28 Prakt. Met. Sonderband 46 (2014) 
with superior performance under these loading conditions. In this size regime, significant as it would 
size effects on material properties are observed [5-8]. Thus, in order to enable a reliable techniques su 
design of novel products, it is detrimental to experimentally measure the stresses 
sustained by such small scale structures at their native length scale. To get an idea of the 
approximate dimensions, Fig. 1 presents the comparison of a human hair with a typical 3 21 ELECTR 
um micro-tensile sample and a 300 nm nano-tensile sample. 
Using electroc 
respect to the 
micrometers. 
fabricate near 
interest allows 
cutting or grin 
[19], or thin la 
[17, 20]. Subs 
thin freestand 
be removed t 
Moreover, the 
any material £ 
in situ micron 
required prej 
electrochemic 
2.2 BROAD 
Fig. 1: Comparison between a human hair, a 3 um Cu micro-tensile sample, and a 300 nm A limitation of 
nano-tensile specimen. or phases are 
even impossik 
It is obvious that such structures cannot be produced by a classical top-down approach. applied. We u 
There are some special bottom-up processes that can grow such small structures, such as an acceleratic 
whisker growth [10, 11], electrodeposition [12, 13], or lithographic processes [14]. While beam is in the 
these processes are capable of delivering a large number of samples, they are typically sputter yield i 
silicon-based techniques that require dedicated fabrication facilities and suffer flexibility current (4-7 
with respect to the materials systems that can be processed straight forward. Therefore, magnitude). T 
we will place our focus on alternative approaches that can be realized with common time. The spt 
metallographic preparation techniques such as electrochemical etching, broad beam ion [23]. The ma 
milling/polishing, and focussed ion beam (FIB) milling. protected by 
microscope. 1 
accuracy of pt 
2. MINIATURIZED SAMPLE PREPARATION selectivity of 1 
cross-sections 
The major strength of the FIB is the ability to machine almost any vacuum compatible > 2 fow mir 
material to various geometries with the precision of a few nanometers [15, 16]. However, materi os ot 
the amount of removable material with a 30 keV Ga* ion beam is in the order of 1 pm? nA” in som yste 
s™' [15, 17, 18], depending significantly on the sputtered material and sputtering angle, as useful : N: En 
well as the required production precision which dictates the final milling current. To the alumini or ms 
, ; or inium, po 
authors’ experience, the production steps within the FIB are the bottleneck of sample following, we ' 
production. Therefore, it is of general interest to minimize the amount of material to be = 
removed by FIB milling. To avoid unwanted material modifications during material thinning.