytically or chemically. All of the

techniques should meet the following criteria:

A. to deliver a good thickness-accuracy

(5 mm);

B. to avoid burr formation;

C. to deliver a specified surface quality (for

example, optical quality).

DESIGN RULE – MASK

TECHNOLOGY

A complete LIGA process database does not yet

exist. Nevertheless, the LIGA centers (CAMD,

BESSY, IMT) have accum ulated working desi gn

rules that enable them to advise customers on the

feasibility of their needs. As an example, numer-

ous parameters influence the quality of the final

parts in terms of lateral dimensions, and rough-

ness of the side wall, some of the most important

being listed here [25,26]:

Secondary radiation. Diverg ence of the elec-

trons and photons; fluorescence emitted by

the membrane, the absorber and the substrate;

secondary electrons emitted by the membrane,

the absorber and the substrate; and the Fresnel

diffraction; it will deposit a certain dose in the

unwanted parts of the resist.

Thermal distortion. The mask, resist and sub-

strate get heated during irradiation. This

leads to thermal distortion which affects the

TABLE 13-4

F&S Plating Solution Characteristics

Property Nickel Nickel/Cobalt Nickel/Iron

Composition 100 80/20 95/5

Bath temp. 40



C40



C52



Plating rate 12 mm/h (1 A/dm

)12mm/h (1 A/dm

)10mm/h (1 A/dm

)

Hardness (Vickers) 280–330 (0.1) 450–500 (0.1) 580–630 (0.1)

Thickness accuracy 50 mm 50 mm 100 mm

FIGURE 13-11 Different mask technology.

CHAPTER 13 Deep X-Ray Lithography 213

accuracy of the copy. Cooling of the mask and

the substrate is very important.

Swelling and thermal expansion of the resist.

X-ray masks are not available from commercial

mask shops as is the case for Cr masks. Therefore

the LIGA centers have developed their own tech-

nology to produce them. A variety of options exist

(see Fig. 13-11 ) requiring different tools using

different mask membranes providing different

performances for different costs (see Table 13-5).

Nevertheless, a standard exists; it is limited to the

characteristics of the support ring. X-ray masks

consist of absorber patterns (generally Au)

supported by highly X-ray transparent membranes,

the characteristics of which are given in Table 13-5.

In Fig. 13-12 is shown a standard X-ray mask

(the membrane being a polished carbon mem-

brane, the layout area having a diameter of

80 mm).

METROLOGY

Micro-system technology requires 3D coordinate

measurements to be performed with measurement

precision and accuracy to within 0.1 mm. Two-

dimensional parts can be measured using

TABLE 13-5

Typical Mask-Blank Characteristics

Material Typical

Thickness

(mm)

Young’s

Modulus

(GPa)

Thermal

Expansion

Coefficient

(10

–6

/K)

Thermal

Conductivity

(W/mK)

Density

(g/cm

)

Price

Titanium 2 116 9 22 4.5 ++

Silicon 100 240 2.3 157 2.32 –

Beryllium 500 318 12 230 1.85 ++

Polished

carbon

150 11 8.8 96 1.8 –

Vitreous carbon 200 28 2.6 6.3 1.4 –

Diamond 30 30 1.2 1000 3.51 ++

FIGURE 13-12 A standard X-ray mask (25 m m Au absorber supported by a 160 mm polished carbon membrane).

214 CHAPTER 13 Deep X-Ray Lithography

one-dimensional measurement devic es. Three-

dimensional parts require 2D information, or a

picture, to describe them. A gear, for instance,

requires multiple measurements, including hub

diameter, pitch, addendum and dedendum circles,

involutes, top- and bottom-surface finish, side-

wall finish, side-wall angle and tooth thickness,

radius of curvature, and circularity. In addit ion,

for reliability, information about the internal

structure of the part is of great importance.

Metrology for micro-features is one of the most

active areas in dimensional metrology. Different

techniques are possible [27]; the most commonly

used being the employment of a coordinate-

measurement machine (CMM) equipped with an

optical-fiber probe [28,29].

UNIQUE FEATURES OF THE LIGA

PROCESS

Features charac terizing this process are listed

below:

1. As a result of their high energy, these

X-rays are c apable of deeply penetrating

thick (e.g. hundreds of micrometers or

even millimeters) layers of polymeric resist,

allowing the formation of tall micro-

structures in one exposure step. Very pre-

cise shape definition of parts, both laterally

in terms of dimensional control and in term

of the straightness and planarity of side

walls, are available.

2. The short wavelengths of X-ray photons pro-

vide high resolution for patterning due to low

diffraction effects. Smallest lateral dimension

of a few micrometers with structural details

in the sub-micrometer range can be manu-

factured.

3. The very small vertical angular divergence of

the X-ray beam achieves high accuracy in pat-

tern transfer from the mask. Due to their

excellent collimation, the X-rays penetrate

thick resists with extremely low horizontal

run-out (less than 0.1 mm/100 mm thickness),

thereby producing the substantially vertical

walls for which LIGA structures are well

known.

4. The almost parallel (well-collimated) light of X-

ray beams produced by synchrotron radiation

sources also allows printing with a large depth-

of-field. A large working gap between the mask

and the substrate can then be used in non-tra-

ditional pattern transfer as for the manufacture

of slanted structures or for pattern formation

on substrates presenting a large topography.

5. The vertical side walls are optically smooth

with a typical local roughness of the order of

10 nm an d longer-range waviness such as

slope errors or steps determined solely by the

accuracy of mask writing.

MARKET SITUATION OF DEEP

X-RAY LITHOGRAPHY

As mentioned earlier, the LIGA technique offers the

possibility to manufacture micro-structures with a

number of unique features. With these properties,

LIGA is on the leading edge of micro-fabrication.

This is a well-known issue to most participants in

the micro-sector. However, most of them consider

LIGA to be very expensive, to employ much time

from design-to-realization, and to invoke a number

of quality problems. They therefore try to avoid

LIGA, and only consider LIGA if the alternative

fabrication methods fail to fulfill the requirements.

These requirements, however, are often very chal-

lenging, even for LIGA, resulting in long develop-

ment times, high failure rates, and, of course, high

cost. This reinforces the above-described percep-

tion, resulting in a sort of vicious circle.

The LIGA road-map (see Fig. 13-13)defined

by IMT (Institut f

€

ur Mikrostrukturtechnik

FIGURE 13-13 Road-map to commercialization.

CHAPTER 13 Deep X-Ray Lithography 215

Karlsruhe Institute of Technology) consists of

continuous process innovation and the ability

to continually improve internal processes in the

direction of greater economy, productivity and

cost effectiveness, so that products can be offered

at lesser cost. Products that are complex and

intricately designed are often produced initially

in a prototype, low volume mode. This is espe-

cially true if the end product is very expensive. A

project named MODULIGA (modular produc-

tion method using the LIGA process) [30] linked

together the a ctivitie s of the German LIGA cen-

ters:IMT,IMM(Institutf

€

ur Mikrotechnik,

Mainz), and BESSY (Berliner Elektronenspei-

cherring-Gesellschaft f

€

ur Synchrotronstrahlung,

Berlin). The LIGA proce ss steps were standard-

ized and interface specifications were defined to

establish a modular production method. The aim

of a project named LIGA 4H (High aspect ratio,

High structure height, High dimensional accu-

racy and high sidewall quality), was to obtain a

stable manufacturing process. The next step

consists of increasing the production and of

decreasing the cost, and theoretically it should

be a s imple procedure then to move from the low

volume prototype production to high volume

production by simply adding equipment and

labour. However, the transition is much more

complex. A fully automated fabrication line

(FELIG) for dire ct LIG(A) parts in a cleanroom

environment has been built up to demonstrate

the high throughput potential of LIGA at the

ANKA synchrotron.

COMMERCIAL APPLICATIONS

Three examples are now given of commercial pro-

ducts representing the different uses of the pro-

cess. For a more complete review of applications,

see [4].

Polymer Compound Refractive X-ray

Lenses – Direct Resist Structure [5]

For hard X-rays the refractive index r in matter is

slightly less than unity. This implies a focal length

F (given as F = R/2Nr where R is the radius of the

lens) of a single concave lens (N = 1) in the range

of so me 10 meters. A compound refractive lens,

consisting of a linear arrangement of N single

lenses (see Fig. 13-14), significantly reduces the

focal length and thus can overcome this problem.

For 14 keV photons a focal spot of 0.32 mm

FIGURE 13-14 Compound refractive lens for focusing X-rays in one direction.

216 CHAPTER 13 Deep X-Ray Lithography

(FWHM) was achieved using a focal distance of

242 mm.

Micro-spectrometer [31] –Mold

The spectrometer, as shown in Fig. 13-15, con-

sists of a molded base-plate where the central

hollow area, the gating profile, the location of

the input fiber and the 45



mirror surface, are

produced with a single pressing. The molded

part is then placed in a vacuum chamber and

an aluminium film is deposited on both the grat-

ing surface and the mirror surface. This is then

assembled into the complete spectrograph by the

fitting of the input fiber, the central hollow

being filled with a polymer of different refractive

index t o that of the molding and a l id of the same

refractive index as that of the base that is glued

into place. The assembly is then aligned with,

and glued to, the diode array.

Micro-gears – Direct Metal Parts [32]

The search for the perfect product is one of the

major preoccupations of the luxury watch-

making industry. The combination of lithogra-

phy with electroforming offers unparalleled

machining precision and extended design f ree-

dom for the manufacture of fine parts. The SU-

8-based UV LIGA technology undoubtedly

attracts interest, in particular when competing

technologies fail to meet the precision and

quality requirements. Deep X-ray lithography

enables greater precision, better side-wall quality

and fewer restrictions, as almost any design can be

realized. As an example, the quality of the produc-

tion of gearwheels (thickness: 180 mm) in nearly

pure, but sufficiently hard, gold was addressed.

From a fabrication point of view, dimensional

quality refers to all of the parts that can be pro-

duced, the deep X-ray LIGA process showing a

process reproducibility of better than 1 mm in the

case of the example gearwheel. In Fig. 13-16, the

reproducibility of the measurement process of a

gear (in PMMA) used as a reference is presented:

it is excellent, the standard deviation being 0.1 mm

In Fig. 13-17, the variation between about 20

wafers is presented. The diameter range min-max

is 2 mm. The FT for all the measurements is better

than 0.5 mm The process shows a reproducibility

of better than 1 mm (standard deviation:

0.5 mm). Measur ements of side-wall roughness

were performed using an atomic-force micro-

scope: all the data (R

) obtained are of better than

50 nm accuracy.

CONCLUSIONS

Being invented in the 1980s, the potential of LIGA

became evident in t he 1990s when numerous

examples of structures and devices were pre-

sented. The lead in MEMS technologies, how-

ever, was taken by silicon micro-machining,

and not by LIGA, despite its technical superior-

ity for many applications. The main reasons for

this are the required technical infrastructure,

FIGURE 13-15 Microspectrometer.

CHAPTER 13 Deep X-Ray Lithography 217

process knowhow, and cost issues. Silicon-

based technologies could exploit the vast tech-

nology base developed for chip-making, with

billions of dollars invested. LIGA, on the other

hand, was new a nd confined to research labo-

ratories. Furthermore , as a key process step,

LIGA required access to a synchrotron-radia-

tion facility, again a research laboratory provi-

sion, and often unacceptable for industries

establishing manufacturing plants. Neverthe-

less, and this is a demonstration of the technical

strength and superiority of LIGA, several indus-

trial LIGA products have been launched by

industries and synchrotron-radiation facilities

are currently widely used for micro-fabrication.

In addition to the described efforts to make

LIGA acceptable as a manufacturing technology

for a large variety of industrial products, cut-

ting-edge r esearch in, and with, LIGA remains a

topic of serious interest. The goals include

FIGURE 13-16 Reproducibility of the measurement process of a gear.

FIGURE 13-17 Variation between about 20 wafers.

218 CHAPTER 13 Deep X-Ray Lithography

research into new materials, new replication

techniques, and new lithographic approaches

exploiting the short-wavelength nature of

X-rays. Questions such as ‘how small can we

really get?’ and ‘can we overlap our top-down

technologies with the typical bottom-up approach

in nano-technology? At which dimensions?’ are

currently being addressed. Making devices much

smaller than is possible today will open up entirely

new fields in research and applications.

REFERENCES

[1] E.W. Becker, W. Ehrfeld, P. Hagmann, A. Maner,

D. M

€

unchmeyer, Fabrication of microstructures with

high aspect ratios and structural heights by synchro-

tron radiation lithography, galvanoforming, and

plastic moulding (LIGA process), Microelectronic

Engineering 4 (1986) 35–36.

[2] R.K. Kupka, F. Bouamrane, C. Cremers, et al., Micro-

fabrication: LIGA-X and applications, Appl. Surf. Sci.

164 (2000) 97–110.

[3] J. Hruby, LIGA technologies and applications, MRS

Bulletin 26 (4) (2001) 337–340.

[4] C.K. Malek, V. Saile, Applications of LIGA technol-

ogy to precision manufacturing of high-aspect-ratio

micro-components and -systems: a review, Micro-

electronics J. 35 (2004) 131–143.

[5] V. Nazmov, E. Reznikova, J. Mohr, A. Snigirev,

I. Snigireva, S. Achenbach, V. Saile, Fabrication and

preliminary testing of X-ray lenses in thick SU-8 resist

layers, Microsystem Technologies 10 (2004) 716–721.

[6] C. Becnel, Y. Desta, K. Kelly, Ultra-deep X-ray

lithography of densely packed SU-8 features: I. An

SU-8 casting procedure to obtain uniform solvent

content with accompanying experimental results,

J. Micromech. Microeng. 15 (2005) 1242–1248.

[7] L.J. Guerin, The SU8 Homepage, http://www.geoci-

ties.com/guerinlj/.

[8] C. Cremers, F. Bouamrane, L. Singleton, R. Schenk,

SU-8 as resist material for deep x-ray lithography,

Microsyst. Technol. 7 (2001) 6–11.

[9] M. Stewart, H. Tran, G. Schmid, T. Stachowiak,

D. Becker, G. Willson, Acid catalyst mobility in resist

resin, J. Vac. Sci. Technol. B 20 (2002) 2946–2952.

[10] H. Winick, Synchrotron Radiation Research, Chap-

ter 2: Properties of synchrotron radiation, Plenium

Press, New York (1980).

[11] G. Margaritondo, Introduction to Synchrotron Radi-

ation, Chapter 2.2: Bending magnet radiation,

Oxford University Press, New York, Oxford (1988).

[12] D. Attwood, Univ. California (Berkeley), Intro Syn-

chrotron Radiation, Bending Magnet Radiation,

EE290F (Feb. 8, 2007).

[13] A. El-Kholi, P. Bley, J. G

€

ottert, J. Mohr, Examination

of the solubility and the molecular weight distribu-

tion of PMMA in view of an optimised resist system

in deep etch x-ray lithography, Microectronic Engi-

neering 21 (1983) 271–274.

[14] M. Gad-el-Hak, The MEMS Handbook, CRC Press,

New York, USA (2001).

[15] O. Schmalz, M. Hess, R. Kosfeld, Structural changes

in poly(methyl methacrylate) during deep-etch X-ray

synchroton radiation litography. Part II: Radiation

effects on PMMA, Die Angewandte makromoleku-

lare Chemie 239 (1996) 79–91.

[16] W. Glashauser and G.V. Ghica, German patent,

3039110 (1982).

[17] J.S. Greeneich, Developer characteristics of poly-

(methyl electron resist methacrylate), J. Electrochem.

Soc. 122 (1975) 970–976.

[18] Z. Liu, F. Bouamrane, M. Rouilliay, R. Kupka,

A. Labeque, S. Metgert, Resist dissolution rate and

inclined-wall structures in deep x-ray lithography, J.

Micromech. Microeng. 8 (1998) 293–300.

[19] P. Meyer, A. El-Kholi, J. Mohr, C. Cremers,

F. Bouamrane, S. Metgert, Study of the development

behavior of irradiated foils and microstructure, SPIE

3874 (1999) 312–320.

[20] P. Meyer, A. El-Kholi, J. Schulz, Investigations of the

development rate of irradiated PMMA microstruc-

tures in deep X-ray lithography, Microelectronic

Engineering 63 (2002) 319–328.

[21] F.J. Pantenburg, S. Achenbach, J. Mohr, Influence of

developer temperature and resist material on the

structure quality in deep x-ray lithography, J. Vac.

Sci. Technol. 16 (1998) 3547–3551.

[22]P.Meyer,J.Schulz,L.Hahn,DoseSim:MS-

Windows Graphical User Interface for using synchro-

tron X-ray exposure and subsequent development in

the LIGA process, Review of Scientific Instruments

74 (2) (2002) 1113–1119.

[23] S. Hafizovic et al., X3D: 3D X-ray lithography and

development simulation for MEMS, Transducers’03

(2003) 1570–1573.

[24] M. Paunovic and M.Schlesinger, Fundamentals of

Electrochemical Deposition. The Electrochemical Soci-

ety Series, John Wiley and Sons 2nd edition (2006).

[25] S.K. Griffiths, J.M. Hruby, A. Ting, The influence of

feature sidewall tolerance on minimum absorber

thickness for LIGA x-ray masks, J. Micromech.

Microeng. 9 (1999) 353–361.

[26] S.K. Griffiths, Fundamental limitations of LIGA x-ray

lithography: sidewall offset, slope and minimum feature

size, J. Micromech. Microeng. 14 (2004) 999–1011.

[27] D. Chinn, P. Ostendorp, M. Haugh, R. Kershmann,

T. Kurgess, A. Claudet, T. Tucker, Three dimen-

sional imaging of LIGA-made microcomponents, J.

of Manufacturing Science and Engineering 126

(2004) 813–821.

CHAPTER 13 Deep X-Ray Lithography 219

[28] S. Cao, U. Brand, T. Kleine-Bestent, W. Hoffmann,

H. Schwenke, S. B€utefisch, S. B€uttgenbach, Recent

developments in dimensional metrology for micro-

system components, J. Microsystem Technologies 8

(2002) 3–6.

[29] U. Brand, J. Kirchhoff, A Micro-CMM with metrol-

ogy frame for low uncertainty measurements, mea-

surement science and technology, Meas. Sci. Tech-

nol. 16 (2005) 2489–2497.

[30] L. Hahn, P. Meyer, K. Bade, H. Hein, J. Schulz,

B. L

€

ochel, H. Scheunemann, D. Schondelmaier,

L. Singleton, MODULIGA: the LIGA process as a

modular production method-current standardization

status in Germany, Micr. Tech. 11 (2005) 240–

245.

[31] C. M

€

uller, J. Mohr, Microspectrometer fabricated by

the LIGA process, Interdisciplinary Science Reviews

18 (3) (1993) 273–279.

[32] P. Meyer, J. Schulz, L. Hahn, V. Saile, Why you will use

the deep X-ray LIG(A) technology to produce

MEMS? Microsystem Technologies 14 (9–11) (2008)

1491–1497.

220 CHAPTER 13 Deep X-Ray Lithography