Content uploaded by Mario Birkholz
Author content
All content in this area was uploaded by Mario Birkholz on Feb 12, 2019
Content may be subject to copyright.
For Peer Review
Technology modules fro
m micro
-
and nano
-
electronics for
the life sciences
Journal:
WIREs Nanomedicine & Nanobiotechnology
Manuscript ID:
Draft
Wiley - Manuscript type:
Overview
Date Submitted by the Author:
n/a
Complete List of Authors:
Birkholz, Mario; IHP,
Mai, Andreas; IHP,
Wenger, Christian; IHP,
Meliani, Chafik; IHP,
Scholz, Rene; IHP,
Keywords:
bioelectronics, semiconductor technology, biosensor, lab-on-chip,
CMOS/BiCMOS
Choose 1-3 topics to
categorize your article:
Nanoscale Systems in Biology (2NSB) < Nanotechnology Approaches to
Biology (NAB), Biosensing (2BIS) < Diagnostic Tools (DAT), Emerging
Technologies (2EMT) < Therapeutic Approaches and Drug Discovery (TAD)
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
For Peer Review
Technology modules from micro- and nano-electronics
for the life sciences
M. Birkholz*, A. Mai, C. Wenger, C. Meliani, R. Scholz
IHP, Im Technologiepark 25, 15236 Frankfurt (Oder), Germany
*corresponding author: birkholz@ihp-microelectronics.com
Abstract
The capabilities of modern semiconductor manufacturing offer remarkable possibilities to be applied in life
science research as well as for its commercialization. In this overview, the technology modules available in
micro- and nano-electronics are exemplarily presented for the case of 250 and 130 nm technology nodes.
Preparation procedures and the different transistor types as available in CMOS and BiCMOS technologies are
introduced as key elements of comprehensive chip architectures. Techniques for circuit design and the
elements of fully integrated bioelectronics systems are outlined. The possibility for life scientists to make use of
these technology modules for their research and development projects via so-called multi-project wafer
services is emphasized. Various examples from diverse fields like (1) immobilization of biomolecules and cells
on semiconductor surfaces, (2) biosensors operating by different principles like affinity viscosimetry,
impedance sprectroscopy and dielectrophoresis, (3) complete systems for human body implants and monitors
for bioreactors, and (4) the combination of microelectronics with microfluidics either by chip-in-polymer
integration as well as Si-based microfluidics are demonstrated from joint developments with partners from
biotechnology and medicine.
Keywords: bioelectronics, semiconductor technology, CMOS/BiCMOS, biosensor, lab-on-chip
1 Introduction
The usage of micro- and nano-electronics in the life sciences has steadily increased over the last years
1-6
. For
one reason, this development is due to the molecular nature of the mechanisms of living, for the study of
which it often suffices to manipulate only a small number of molecules. This fact established a high demand for
miniaturized systems acting as links between men and microbiology and led to medical and biotechnological
devices enabling an increased comfort for patients, the on-line monitoring of bioprocesses and various high-
throughput processes to mention only a few examples. For another reason, the demand for micro-electronics
in the life sciences derives from the fact that both disciplines increasingly stronger apply nano-technological
methods, such that they start to share a common set of technical equipment and tools.
Modern microelectronics relies on the elementary semiconductor silicon and the CMOS technology based upon
it. In the latter, electrons and missing electrons or holes may equally act as mobile charge carriers in
complementary metal-oxide-silicon devices. Since the introduction of integrated circuits (ICs) in 1958 and, in
particular, since G. Moore realized the number of devices per IC to double every 1½ - 2 years
7
, semiconductor
technology followed this rule like a self-fulfilling prophecy. Meanwhile, the number of transistors in computer
ICs has reached the order of magnitude of billions and their dimensions have advanced from the µm- into the
nm-range. The continuous shrinking of device dimensions and increase of IC performance is denoted as scaling
8, 9
. Its significance becomes obvious from the fact that the parameters of tomorrows ICs are already defined
today by the international semiconductor industry
10
.
Next to CMOS technology and scaling another direction of research and development has evolved denoted as
“More-than-Moore”
11
. It aims at introducing additional functions into integrated circuits in order to realize
Page 1 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 2
new applications in consumer electronics, sensoric devices or communication technology – to mention only a
few examples. The More-than-Approach is also motivated by cost considerations. Modern CMOS fabs have
recently reached the 32 and 22 nm node requiring investments for appropriate clean rooms on the order of
billions of Euros that lie beyond the funding options of public research institutes.
Having worked in various projects with partners from the life sciences, it seems to the authors that the
interdisciplinary dialog between both disciplines suffers from disjunctive terminology and a mutual lack of
knowledge of the techniques on the other side. This review endeavors to outline the technical apparatus avai-
lable in micro- and nano-electronics on a non-specialist, but specific level. The presentation will focus on the
capabilities of 250 and 130 nm technologies as available at the authors’ institute. It should be mentioned that
a substantial body of work has been accumulated in the field of hybrid junctions from neuronal cells and
semiconductor devices starting with the seminal work of Fromherz and co-workers in 1991
12
. The following
developments were reviewed in Ref.
13
, and important progresses on neuro-microelectronic contacts and their
perspective applications as implants and neural tissue interfaces are continuously reported, e.g. Ref.
14-16
. The
focus of this work, however, will be on the fabrication of microelectronic chips, the designs of electronic
circuits and systems and their applications in biosensors and biotechnology. The authors hope that the given
examples will make more transparent, what life scientists may expect from the presented micro- and nano-
electronics platform and how the techniques can be applied to their research and system developments.
2 Preparation technologies for microelectronic chips
2.1 Clean room processing
Microelectronic chips are produced in clean rooms categorized by the number of detectable contaminants
within certain size classes. Many of them rank between class 1 and 100, which indicates either less than 35 or
3520 particles with diameter > 0.5 µm per m
3
, see Figure 1A for a view into the IHP clean room, in which a 130
nm pilot line is run. Wafer processing is almost fully automated in clean rooms operating on smaller technology
nodes and staff presence is mainly restricted to maintenance.
The technical equipment or tools are adapted to the processing of silicon wafers with a specific diameter. At
the time of writing, mostly 300 mm wafers are used in modern semiconductor technology. Previous IC
generations made use of 200 mm (Figure 1B), 6”, 4” wafers etc., while the next 450 mm wafer generation is
currently under development. Silicon wafers are perfect single crystals exhibiting the highest ratio of crystalline
perfection per EUR of all materials available. Silicon atoms constituting the wafer are arranged in the diamond
crystal structure (T
h
6
) having a density of N
Si
= 5 × 10
22
cm
-3
at ambient conditions. Most wafers exhibit an
orientation with the crystallographic 〈001〉 axis being perpendicular oriented to the wafer surface.
Nevertheless, wafers with different orientations like 〈111〉 may also be processed within the same clean room;
the diameter, however, is generally fixed for the set of tools available. The thickness of a 200 mm wafer
amounts to 0.75 mm and is much larger than the microelectronic circuits exhibiting thicknesses between 10
and 20 µm according to the technology used.
Prior to processing the electrical conductivity σ of the wafer typically amounts to 2 (Ωm)
-1
, a value which lies
between those of metals of about 10
7
(Ωm)
-1
and those of isolators on the order of 10
-12
(Ωm)
-1
. Type and
magnitude of conductivity may be adjusted by introducing doping elements into the Si crystal lattice. For
instance, doping with group V elements like P and As will introduce additional mobile electrons and thus
enhances the conductivity with negative charge carriers, i.e. n-type. Doping with group III elements like boron,
on the other hand, will produce a p-type conductive Si crystal due to excess and mobile positive charge carriers,
i.e. missing electrons. Smallest amounts of dopants can vary σ by orders of magnitude. The concentration of
dopants N is specified by the dimensionality of cm
-3
. While σ = 3.2 × 10
-4
(Ωm)
-1
is valid for undoped Si, the
substitution of every millionth Si by one B atom (N
B
= 5 × 10
16
cm
-3
) causes σ to increase to 145 (Ωm)
-1
.
Wafer processing is started with a base doping on the order of 10
15
cm
-3
, while any other doping level is
obtained by introducing higher concentrations of dopants on the order of 10
16
to 10
20
cm
-3
within selected
areas. The tuning of conductivity in selected volume elements of the wafer compares to the situation in
biological cells, where spatially varying pH values or concentrations of electrolytes etc. are organized by
compartmenting the cell volume into cell organelles.
Page 2 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 3
In modern semiconductor technology doping is performed via ion implantation. For this purpose, doping atoms
are evaporated from a reservoir, ionized and accelerated by an electrical field towards the wafer surface. For
acceleration energies up to 900 keV an average depth of 0.6…1.5 µm below the wafer surface can be reached
depending on dopant and level of ionization. Since the Si crystal lattice is severely damaged during
implantation, a subsequent recrystallization is performed at high temperatures with huge heating and cooling
rates up to several hundred °C s
-1
in order to restrict possible diffusion effects.
For processing, a 200 mm wafer is partitioned in about 70 - 120 rectangles, so-called test fields, having a size
between 10 × 26 and 16 × 26 mm (Figure 1C). The microchips prepared on every test field are identical so that
many equal ICs are processed on one wafer. Integrated circuits are produced by a sequence of typically a few
hundred process steps belonging to the groups of coating, structuring, etching, doping and clean processes.
Coatings are applied in the form of thin films having thicknesses between a few nm and some µm. The wafer is
covered over its full size after depositing the coating and the small areas, from which device components like
the transistor gate or metal lines are prepared, still have to be excised. This is done by lithographic structuring
and subsequent etching steps.
Figure 2 displays a cross section SEM micrograph of a developed and etched photoresist layer at the resolution
limit accessible to the lithography tools used. The investigation aimed at implanting as small as possible n-
doped bars in a p-type wafer. The photoresist coating was illuminated through a mask, in which transparent
and in-transparent regions alternated in order to generate stripes of light and shadow having 130 nm width.
The micrograph was taken after implanting n-areas and prior to removal of structured photoresist and
annealing
17
. Doped areas were still in an amorphous state and prone to etching, which allowed achieving a
good SEM contrast. It can be seen that the intended periodicity (or pitch) of 260 nm was arrived at, albeit slight
roundings of resist edges are recognized to occur. In general, the resolution achievable by photolithography
determines the minimum feature dimensions of active devices that may be fabricated. The remarkable aspect
of semiconductor technology is not only due to the precision achievable, but also due to the reproducibility,
meaning in the example shown here that the doping bars are equably repeated more than 760,000 times along
the wafer’s diameter.
2.2 Transistor types and chip architecture
The most important devices in microelectronic circuits are metal-oxide-semiconductor field-effect transistors
or MOSFETs, which operate as amplifiers or switches. TEM micrographs of a MOSFET are given in Figure 3.
Basically they act like a valve with currents flowing from the source S to the drain D, and its magnitude being
controlled by the voltage applied to the gate G. When CMOS technology was introduced in the 1970 it was its
success to allow for the complementary preparation of both n-channel as well as p-channel MOSFETs on the
same wafer, where the latter are characterized by a channel current that is due to missing electrons or holes.
The preparation of MOS devices requires several thin film depositions, e.g. gate oxide, polycrystalline silicon
gate and source/drain spacer material and contact electrodes.
The geometrical distance between source and drain is denoted as channel length L. It represents a key
parameter for the performance of the technology used, since it is a measure for the speed of switches that may
be designed with the underlying MOSFETs. The smaller L is the earlier D is reached by charge carriers starting
from S. Between 1975 and 2012 the channel length L was reduced from 3 µm to about 22 nm in modern CMOS
fabs, while the clock frequency of processors increased from a few MHz into the GHz range.
An important role in semiconductor manufacturing is played by the tools for thin film deposition. The applied
processes are often denoted as vapor deposition, since they are usually performed via condensation from the
gas phase. Two different groups may essentially be distinguished with the first one encompassing physical
processes like evaporation or sputtering, while the processes from the second group are of basically chemical
nature. Depending on whether a process belongs to either the first or the second group, it is either denoted a
PVD or a CVD process.
For instance, the MOSFET gate from polycrystalline silicon depicted in Figure 3B, is prepared by a CVD process,
in which gaseous silane SiH
4
is introduced into the deposition chamber to decompose into Si and H
2
. The wafer
temperature lies between 500-600°C causing a thermal decomposition of SiH
4
. While H
2
returns to the gas
phase, the remaining Si atoms condensate on the surface to form the layer. Typical growth rates are on the
order of nm s
-1
.
Page 3 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 4
Next to field-effect transistors FETs, where the charge carrier transport is regulated via a voltage applied to the
gate, another type is represented by hetero-junction bipolar transistors HBTs. They differ from FETs by the
direction of current, which is mainly oriented perpendicular to the wafer surface instead of being parallel. In
addition, start and end point of the current to be amplified are named emitter and collector, with the two of
them being separated by the so-called base (Figure 3C). Most importantly, however, the emitter-collector
current is not adjusted by a voltage but by another current that is introduced laterally into the base.
While only one charge carrier type n or p is flowing from source to drain of a MOSFET, the charge transport in
HBTs is related to both electron and holes that are transported in the active state and from which the notion of
a bipolar device derives. Extreme high cutoff frequencies above 200 GHz may be obtained with the HBT
configuration, with the base region acting as key element. It is formed from the solid binary phase Ge
0.2
Si
0.8
containing about 20% germanium, with Ge atoms distributed randomly over Si crystal lattice sites. The
preparation is performed as a thin layer by CVD from SiH
4
and GeH
4
by the technique of epitaxial growth,
where the crystal structure of the growing film is imposed from the underlying crystal lattice. Figure 5 displays
the TEM micrograph of a HBT configuration.
A focus of HBT research was to increase the maximum frequency of operation f
max
, by which the device may be
operated. The approach to achieve this goal was to stabilize the boron doping of the SiGe base by adding small
amounts of carbon to reduce the out-diffusion of B atoms during subsequent annealing steps
18
. This measure
allowed to substantially increasing the steepness of B doping profile, which turned out as a key for high-
frequency performance and, for the time of writing, HBT limiting frequencies f
max
of up to 500 GHz have already
been achieved
19
.
Figure 4A displays the schematic architecture of a microelectronic chip fabricated by a 0.25 µm technology that
allows for the preparation of MOSFETs having L = 0.25 µm. Various horizontal metal layers and vertical
connections between them can be recognized that are embedded into electrically isolating glass, SiO
2
,
indicated by blue color. At the lower edge of the figure, active devices like MOSFETs and HBTs can be seen that
are prepared directly upon the silicon wafer. The process part used for the fabrication of active transistors is
denoted as Frontend-of-Line (FEoL) and distinguished from the subsequent process parts that are named as
Backend-of-Line (BEoL).
For the technologies considered horizontal metal layers are essentially made up from aluminum alloyed with a
few percent of copper (Al:3%Cu). They are deposited by a magnetron sputtering process at 0.27 Pa and 200°C.
Five to seven metal layers are prepared in IHP’s technology flows with the precise number depending on the
technology used. The top-most layers TM1 and TM2 exhibit thicknesses of 2 and 3 µm being significantly
thicker than the underlying metal layers of only 0.58 to 0.72 µm denoted by M1..M3.
The latter are shown in a cross section micrograph in Figure 4B. Also, vertical tungsten plugs may be recognized
that are introduced into previously etched holes in the inter-layer dielectric (ILD). The W plug preparation is
performed with a CVD process using W(CO)
5
as a precursor to assure a continuous hole filling starting from the
bottom and the side walls. Such conformal deposition may not be achieved by PVD processes. The tungsten
plugs have no direct contact with Al:Cu metal layers, since the latter are covered on top and at the bottom by
titanium nitride TiN, an electrically conducting ceramic acting as a diffusion barrier.
From the materials science point of view, TiN is a remarkable solid that behaves in mechanical respect like a
ceramic and, concomitantly, disposes of an electrical conductivity of 5 x 10
6
Sm
-1
comparable to those of metals
20
. TiN layers in CMOS architectures obey only a few nm thicknesses and are usually deposited by the PVD
technique of magnetron sputtering. TiN remains practically un-corroded when brought in contact with biogenic
fluids. Altogether, metal layers and W plugs act as lateral and vertical electrical connections. They are
electrically isolated by embedding them in ILD. The latter are essentially formed from SiO
2
deposited by CVD
techniques with SiH
4
, O
2
, O
3
or other gaseous precursors that are introduced into the deposition chamber and
reacted at temperatures of up to 400°C on the wafer surface.
Another group of processes used in microelectronic clean rooms are etching processes. They mainly apply to
selected areas of thin films deposited over the full wafer that are not required for the circuit to be configured.
On the one hand, etching may be based on wet-chemical processing, where the wafer is dived into an etching
solution. Well known, for instance, is the HF dip, by which the wafer is submerged for a few minutes in 2% HF
in order to dissolve the native oxide SiO
2
layer that covers every elemental Si surface under ambient conditions
with a thickness of 1..2 nm.
Page 4 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 5
On the other hand, plasma etching techniques are moreover applied operating with ions accelerated towards
the wafer in order to sputter unwanted layers off the surface. These processes may only be performed in
vacuum chambers, since plasma may only be ignited under low-pressure conditions. Etching processes can be
performed selectively by covering the surface with dedicated layers as e.g. photoresist and opening only over
areas intended for etching. However, the protection layer has to be chosen such that a high selectivity of the
process is given, i.e. that the protection is more resistant to etching than the layer to be etched off.
Essentially, all processes have now been presented for fabricating microchip architectures as shown in Figure 4.
Several of the structures mentioned like irradiated photoresist or etched-off layers are measured during wafer
processing by SEM and other inspection techniques. Special control structures are designated for this purpose
on every test field, by which, for instance, the critical dimensions of MOSFET gate lengths L, widths W and
other parameters are controlled.
In addition, various test devices are subjected to electrical control. They are denoted as process control
monitors PCMs and encompass e.g. MOSFETs and HBTs, devices for determination of sheet resistances or
capacitors. A full measurement of a wafer thus yields between some thousand measurement points that are
automatically valued, the results of which decide on releasing the wafer to the customer.
3 Surface conditioning
An important presupposition to be fulfilled relates to the state of the particular surface that interacts with the
bio-milieu
21, 22
. In most cases the interaction has to act through defined material windows, because of the high
corrosivity of electrolyte solutions versus semiconductors.
An illustrative example is given in Figure 6 showing an optical microscope view of the surface and a SEM cross
section view of a microelectronic chip, where a defect in the surface passivation has allowed the surrounding
electrolyte solution to reach the TM1 level. It may be recognized that the metal line made of Al:Cu has been
damaged along a path of many µm causing a total failure of the sensor chip. Such defects have to be avoided
without restraints and various measures can be taken for this purpose like planarization of the surface
23
or
usage of particular nitrogen-rich silicon nitride
24
.
3.1 Metallic electrodes
In most bioelectronics devices a set of metallic electrodes has to interact electrically with the bio-milieu. This is
a particular challenging task, since the presence of electrical fields was often observed to intensify corrosion
effects. In these situations titanium nitride may be the material of choice, since it turned out remarkably
corrosion-resistant. Its corrosion-resistance in biotechnological applications is expressed by negligible redox
rates in cyclic voltammograms when TiN was used as working electrodes
25
, compare Figure 6C. Accordingly,
TiN electrodes are applied in biomedical applications like the artificial retina implant
21
or IHP’s glucose sensor
chip
26, 27
. TiN should always be considered as the top-most electrode material, when the electrodes are
intended to interact with bio-milieus.
Introducing other metals into the clean room than those already processed is not an easy task. This is due to
the fact that various metal atoms in semiconductors act as deep recombination centers for free charge carriers.
The associated reduction of current densities is detrimental to many devices and may cause their failure.
Additional materials thus have to pass an extensive qualification procedure, before releasing them for
preparation processes in the clean room.
The usage of some materials, however, is excluded from the outset. This holds, for instance, for gold acting as
efficient recombination center in silicon. On the other hand, gold is multiply used for the immobilization of
organic molecules on technical surfaces by forming a covalent bonding via sulfur atoms
28, 29
. Post-CMOS
processes have thus been developed that allow for the deposition of gold electrodes on otherwise fully
processed microelectronic chips
30
. For this purpose, the wafer is covered with photoresist, exposed and
developed and finally discharged from the clean room to receive a full-covering gold deposition outside. In a
subsequent lift-off process the wafer is bathed in resist-dissolving solution leaving only the areas with an Au
layer that were not covered by resist.
Page 5 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 6
3.2 Nano-structured surfaces for immobilization
Alternative protocols may apply for the covalent immobilization of biomolecules that were initially developed
for glassware, since the native SiO
2
surface layer exposes hydroxide groups Si-OH into aqueous media. Here,
the well-known cross-linking techniques based on organosilanes like aminopropyl-triethoxy-silane (APTS)
should be considered or double cross-linking that additionally make use of glutaraldehyde (GD)
31, 32
.
It appears more elegant to immobilize biomolecules on semiconductor surfaces by other methods like physical
adsorption. In particular, the interesting perspective arises for microelectronics to make use of electrical fields
generated within the chip itself. In principle, this approach will enable to map the structures inscribed into the
silicon surface to the immobilized molecular layer
33
. A first approach for electrically-assisted immobilization is
offered by the structuring of a silicon surface by alternating p- and n-doped stripes. Due to the diffusion of
electrons and holes, a built-in electric field is formed – as well-known from the pn junction, with the electrical
field lines will not confine to the silicon volume, but also extend above the wafer surface.
Figure 6 A/B shows the results as obtained from a finite-element simulation (FEM) of a line lattice of n-doped
stripes that were implanted into a p-doped wafer having a base doping N
A
of some 10
15
cm
-3
. The width of the
stripes amounted to 130 nm, while the depth of n-doped areas was adjusted by the implantation energy to
20 nm. Electrical fields are seen to extend up to 100 nm into the space above the surface
34
, where they can
cause the orientation and attraction of dipolar molecules like amino acids, proteins etc. It has to be mentioned,
however, that electrical fields in aqueous solutions will cause the formation of a Helmholtz layer and thereby
reducing the range of surface fields compared to the vacuum case
35
. An intermediate state during the
preparation of a doping line lattice is displayed in Figure 2C
17, 36
.
Such doping lattices may not only have an orienting effect on biological macromolecules, but also on complete
cells. The intentional orientation of cells is essential for biomechanics, cell biology, tissue engineering and
regenerative medicine applications
37
. The orienting effect of a surface doping lattice has been investigated in a
model system for the case of human cancer cells. Figure 6C demonstrates the successful orientation of MG-63
osteoblasts along a doping lattice with a width of p- and n-doped stripes of 180 nm each
38
.
Various other methods for the structuring of semiconductor surfaces are possible, which all offer the
advantages of semiconductor processing, i.e. to operate at the lowest level of the length scale accessible by
technical means today and to be reproducible in large sample numbers. One variant of particular interest
makes use of nanometer sized or micrometer sized electrodes that are embedded in an isolator. This may, for
instance, be realized by configuring the tungsten vias in regularly ordered arrays. The preparation scheme and
micrographs of such an array are shown in Figure 7 with W electrodes of 500 nm diameter and 2 µm distance
39
. Such arrays may effectively be used to pursue an electrical immobilization of small objects by
dielectrophoresis (DEP) as shown in the next section.
4 Biosensor modules
Sensor technology represents the main field of applications, to which microelectronics currently delivers the
most relevant contributions to biotechnology. Firstly, this relates to the miniaturization enabled by the use of
microchips for the measurement of physico-chemical quantities like temperature, pressure, conductivity, ion
density, pH or viscosity etc. Secondly, micro-sensors are multiply applied for the detection and measurement of
concentrations of biochemical analytes in samples of serum, pure bred or cell culture media etc. The
miniaturization of biosensors is of particular interest, since it paves the way for integrating it close to the point
of interest or point of care and to monitor the analyte concentration regularly. The obtained transients of
metabolites or other small biomolecules may be of great interest to medical research, the individual health of a
patient or the understanding of a bioprocess.
4.1 Microelectromechanical systems (MEMS)
Also micro-electromechanical systems (MEMS), which combine microelectronics with mechanically excitable
parts, have been developed for this purpose
40, 41
. A microviscosimeter may serve as an example that operates
by affinity sensorics to determine low-molecular weight analytes. The detection scheme relies on an assay
containing a receptor and a polymer formed from analyte monomers and being spatially separated by a
semipermeable membrane from the test solution. The receptor is acting as a weak antibody with a highly
Page 6 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 7
specific paratop for the target molecule. However, its dissociation constant K
d
is only on the order of 10
-5
M
and the receptor-analyte binding then is of reversible nature.
The plant lectin Concanavalin A (ConA) has been used in many investigations for glucose detection due to its
specific binding pocket for glycosyl and mannosyl residues. The protein is isolated from Canavalia ensiformis,
which was one of the first, purely isolated proteins
42
and encompasses 235 amino acids with 26.5 kDa
43
. The
formation of the active tertiary structure requires the inclusion of Mn
2+
and Ca
2+
into the functional center and
for physiological pH values ConA configures into a homo-tetramer with sugar binding sites being exposed at its
periphery.
Mixing ConA with glucose polymers like dextran causes the formation of a macromolecular network between
lectin tetramers and polymer molecules exhibiting a gel-like character. Mixtures with high viscosities on the
order of a few 100 mPa⋅s may be prepared by varying the size of dextran molecules, the ConA or dextran
concentration
44
. The usage of such mixtures for the determination of glucose concentrations appears of having
been investigated for the first time by Schultz et al. in 1978
45
. Various approaches to measure glucose by
affinity assays have been tested since then, which initially relied on fluorescence detection
46-49
, to later
encompass impedance-based
50-52
and viscosimetric techniques
53-57
.
The variation of viscosity due to glucose variation is made use of in a recently developed variant. A BioMEMS
has been developed for this purpose, in which a bendable beam is moved through the ConA-dextran-glucose
assay and from the velocity of which the viscosity is deduced, s. Figure 8A. The beam is bend in a quasi-
electrostatic operation mode, i.e. a high-frequency voltage to attract the beam to the ground plate. The
frequency amounts to 3.2 GHz and is situated between the absorption maximum of water at about 17 GHz and
of protein solutions at some 100 MHz
58
.
During the measurement the beam is moved towards the ground plate and from the time t
sw
it takes to reach a
defined position the viscosity is derived. Figure 8B-D show one of the measuring chambers realized (without
the assay) and a four-fold clamped beam exhibiting a thickness of only 50 nm. The restoring mechanical
element in the middle of the beam is designed in the form of a double-U. Beam and ground plate are made
from TiN due to its demonstrated in vitro and in vivo stability
23, 25
. Glucose may continuously be monitored
with this BioMEMS in the relevant physiological range between 0.3 and 30 mM L
-1
with a precision of a few
percent
57
(Figure 11D).
4.2 Label-free immunoassays
Microelectronics may also be applied for the classical approach of immuno-assays with very small dissociation
constants K
d
of the receptor-analyte complex in order to realize the detection with a miniaturized system.
Again, the success is decided about by the performance of the interface, i.e. by the immobilization of
antibodies on the semiconductor surface and its biocompatible design. Various sensoric techniques have been
investigated for the microelectronic transformation of the binding event into an electrical quantity.
One approach makes use of dielectrophoresis
59
and the tungsten arrays presented in the previous section. It
was shown in Ref.
60
that the functionality of anti-bodies could be conserved after their dielectrophoretic
immobilization. An array as displayed in Figure 7 was applied for this purpose with all tungsten nano-cylinders
being subjected to the same alternating voltage by connecting them via a bottom metal layer. DEP with such an
array appears particularly suited for molecules compared to cells, because their small dimensions and
curvature (30 nm in the case considered here) allow for the generation of sufficiently strong electrical field
gradients. Frequencies in the 10..100 kHz range turned out as optimal with respect to avoiding thermally
induced effects and formation of gas bubbles. These frequencies were successfully applied for investigations
carried out with intrinsically fluorescing protein R-phycoerythrin RPE that was brought in contact with mouse
anti-RPE-IgG1 anti-bodies. The array was positioned under a fluorescence microscope and covered with ITO-
coated glass serving as a counter electrode. Anti-bodies were first immobilized on W electrodes by applying a
100 kHz voltage with 18 V
rms
, which was followed by a rinsing step. Finally, REP was introduced and its binding
to immobilized anti-bodies can clearly be recognized by inspecting the fluorescence micrograph (Figure 9B)
39
.
4.3 FET-derived sensors
In addition to BioMEMS, SAWs and MRRs the classical bio-analytical devices of microelectronics have to be
mentioned, which are ion-selective field effect transistors ISFETs. Various derivatives were developed from
them
61
like ChemFET, EnzymeFETs etc., which all operate by the same principle: instead of using the gate
voltage V
SG
to tune the source-drain current of a MOSFET, an ion-sensitive or analyte-sensitve layer above the
Page 7 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 8
FET channel is used to modulate the channel current I
SD
. Molecular receptors have to be immobilized upon the
gate that interact with the analyte via diffusion or chemical bonding to vary I
SD
, and from the variation of which
the concentration may be deduced. ISFETs have found their main field of application for pH measurements
62
,
for which the largest variation I
SD
/pH of 59 mV/pH has been observed for Ta
2
O
3
layers covering the channel
63
.
The ISFET became largely acquainted to biotechnology in 2011, when IonTorrent presented their next
generation sequencing tool for DNA sequencing
64
.
Spatio-temporal variations in various fields of biotechnology may be investigated by a recently developed ion
camera that also operates with an ISFET array and achieving resolutions of 66 µs and 70 ms in space and time
65
. ISFETs have been adopted for the detection of alkali metal ions, penicillin in combination with pH and
functional hybrid systems with living cells for both fundamental studies and biosensoric applications to
mention only a few examples
66-68
.
4.4 Impedimetric sensors
Next to classical biosensors that operate via aptamers, lectins or enzymes by the principle of steric comple-
mentarity, the detection may also be performed by electromagnetic procedures taking a kind of finger print
from the target molecules. Rather promising appears the extension of impedance spectroscopy or dielectric
spectroscopy from kHz and MHz into the GHz range. Appropriate sensor structures may be prepared by
semiconductor technology in a comparatively simple manner, since oscillator circuits only have to supply their
output signals to interdigitated electrodes (IDE)
69-71
or microstrip lines
72
in the vicinity of the biomilieu to let
them interact with the biochemical assay under investigation. The dependence on angular frequency ω = 2πf is
accounted for by the dielectric function ε(ω), for which the real and imaginary parts ε‘ and ε“ for water are
shown in Figure 11A. The course of the dielectric function of a typical protein solution is overlaid to the Figure
showing an additional resonance peak at about 100 MHz, i.e. at lower frequencies due to the higher mass of
proteins compared to H
2
O.
The analysis of impedance spectra always requires a dielectric modelling for interpreting the measured data,
which recursively allows for deducing the molecules and their concentration. Compared to an enzyme test, the
technique is of limited specificity and requires thorough pre-investigations with respect to the assignment of
possibly occurring spectral patterns. Advantageously, impedance spectroscopy also allows for the monitoring
of full cells and cell ensembles
71
.
As in the case of SAW sensors, the geometry parameters of interdigitated electrodes have to be adapted to the
frequency range used in impedance spectroscopy. They accordingly shrink with increasing frequency. For a
frequency of 12 GHz, for instance, the penetration depths of electrical fields maximally amounts to a few mm
into a water solution for an optimally designed IDE geometry. This aspect has to be considered, whenever
spatial inhomogeneities of molecules or cell densities may be expected within the solution under investigation.
The frequency range above 1 THz increasingly comes into focus of bio-analytics, but comparatively few
spectroscopic studies were performed. In general, the penetration depth up to which THz sensors may gain
information from biological samples is in the sub-mm range. The investigations performed pointed to
characteristic peak patterns of nucleosides, nucleotides, saccharides and proteins that compare to the
detection specificity in UV-VIS spectroscopy
73, 74
. At the moment, Si-based microelectronic circuits are still
unavailable for such high frequencies and investigations have to be executed with III-V semiconductor wave
generators
73, 75
. This field of application may offer interesting perspectives for BiCMOS technologies including
high-frequency HBTs. So far, first studies with sub-THz radiation generated by SiGe:C HBT circuits were carried
out for gas sensorics
76
.
5 Microchips for molecule sensors and cytometry
Next to the sensor modules, that are in direct contact with the bio-milieu, a microelectronic chip encompasses
additional devices and modules, which serve for the generation of excitation signals or the processing of
measurement data.
Figure 10A schematically depicts the central elements of the electrical circuits of the affinity sensor chip
introduced in the previous section. A nearly mirror symmetry can be recognized to divide the upper and the
lower part, which stands for the division of the circuit into a measuring MEMS (top) and a reference MEMS
(bottom). The cavities of both microelectromechanical systems are blue shaded and their properties are
Page 8 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 9
modelled in this equivalent circuit by parallel operating capacitors (C) and resistors (R). Both devices are
marked by an arrow to indicate that their absolute values are varying during beam bending. This holds,
however, only for the measuring MEMS and not for the reference one, since the latter is mechanically rigid and
does not bend, although it is subjected to the same voltage.
A high-frequency (HF) voltage of 3.2 GHz is applied to the beam, while the chip is powered with a DC voltage of
3 V. The required HF voltage thus has to be generated on-chip, for the purpose of which three inverters are
used forming a ring-oscillator circuit. An inverter is simply build up from a p- and an n-transistor. Electrical
circuits composed of an uneven number of inverters have the remarkable property to convert a DC supply into
an oscillating resonance state with resonance frequency f
0
depending on the design parameters of the
transistors. The capabilities of the sensor-chip with a foot print of ½ mm
2
only (Figure 10B) have already been
demonstrated for in vitro glucose measurements (Figure 10C).
An important step on the way to realize a microelectronic biochip is the design of the circuit layout. Here, the
design essentially is the partition of lithography masks into dark and transparent areas. Various software
techniques have been developed for this purpose that all run under the title of Electronic Design Automation
(EDA). The design of circuit layouts takes place in a modular fashion with modern EDA tools, i.e. in a particular
technology usable elements like C, R, MOSFETs, HBTs etc. are stored in libraries as so-called parameter cells.
Figure 12B shows the micrograph of an impedance measurement chip operating in the frequency range around
28 GHz
72
. Its key element is an open microstrip line probing the solution above the chip, the dielectric function
of which is measured by affecting the electrical capacitance of an oscillator circuit. Such high-frequency
impedance chips may advantageously be applied for monitoring the density of biomolecules or microorganisms
in culture solutions. Parameter cells from SiGe:C HBTs, MOSFETs, pc-Si resistors and MIM capacitors as well as
dedicated device developments have to be combined for the circuit design and optimized with respect to the
intended application
69, 70, 72
. The chip footprint still is on the order of mm to allow for a sufficient spatial
separation of sensor functionality in the middle of the chip and bond pads at the periphery.
The function of the designed integrated circuits is examined by simulation runs in order to recognize possible
errors prior to chip production and cost-intensive usage of clean room resources. Simulations and design
optimization are often iteratively performed, until the circuit exhibits the intended behavior. Various
constraints have to be considered during circuit design with respect to minimum distances between devices or
vias or between devices and metal lines, which have been set-up to facilitate optimum chip operation. The set
of geometrical constraints is denoted as design rules and a so-called design rule check (DRC) is consequently
performed for every submitted chip design. Chip preparation is generally rejected by foundries for designs
failing the DRC.
Designed and simulated circuits are finally submitted to the foundry for production. After successful DRC they
are included, for instance, in a full test field of an MPW shuttle. A recently processed test field in an IHP 0.25
µm MPW run is exemplarily displayed in Figure 11C.
6 System integration
An essential development step is related to the integration of the microchip into an operative system, which
must be preceded by a technical and electronic design. Also a user scenario must be defined that has to include
specifications of electronic components and the system integration and has to include the important question
of sterilization.
The challenges to be expected can be illustrated by the system integration of the affinity BioMEMS presented
above into a medical implant intended for a continuous monitoring of glucose levels in human tissue. The
sensor is scheduled by the operation scenario to determine one glucose concentration value c
g
every 5 min and
to transmit the measured data five times per day. An operating time of at least half a year is planned for the
sensor system. Figure 13A schematically depicts the implant architecture
77
.
The control of measurement cycles and intermediate data storage is usually performed by a microcontroller
µC. This is a minicomputer integrated into single chip disposing of all computer components in miniaturized
form, with micro-signal processors TI MSP 430 having found the broadest dissemination. In case of a wireless
data transmission the appropriate frequency band and the transmission protocol have to be chosen, by which
the sensor communicates with the base station. For medical implants the 403-405 MHz frequency range has
Page 9 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 10
been approved by regulatory bodies
78, 79
, which is denoted as Medical Implant Communication Service (MICS)
and used by some 10.000 cardio implants worldwide.
System components are configured and electrically connected on a printed circuit board (PCB), which is made
of highly isolating materials and exhibits only little absorption. Figure 12B shows the PCB developed for the
glucose sensor implant. Electrical connections are usually implemented from copper or – for life science
applications – frequently from gold. In the Figure, the upper part of the PCB is seen to carry the JTAG port for
µC programming, which may be cut off afterwards in order to reduce the form factor.
A battery as used in cardio implants is scheduled for the power supply
80
that has to encompass a few 100 mAh
in order to enable the number of 288 measurements and 5 data transmissions per day. The battery then
becomes the form-factor-determining component of the implant. Data transmission requires an antenna and
the sensor chip will be connected via a so-called zero-insertion-force connector.
The BioMEMS chip was integrated into a cooling body that also allowed for the integration of the
semipermeable membrane at very small distance (ca. 150 µm) above the surface of the sensor chip. It also
offered sufficient space for the reception of the affinity assay encompassing Con A and dextran in an
electrolyte solution. All sensor components were integrated into a full sensor system by a silicone casting
81
,
see Figure 12. This system offers a platform to investigate the sensor function and stability under in vitro and in
vivo conditions.
Another example relates to an autonomous sensor capsule for algae bioreactors. Goal of the project was to
follow the transients of pH, electrolyte and glucose in a photo-bioreactor build up from some 100 m of
translucent tubes with an inner diameter of 5 cm. It is still unresolved, which inhomogeneities occur in such
reactors and how they affect growth and yield of the algae culture
82
. Mobile capsules being transported with
the fluid stream appear to be the system of choice. In this case, wireless data transmission must not obey the
strict regulations of medical implants, but may follow usual industry standards
83
.
The outer diameter of the sensor capsule was set to 44 mm, which is a little smaller than the inner diameter of
the reactor tubes
84
. The system PCB is mounted in the equatorial plane of the capsule with its upper layer
exposing the sensor surfaces, while the other components of the electrical circuit were installed on the
opposite side. The upper sphere of the capsule is permeable to the algae medium, while the lower part is
hermetically sealed as is the PCB such that the liquid medium cannot enter the electronics-carrying side.
Potentiometric sensors were scheduled for measuring pH and Na
+
, while glucose was intended to be monitored
by an amperometric sensor. Figure 13 schematically depicts the PCB layout, where also the electrode surface
for conductivity measurements can be recognized. In this example, the transceiver CC1101 is used as radio chip
operating in the 433 MHz band
84
.
7 Microfluidics and lab-on-chip systems
Microfluidic techniques developed in the last decades offer a suitable platform for the application of
microelectronics in the life sciences
85
. In principle, they may be realized from polymeric solids, while silicon
was of less importance, because the access to semiconductor technology was not simple and associated with
higher costs
86
. However, the perspective appears attractive to perform biotechnological experiments on
combined microelectronic-microfluidic platforms and – concomitantly – the first steps of data accumulation
and “intelligent” processing. In general, the combination of both micro-techniques is a central approach to lab-
on-chip technologies.
Remarkably, the first microfluidic channels were introduced in silicon wafers for chromatographic applications
87
. Today, the decisive technological process makes use of deep reactive ion etching (DRIE) and is also known as
Bosch process. It can be applied for introducing the channels for fluid transport into the Si wafer and operates
by a plasma-assisted etching from the gas phase. The process is capable to produce some 100 µm deep holes
and ditches
88
. The common integration of fluidics and sensor elements is of particular importance for
impedance spectroscopy at high frequencies, since the damping of electromagnetic radiation severely
increases with increasing frequency. In particular, the objects of interest have to be very close to the sensor
chip, when frequencies in the GHz range and in the THz range shall be applied for the investigation, i.e. the
distance must be less than about 100 µm, with the precise value depending upon the chosen frequency, the
electrode geometry and the amount of heat tolerable during the investigation.
Page 10 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 11
A two wafer approach as already developed
89
may be applied to position the microfluidics in close proximity to
the sensor elements of the integrated circuit
90, 91
. The wafer carrying the BiCMOS circuit is thereby designated
as BiCMOS wafer or top wafer, whereas the bottom wafer supplies the fluid inlet and outlet that are prepared
as cylindrical holes by the DRIE process. In addition, the top wafer encompasses another microfluidic channel,
which is positioned directly below the integrated circuit, by which the fluid is probed. An alignment technology
for backside integration was to develop in advance to allow for DRIE of the wafer backside to align the micro-
channel under the sensor circuit with a precision in the 1-2 µm range
90
.
It can be seen from Figure 14 that the silicon below the sensor area is completely etched off, facilitating the
fluid to stream along the bottom side of the electronic circuit. The Figure also displays the stacking of top and
bottom wafer in the area of the active circuit. Critical parameters are the fracture toughness of the BEoL stack
and the Reynolds number of the flow in the top wafer. It is outlined in Ref.
91
that flow rates of some 100 µL s
-1
might be realized with the two-wafer configuration, even for channel heights of only 50 µm in the BiCMOS
wafer.
Various projects have thus demonstrated the feasibility to integrate microelectronic circuits into microfluidic
platforms. Life scientists may formulate their experimental constrains and intention to configure in cooperation
with semiconductor technologists the appropriate lab-on-a-chip devices with combined components.
8 Multi-project-wafer service
In order to enable access to its technology for external users so-called multi-project wafer MPW shuttles are
regularly processed at IHP (in general, such a service is offered by various semiconductor fabs). The ICs of
several customers are prepared in these runs for one of the offered technologies that differ with respect to the
performance parameters of MOSFETs and HBTs. Table 1 gives an overview of the technologies offered by IHP’s
MPW service and device characteristics.
For participation the MPW service, a potential user has to login and upload his layout data for a chosen
technology run via the IHP webserver. After a formal procedure, the design rules as well the parameter cells for
the most important devices will be supplied allowing for the design of customer-specific circuits. Generally,
these tools have to be applied by scientists and developers from the electronic disciplines. Life scientists with
interest in biochips are thus recommended to cooperate with circuit designers.
Submitted designs are combined into a common test field, which also comprises several hundred test
structures and alignment markers, see Figure 11C. After this the set of lithography masks may be prepared.
Prior to clean room processing the appropriate wafer type has to be selected, whereas for product flows
typically wafers with 2 Sm
-1
und 〈001〉 orientation are chosen. Wafers are collected into a lot that usually
comprises of 24 pieces, see Figure 1B. Processing starts with the labelling of each wafer, which is followed by
various cleaning steps. Subsequently, the essential set-up of the integrated circuit begins. It is divided into
different modules like TRENCH, TRCHFI, BLAYER etc., which stands for the separation of active areas,
preparation of buried layers and so on.
Wafer processing is finished for 0.25 and 0.13 µm technologies about 2½ - 3 month after starting the lot and
after a few 100 process steps. Chips then have to be separated, which is usually performed by automated
circular-blade saws. Beforehand, the wafer is subjected to thinning to reduce its original thickness to 300 µm or
another customer-specified value. Next to conventional sawing the technique of laser dicing may be applied for
chip separation
92
, which practically avoids any particle generation. Finally, the separated microchips are
delivered for experimentation or system integration.
9 Conclusions
Next to micro- and nano-electronics no other technology exists that operates on length scales so close to the
molecular structures of living. New bio-analytical and bio-sensor principles are enabled by them that may
directly be integrated into microelectronic chips allowing for the detection, transduction, analysis and
otherwise intelligent processing of measured data within smallest regions of space. Various semiconductor fabs
offer their assistance to life scientists for the fabrication of BioASICs by usage of multi-project wafer services.
The pilot line of IHP disposes of a set of 0.25 and 0.13 µm CMOS/BiCMOS enabling, for instance, the
preparation of MEMS devices with mechanical membranes from bio-stable TiN, di-electrophoretic electrode
Page 11 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 12
arrays, impedimetric measurement chips in the GHz range, HBTs with transient frequencies in the few 100 GHz
range and Si-integrated microfluidics to mention only a few technology modules, from which full bio-systems
might be configured. The usage of microelectronic techniques is expected to pave the way for substantial
progress in various fields of life science research and development. It still suffers from the so-far disjunctive
terminology between both disciplines, but the authors hope that the presented review might foster a broader
utilization of microelectronics by life scientists.
Acknowledgement
We thank our colleagues from IHP that helped during the collection of this overview and which enabled to offer
the technology modules presented above to external customers. We likewise thank our partners that
participated in project translation and the German Federal Ministry of Research and Technology (grant
numbers 0313862B and 16SV3934), German Federal Ministry of Economics (KF 0653901 UL8), the European
Regional Development Fund and the German state Brandenburg (TeraSens project) for funding cooperation
projects, the results of which were presented here. We particularly thank Aktionszentrum BioTOP, the
diagnostic network Berlin-Brandenburg, IGE from TU Berlin and the center for molecular diagnostics and
biotechnology ZMDB for long-standing and manifold support in transferring microelectronics to the life
sciences, and last, but not least, the Chair of Bioprocess Engineering of TU Berlin, with which the Joint Lab
Bioelectronics IHP/TUB was founded in 2012.
Page 12 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 13
References
1. Cumming DRS. ISFET technology for multi-sensor technology: from lab-in-a-pill to massively parallel
measurements. Electronics Letters 2011. Available at: http://digital-
library.theiet.org/content/journals/10.1049/el.2011.2891.
2. Graham AHD, Robbins J, Bowen CR, Taylor J. Commercialisation of CMOS Integrated Circuit
Technology in Multi-Electrode Arrays for Neuroscience and Cell-Based Biosensors. Sensors 2011,
11:4943-4971.
3. Rothe J, Lewandowska MK, Heer F, Frey O, Hierlemann A. Multi-target electrochemical biosensing
enabled by integrated CMOS electronics. J. Micromech. Microeng. 2011, 21:054010.
4. Pan T, Wang W. From Cleanroom to Desktop: Emerging Micro-Nanofabrication Technology for
Biomedical Applications. Annals of Biomedical Engineering 2011, 39:600-620.
5. Kumar S, Prakash R, Chouhan S, Salhan AK. CMOS sensors for microscopy, spectrophotometry and as a
transducer in biosensors. In: Point-of-Care Healthcare Technologies (PHT): IEEE; 2013.
6. Ghallab YH, Ismail Y. CMOS Based Lab-on-a-Chip: Applications, Challenges and Future Trends. Circuits
and Systems Magazine, IEEE 2014, 14:27-47.
7. Moore GE. Cramming more components onto integrated circuits. Electronics 1965, 38:114-117.
8. Dennard RH, Gaensslen FH, Yu H-N, Rideout VL, Bassous E, LeBlanc A. Design of Ion-Implanted
MOSFETs with Very Small Physical Dimensions. IEEE J. Solid-State Circuits 1974, 9:256-268.
9. Bohr M. A 30 Years Retrospective on Dennard's MOSFET Scaling Paper. IEEE Solid State Circuits Society
News 2007, 12:11-13.
10. ITRS EC. 2013 Edition - Executive Summary. International Technology Roadmap for Semiconductors
2013. Available at: http://www.itrs.net/Links/2013ITRS/2013Chapters/2013ExecutiveSummary.pdf.
11. Arden W, Brillouët M, Cogez P, Graef M, Huizing B, Mahnkopf R. More-than-Moore - White Paper.
2011.
12. Fromherz P, Offenhausser A, Vetter T, Weis J. A neuron-silicon junction: a Retzius cell of the leech on
an insulated-gate field-effect transistor. Science 1991, 252:1290-1293.
13. Fromherz P. Joining microelectronics and microionics: Nerve cells and brain tissue on semiconductor
chips. Solid-State Electronics 2008, 52:1364-1373.
14. Panaitov G, Thiery S, Hofmann B, Offenhäusser A. Fabrication of gold micro-spine structures for
improvement of cell/device adhesion. Microelectronic Engineering 2011, 88:1840-1844.
15. Schuettler M, Donaldson N, Seetohul V, Taylor J. Fibre-selective recording from the peripheral nerves
of frogs using a multi-electrode cuff. Journal of Neural Engineering 2013, 10:036016.
16. Ballini M, Muller J, Livi P, Yihui C, Frey U, Stettler A, Shadmani A, Viswam V, Lloyd Jones I, Jackel D, et
al. A 1024-Channel CMOS Microelectrode Array With 26,400 Electrodes for Recording and Stimulation
of Electrogenic Cells In Vitro. Solid-State Circuits, IEEE Journal of 2014, 49:2705-2719.
17. Birkholz M, Zaumseil P, Bauer J, Bolze D, Weidner G. Small-Angle X-ray Reciprocal Space Mapping of
Surface Relief Gratings. Mater. Sci. Eng. C 2007, 27:1154-1157.
18. Osten HJ, Lippert G, Knoll D, Barth R, Heinemann B, Rucker H, Schley P. The effect of carbon
incorporation on SiGe heterobipolar transistor performance and process margin. In: Electron Devices
Meeting, 1997. IEDM '97. Technical Digest., International; 1997.
19. Heinemann B, Barth R, Bolze D, Drews J, Fischer GG, Fox A, Fursenko O, Grabolla T, Haak U, Knoll D, et
al. SiGe HBT Technology with fT/fmax of 300 GHz/500 GHz and 2.0 ps CML, Gate Delay. In:
International Electron Devices Meeting (IEDM); 2010.
20. Martienssen W, Warlimont H, eds. Handbook of Condensed Matter and Materials Data. Berlin:
Springer; 2005, 1120.
21. Hämmerle H, Kobuch K, Kohler K, Nisch W, Sachs H, Stelzle M. Biostability of micro-photodiode arrays
for subretinal implantation. Biomat. 2002, 23:797-804.
22. Graham AHD, Bowen CR, Taylor J, Robbins J. Neuronal cell biocompatibility and adhesion to modified
CMOS electrodes. Biomed. Microdev. 2009, 11:1091-1101.
23. Fröhlich M, Birkholz M, Ehwald K-E, Kulse P, Fursenko O, Katzer J. Biostability of an implantable
glucose sensor chip. IOP Conference Series: Materials Science and Engineering 2012, 41:012022.
24. Lisker M, Trusch A, Krüger A, Fraschke M, Tillack B, Weimann N, Ostermay I, Krüger O. Silicon Nitride
Stop Layer in Back-End-of-Line Planarization for Wafer Bonding Application. International Conference
on Planarization/CMP Technology (ICPT) 2014.
Page 13 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 14
25. Birkholz M, Ehwald K-E, Wolansky D, Costina I, Baristiran-Kaynak C, Fröhlich M, Beyer H, Kapp A, Lisdat
F. Corrosion-resistant metal layers from a CMOS process for bioelectronic applications. Surf. Coat.
Technol. 2010, 204:2055-2059.
26. Birkholz M, Ehwald K-E, Kulse P, Drews J, Fröhlich M, Haak U, Kaynak M, Matthus E, Schulz K,
Wolansky D. Ultrathin TiN membranes as a technology platform for CMOS-integrated MEMS and
BioMEMS devices. Adv. Func. Mat. 2011, 21:1652-1656.
27. Kulse P, Birkholz M, Ehwald K-E, Bauer J, Drews J, Haak U, Höppner W, Katzer J, Schulz K, Wolansky D.
Fabrication of MEMS actuators from the BEOL of a 0.25 µm BiCMOS technology platform. Microelectr.
Eng. 2012, 97:276-279.
28. Hrapovic S, Liu Y, Enright G, Bensebaa F, Luong JHT. New Strategy for Preparing Thin Gold Films on
Modified Glass Surfaces by Electroless Deposition. Langmuir 2003, 19:3958-3965.
29. Lee JM, Park HK, Jung Y, Kim JK, Jung SO, Chung BH. Direct Immobilization of Protein G Variants with
Various Numbers of Cysteine Residues on a Gold Surface. Anal. Chem. 2007, 79:2680-2687.
30. Kitzmann J. Strukturierte Goldabscheidung mittels Galvanik und PVD. 2011.
31. Schnyder B, Kötz R, Alliata D, Facci P. Comparison of the self-chemisorption of azurin on gold and on
functionalized oxide surfaces. Surf. Interface Anal. 2002, 34:40-44.
32. Birkholz M, Zaumseil P, Kittler M, Wallat I, Heyn MP. Structure of biomembrane-on-silicon hybrids
derived from X-ray Reflectometry. Mater. Sci. Eng. B 2006, 134:125-129.
33. Birkholz M. A thin film approach to protein crystallography. Nucl. Instr. Meth. Phys. B 2009, 268:414-
419.
34. Ratzke M. Kraftmikroskopie als Werkzeug zur Charakterisierung elektrischer
Oberflächeneigenschaften. Physik 2011. Vol. Doktor.
35. Israelachvili J. Intermolecular and Surface Forces. 2nd ed. Amsterdam: Academic Press; 1991.
36. Zaumseil P, Birkholz M, Weidner G. X-ray Characterization of Periodic Sub-nm Surface Relief Gratings.
phys. stat. sol. a 2007, 204:2657-2661.
37. Li Y, Huang G, Zhang X, Wang L, Du Y, Lu TJ, Xu F. Engineering cell alignment in vitro. Biotechnology
Advances 2014, 32:347-365.
38. Körtge A, Stählke S, Lange R, Birkholz M, Fraschke M, Schulz K, Nebe B, Elter P. Alignment of MG-63
Osteoblasts on Fibronectin-Coated Phosphorous Doping Lattices in Silicon. World Academy of Science,
Engineering and Technology 2013, 73:801-804.
39. Otto S, Kaletta UC, Bier F, Wenger C, Hölzel R. Dielectrophoretic Immobilisation of Antibodies on
Microelectrode Arrays. Lab on a Chip 2014, 14:998-1004.
40. Bashir R. BioMEMS: state-of-the-art in detection, opportunities and prospects. Adv. Drug Delivery Rev.
2004, 56:1565-1586.
41. Grayson ACR, Shawgo RS, Johnson AM, Flynn NT, Li Y, Cima MJ, Langer RS. A BioMEMS Review: MEMS
Technology for Physiologically Integrated Devices. Proc. IEEE 2004, 92:6-21.
42. Sumner JB, Gralén N, Eriksson-Quensel I-B. The molecular weights of Concanavalin, Concanavalin A
and Concanavalin B. J. Biol. Chem. 1938:45-48.
43. Bouckaert J, Dewallef Y, Poortmans F, Wyns L, Loris R. The Structural Features of Concanavalin A
Governing Non-proline Peptide Isomerization. J. Biol. Chem. 2000, 275:19778-19787.
44. Ehwald R, Ballerstadt R, Dautzenberg H. Viscosimetric Affinity Assay. Anal. Biochemistry 1996, 234:1-8.
45. Schultz JS, Sims G. Affinity Sensors for Individual Metabolites. Biotechnol. Bioeng. Sympos. 1979. Vol.
9, Pages 65-71.
46. Schultz JS, Mansouri S, Goldstein IJ. Affinity sensor: a new technique for developing implantable
sensors for glucose and other metabolites. Diabetes Care 1982, 5:245-253.
47. Ballerstadt R, Evans C, Gowda A, McNichols R. Fiber-Coupled Fluorescence Affinity Sensor for 3-Day in
Vivo Glucose Sensing. J. Diabetes Sci. Technol. 2007, 1:218-227.
48. Nielsen JK, Christiansen JS, Kristensen JS, Toft HO, Hansen LL, Aasmul S, Gregorius K. Clinical Evaluation
of a Transcutaneous Interrogated Fluorescence Lifetime-Based Microsensor for Continuous Glucose
Reading. J. Diabetes Sci. Technol. 2009, 3:98-109.
49. Dutt-Ballerstadt R, Evans C, Pillai AP, Orzeck E, Drabek R, Gowda A, McNichols R. A human pilot study
of the fluorescence affinity sensor for continuous glucose monitoring in diabetes. J Diabetes Sci
Technol. 2012, 6:362-370.
50. Huang X, Siqi L, Schultz J, Qian W, Lin Q. A Capacitive MEMS Viscometric Sensor for Affinity Detection
of Glucose. Microelectromechanical Systems, Journal of 2009, 18:1246-1254.
51. Huang X, Li S, Schultz JS, Wang Q, Lin Q. A dielectric affinity microbiosensor. Applied Physics Letters
2010, 96:033701.
Page 14 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 15
52. Xian H, Siqi L, Davis E, Dachao L, Qian W, Qiao L. A MEMS Dielectric Affinity Glucose Biosensor.
Microelectromechanical Systems, Journal of 2014, 23:14-20.
53. Ballerstädt R, Ehwald R. Suitability of aqueous dispersions of dextran and Concanavalin A for glucose
sensing in different variants of the affinity sensor. Biosens. Bioelectr. 1994, 9:557-567.
54. Beyer U, Schäfer D, Thomas A, Aulich H, Haueter U, Reihl B, Ehwald R. Recording of subcutaneous
glucose dynamics by a viscosimetric affinity sensor. Diabetol. 2001, 44:416-423.
55. Zhao Y, Li S, Davidson A, Yang B, Wang Q, Lin Q. A MEMS viscosimetric sensor for continuous glucose
monitoring. J. Micromech. Microeng. 2007, 17:2528-2537.
56. Birkholz M, Ehwald K-E, Ehwald R, Kaynak M, Borngräber J, Drews J, Haak U, Klatt J, Matthus E, Schoof
G, et al. Mikroviskosimeter zur kontinuierlichen Glucosemessung bei Diabetes mellitus.
Mikrosystemtechnik Kongress 2009 2009:124-125.
57. Birkholz M, Ehwald K-E, Basmer T, Reich C, Kulse P, Drews J, Genschow D, Haak U, Marschmeyer S,
Matthus E, et al. Sensing glucose concentrations at GHz frequencies with a fully embedded BioMEMS.
J. Appl. Phys. 2013, 113:244904.
58. Gabriel G, Gabriel S, Grant EH, Halstead BSJ, Mingos DMP. Dielectric parameters relevant to
microwave dielectric loss. Chem. Soc. Rev. 1998, 27:213-223.
59. Pethig R. Review Article—Dielectrophoresis: Status of the theory, technology, and applications.
Biomicrofluidics 2010, 4:022811.
60. Laux EM, Wenger C, Kaletta UC, Bier FF, Hoelzel R. Functionality of Dielectrophoretically Immobilized
Enzyme Molecules. Electrophoresis 2014, 35:459.
61. Schöning MJ, Poghossian A. Bio FEDs (Field-Effect Devices): State-of-the-Art and New Directions.
Electroanalysis 2006, 18:1893-1900.
62. Hammond PA, Ali D, Cumming DRS. Design of a single-chip pH sensor using a conventional 0.6-μm
CMOS process. Sensors Journal, IEEE 2004, 4:706-712.
63. Grattarola M, Massobrio G. Bioelectronics Handbook: MOSFETs, biosensors, and neurons. New York:
McGraw-Hill; 1998.
64. Rothberg JM, Hinz W, Rearick TM, Schultz J, Mileski W, Davey M, Leamon JH, Johnson K, Milgrew MJ,
Edwards M, et al. An integrated semiconductor device enabling non-optical genome sequencing.
Nature 2011, 475:348-352.
65. Nemeth B, Symes MD, Boulay AG, Busche C, Cooper GJT, Cumming DRS, Cronin L. Real-Time Ion-Flux
Imaging in the Growth of Micrometer-Scale Structures and Membranes. Advanced Materials 2012,
24:1238-1242.
66. Schöning MJ, Sauke M, Steffen A, Marso M, Kordos P, Lüth H, Kauffmann F, Erbach R, Hoffmann B. Ion-
sensitive field-effect transistors with ultrathin Langmuir-Blodgett membranes. Sensors and Actuators
B: Chemical 1995, 27:325-328.
67. Poghossian A, Schöning MJ. Chapter 14 ‘High-order’ hybrid FET module for (bio)chemical and physical
sensing. In: Alegret S, ed. Comprehensive Analytical Chemistry. Vol. Volume 39: Elsevier; 2003, 587-
623.
68. Poghossian A, Ingebrandt S, Offenhäusser A, Schöning MJ. Field-effect devices for detecting cellular
signals. Seminars in Cell & Developmental Biology 2009, 20:41-48.
69. Guha S, Jamal FI, Schmalz K, Wenger C, Meliani C. CMOS lab on a chip device for dielectric
characterization of cell suspensions based on a 6 GHz Oscillator. In: European Microwave Conference,
European Microwave Week 2013, (EuMW 2013). Nürnberg; 2013.
70. Guha S, Jamal FI, Schmalz K, Wenger C, Meliani C. An 8 GHz CMOS Near Field Bio-sensor Array for
Imaging Spatial Permittivity Distribution of Biomaterials. In: International Microwave Symposium (IMS
2014); 2014.
71. Couniot N, Afzalian A, Van Overstraeten-Schlögel N, Francis LA, Flandre D. Capacitive biosensing of
bacterial cells: Analytical model and numerical simulations. Sensors and Actuators B: Chemical 2015,
211:428-438.
72. Jamal FI, Guha S, Meliani C. A SiGe BiCMOS Dielectric Sensor utilizing an Open-ended Microstrip Line in
a 28 GHz Colpitts Oscillator. In: European Microwave Week (EuMW): IEEE; 2014.
73. Nishizawa J-i, Sasaki T, Suto K, Tanabe T, Saito K, Yamada T, Kimura T. THz transmittance
measurements of nucleobases and related molecules in the 0.4- to 5.8-THz region using a GaP THz
wave generator. Optics Communications 2005, 246:229-239.
74. Saha SC, Grant JP, Ma Y, Khalid A, Hong F, Cumming DRS. Application of terahertz spectroscopy to the
characterization of biological samples using birefringence silicon grating. Journal of Biomedical Optics
2012, 17:0670061-0670065.
Page 15 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 16
75. Wolf A. Aufbau und Funktionsweise eines kontinuierlich-strahlenden Terahertz-Spektrometers. 2013.
76. Schmalz K, Mao Y, Borngräber J, Neumaier P, Hübers HW. Tunable 500 GHz transmitter array in SiGe
technology for gas spectroscopy. Electronics Letters 2015, 51:257-259.
77. Basmer T, Kulse P, Birkholz M. Systemarchitektur intelligenter Sensorimplantate. Biomed.
Eng./Biomed. Tech. 2010, 55:P43.
78. Johansson AJ. Performance of a radio link between a base station and a medical implant utilising the
MICS standard. In: Engineering in Medicine and Biology Society, 2004. IEMBS '04. 26th Annual
International Conference of the IEEE; 2004.
79. ETSI. Electromagnetic compatibility and Radio spectrum Matters (ERM); Short Range Devices (SRD);
Ultra Low Power Active Medical Implants (ULP-AMI) and Peripherals (ULP-AMI-P) operating in the
frequency range 402 MHz to 405 MHz; Part 1: Technical characteristics and test methods. ETSI EN 301
839-1 V1.3.1 (2009-10) 2010. Vol. ETSI EN 301 839-1 V1.3.1 (2009-10), Page 52.
80. Basmer T, Genschow D, Fröhlich M, Birkholz M. Energy budget of an implantable glucose
measurement system. Biomed. Tech./Biomed. Eng. 2012, 57:259-262.
81. Birkholz M, Glogener P, Basmer T, Glös F, Genschow D, Welsch C, Ruff R, Hoffmann KP. System
integration of a silicone-encapsulated glucose monitor implant. Biomed. Eng./Biomed. Tech. 2014,
59(s1):S1089 – S1092.
82. Dillschneider R, Posten C. Closed Bioreactors as Tools for Microalgae Production. In: Lee JW, ed.
Advanced Biofuels and Bioproducts: Springer New York; 2013, 629-649.
83. Todtenberg N, Basmer T, Klatt J, Schmalz K. Estimation of 433 MHz Path Loss in Algae Culture for
Biosensor Capsule Application. In: European Microwave Conference (EuMC 2013); 2013.
84. Todtenberg N, Schmitz-Hertzberg S-T, Schmalz K, Klatt J, Jorde F, Juttner B, Kraemer R. Autonomous
Sensor Capsule for Usage in Bioreactors. Sensors Journal, IEEE 2015, 15:4093-4102.
85. Baratchi S, Khoshmanesh K, Sacristan C, Depoil D, Wlodkowic D, McIntyre P, Mitchell A. Immunology
on chip: Promises and opportunities. Biotechnology Advances 2014, 32:333-346.
86. Kling J. Moving diagnostics from the bench to the bedside. Nat. Biotechnol. 2006, 24:891-893.
87. Terry SC, Jerman JH, Angell JB. A gas chromatographic air analyzer fabricated on a silicon wafer.
Electron Devices, IEEE Transactions on 1979, 26:1880-1886.
88. Marschmeyer S, Fraschke M, Milewski T, Schulze S. Characterization of Fluorocarbon Polymers
Deposited with Different Chamber Pressure and Source Power in the Deposition Cycle of the Bosch
Process. In: Plasma Etch and Strip in Microelectronics (PESM); 2011.
89. Hwang T, Popa D, Sin J, Stephanou HE, Leonard EM. BCB wafer bonding for microfluidics. In:
Micromachining and Microfabrication Process Technology IX. San Jose, CA: SPIE; 2004.
90. Bauer J, Kulse P, Haak U, Kaynak M, Ehwald K-E, Marschmeyer S, Birkholz M, Schulz S, Old G, Scheuring
G, et al. Alignment Technology for Backside Integration. Proceedings SPIE 2011, 7985:798508.
91. Baristiran-Kaynak C, Kaynak M, Wietstruck M, Marschmeyer S, Kulse P, Schulz K, Silz H, Krüger A, Barth
R, Schmalz K, et al. Modeling and Characterization of BiCMOS Embedded Microfluidic Platform for
Biosensing Applications. In: Silicon Monolithic Integrated Circuits in RF Systems; 2014.
92. Birkholz M, Ehwald K-E, Kaynak M, Semperowitsch T, Holz B, Nordhoff S. Separation of extremely
miniaturized medical sensors by IR laser dicing. J. Opt. Adv. Mat. 2010, 12:479-483.
Page 16 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 17
Tables
Table 1
SG25H1
H
igh
-
performance 0.25 µm technology with
npn
-
HBTs up to
f
T
/
f
max
=
180/220 GHz
(compare Figure 4A)
.
SG25H3 0.25 µm technology with a set of npn-HBTs ranging from a higher RF performance (f
T
/f
max
= 110/180 GHz) to
higher breakdown voltages up to 7 V.
SGB25V Cost-effective technology with a set of npn-HBTs up to a breakdown voltage of 7 V.
SG13S High-performance 0.13 µm BiCMOS with npn-HBTs up to f
T
/f
max
= 250/300 GHz, with 3.3 V I/O CMOS and 1.2
V logic CMOS.
SG13G2 0.13 µm BiCMOS technology with the same device portfolio as SG13S, but much higher bipolar performance
with f
T
/f
max
= 300/500 GHz.
Page 17 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 18
Figure Legends
Figure 1. A. View into IHP clean room. B. Wafer lot consisting of 24 Si Wafers having a diameter of 200 mm. C.
An almost full-processed wafer showing the separation of the surface into single dies of the same test field.
Figure 2. SEM micrograph of an intermediate step during the preparation of a doping lattice. The picture was
taken after implantation of n-doped areas and prior to etching off the photoresist.
Figure 3. A. TEM micrographs of increasing magnification from the cross section of a MOSFET having a channel
length of 130 nm as routinely produced in 0.13 µm CMOS technologies. Source and drain regions are situated
directly below the vertical tungsten plugs (W VIAs). The thin SiO
2
film covering the channel exhibits a thickness
of 2 nm. C. TEM cross section micrograph of a most-recent hetero-junction bipolar transistor (HBT); the emitter
and the base, consisting of 20 nm thin SiGe:C, have been emphasized. The electrical connection to a tungsten
VIA and the lowest metal layer M1 can clearly be recognized.
Figure 4. A. Schematic drawing of SG25H1 chip architecture in 0.25 µm technology. Blue areas are formed by
ILD layers that have to establish the electrical isolation between electrical connections. Metal layers M1…M5
are horizontally oriented, while vertical currents are transported by VIAs. Active devices like n-type and p-type
MOSFETs as well as a HBTs are formed directly on the Si wafer and can be seen at the bottom. The layer stack is
terminated on top by a passivation layer of silicon nitride (grey) having openings only in the area of bond pads.
B. SEM cross section micrograph of a chip architecture processed until to ILD3. The framing of metal layers M1
to M3 into a top and bottom TiN layer may clearly be recognized as well as the VIAs connecting them.
Figure 5. A. Optical microscope picture of the surface of a microchip after storing it in electrolyte solution for
some days. A top-most metal layer has been subjected to corrosion as can be recognized by the color change
from yellow to green. SEM cross section micrograph of the same area shows a penetration of the passivation
layer, which turned out to be caused by the topography variation at the edge of TM1-induced surface
protrusion. Such defects can be avoided by usage of an alternative passivation nitride and by a CMP step for
planarization. C. Cyclic voltammogramm of a gold layer and a thin TiN film exposing the same area into 5 mM
K
4
{Fe(CN)
6
}/ 5 mM K
3
{Fe(CN)
6
} in 100 mM pH 7.4 phosphate buffer. The TiN electrode exhibits an almost zero
redox turnover when compared with the currents crossing the gold electrode, the latter exceeding 100 µA.
Figure 6. A. Color-scale representation of electric potential distribution in and above a doping lattice as
calculated by an FEM simulation for N
D
= 10
20
cm
-3
. B. Course of surface potentials for the different cases of
N
D
= 10
16
, 10
17
, 10
18
, 10
19
and 10
20
cm
-3
. Calculations were performed for the vacuum case
34
, which has to be
modified in aqueous solutions due to shielding by electrolyte ions and water dipoles. C. SEM view on a doping
lattice upon which MG 63 osteoblasts were cultivated. The orientation of cells along the direction of the lattice
(white arrow) can clearly be recognized
38
.
Figure 7. Fabrication process of microelectrode array for dielectrophoretic immobilization: a) metal deposition;
b) SiO
2
deposition, CMP and VIA etch; c) tungsten filling of VIAs and subsequent CMP; d) 3D side view; e)
confocal reflection microscopy of a part of the array, scale 50 µm; f) detail view of e), scale 2 µm; g) SEM cross
section micrograph of a single tungsten electrode with embedding SiO
2
and metal layer beneath, scale bar 500
nm
39
.
Figure 8. A. Scheme of MEMS concept and operation of an affinity sensor. A cavity is filled by the assay
encompassing the receptor (small red balls) and the polymer of the analyte (large blue spheres). The network
formed by macromolecular receptors and polymers partially decomposes under inserting the monomeric
analyte, which leads to a change in viscosity. B. SEM picture of an assay-free MEMS with mechanically aving the
shape of an X. Beam and ground plate are prepared from biostable TiN with the beam thickness amounting to
50 nm only. C. The elastically restoring element in the middle of the beam takes the shape of an open double
U. D. The beam is formed immoveable for a parallel reference measurement by closing the double U.
Figure 9. A. Measurement set-up for dielectrophoretic electrode array on a microscope slide (75 x 25 mm). B.
Dielectrophoretic immobilization of anti-RPE anti-bodies and subsequent incubation with RPE. An alternating
voltage of 100 kHz and 18 V
rms
has been applied for 20 min. The Figure depicts the superposition of a reflection
Page 18 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Birkholz et al.: Technology modules … 19
picture of the array (grey values converted to violet) and a fluorescence picture (grey values converted to
yellow).
Figure 10. Affinity-viscosimetric sensor chip. A. Electrical circuit of affinity sensor chip: the DC voltage
introduced is converted to a high-frequency AC voltage of 3.2 GHz via two ring oscillator circuits. The
configuration of beam and ground plate acts as a capacitor C with serial and parallel resistors R (shaded areas).
C and R may vary due to beam bending in the measurement circuit, which is indicated by arrows; both
quantities remain constant, however, during the measurement in the reference circuit. B. Chip photograph,
from which the measurement and reference MEMS can be recognized on top and at the bottom; also the
position of the different components on the chip for the signal transduction cascade are indicated. C. Transient
of the measurement parameter switching time t
sw
for test solutions with varying glucose concentration c
g
and
temperature T as well as D. gauge curves t
sw
(c
g
) derived from them with T as a parameter.
Figure 11. A. Schematic drawing of real and imaginary part ε‘ and ε“ of an aqueous protein solution
58
. The
maximum of ε“ around 10
8
Hz is due to absorbing proteins, while the absorption maximum above 10 GHz is
caused by water molecules. B. Layout of an impedance measurement chip that can be used for the
determination of cell densities. The upper part shows square-sized bond pads, while sensor elements that will
come in contact with the biomilieu are arranged in the middle of the chip. The distance between both was
chosen rather large in order to establish a sufficient blocking of the biomilieu and protecting the electrical
contacts from corrosion. C. Schematic overview of a test field in a 0.25 µm CMOS/BiCMOS technology
occupying an area of 10.7 x 25.9 mm on a 200 mm wafer. There are about 70-120 exemplars of the same test
field to be processed on each wafer of a production lot.
Figure 12. A. General system architecture of a bio-sensor implant. B. PCB of with microcontroller µC, radio
module (ZL70321) and ZIF connector for connecting to sensor chip. The lateral extension of 27 mm fits to that
of the battery positioned below. C. 3D integration scheme for antenna (1), sensor probe (2), flexible cable for
connecting to ZIF (3), printed circuit board (4), antenna adapter (5), µC (6), battery (7) and distance holder (8).
D&E. Top and bottom view of a silicone-encapsulated biosensor implant. The sensor probe is positioned in the
middle of the system in the left figure; in the picture to the right one may recognize the D-shaped battery
through the silicone. The outer dimensions of the implant amount to 38.6 × 49.3 × 15.5 mm.
Figure 13. Autonomous sensor capsule for application in photo-bioreactor with algae cultivation. A. Outward
appearance; B. configuration of sensor surfaces on equatorially positioned PCB; C. photographs of both sides of
the PCB: with microcontroller and transceiver (left) and sensor field (right). The semi-sphere above the sensor
field is permeable to the cultivation broth, while the opposite side is impermeable as is the PCB layer
84
.
Figure 14. Approach for microfluidics integration into microelectronic chips: A. conceptual sketch; B. FEM
simulation of flow; C. cross section micrograph of the produced wafer stack; D. optical microscope view on HF
impedance sensor with microfluidics channel below: unfilled and E. with fluorescence-labeled liquid
91
.
Page 19 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 1
190x275mm (96 x 96 DPI)
Page 20 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 2
190x275mm (96 x 96 DPI)
Page 21 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 3 A and B
190x275mm (96 x 96 DPI)
Page 22 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 3C
190x275mm (96 x 96 DPI)
Page 23 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 4
190x275mm (96 x 96 DPI)
Page 24 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 4B
190x275mm (96 x 96 DPI)
Page 25 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 5
190x275mm (96 x 96 DPI)
Page 26 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 6A & B
190x275mm (96 x 96 DPI)
Page 27 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 6C
190x275mm (96 x 96 DPI)
Page 28 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 7
190x275mm (96 x 96 DPI)
Page 29 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 8
190x275mm (96 x 96 DPI)
Page 30 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 9
190x275mm (96 x 96 DPI)
Page 31 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 10A
190x275mm (96 x 96 DPI)
Page 32 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 10B, C & C
190x275mm (96 x 96 DPI)
Page 33 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 11
190x275mm (96 x 96 DPI)
Page 34 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 11 C
190x275mm (96 x 96 DPI)
Page 35 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 12
190x275mm (96 x 96 DPI)
Page 36 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 12 D & E
190x275mm (96 x 96 DPI)
Page 37 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 13
190x275mm (96 x 96 DPI)
Page 38 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For Peer Review
Figure 14
190x275mm (96 x 96 DPI)
Page 39 of 39
John Wiley & Sons
Wiley Interdisciplinary Reviews: Nanomedicine
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60