biosensors

Article
Silicon Integrated Dual-Mode Interferometer with
Differential Outputs
Niklas Hoppe 1, * ID , Pascal Scheck 1 , Rami Sweidan 1 , Philipp Diersing 1 , Lotte Rathgeber 1 ,
Wolfgang Vogel 1 , Benjamin Riegger 2 , Alexander Southan 2 and Manfred Berroth 1
1 Institute of Electrical and Optical Communications Engineering (INT), University of Stuttgart,
70569 Stuttgart, Germany; pascal.scheck@gmail.com (P.S.); rami.sweidan@student.uni-tuebingen.de (R.S.);
philipp.diersing@outlook.com (P.D.); lotte.rathgeber@int.uni-stuttgart.de (L.R.);
w.vogel@int.uni-stuttgart.de (W.V.); berroth@int.uni-stuttgart.de (M.B.)
2 Institute of Interfacial Process Engineering and Plasma Technology (IGVP), University of Stuttgart,
70569 Stuttgart, Germany; benjamin.riegger@igvp.uni-stuttgart.de (B.R.);
alexander.southan@igvp.uni-stuttgart.de (A.S.)
* Correspondence: niklas.hoppe@int.uni-stuttgart.de; Tel.: +49-711-6856-7918

Received: 30 June 2017; Accepted: 8 September 2017; Published: 14 September 2017

Abstract: The dual-mode interferometer (DMI) is an attractive alternative to Mach-Zehnder

interferometers for sensor purposes, achieving sensitivities to refractive index changes close to
state-of-the-art. Modern designs on silicon-on-insulator (SOI) platforms offer thermally stable and
compact devices with insertion losses of less than 1 dB and high extinction ratios. Compact arrays
of multiple DMIs in parallel are easy to fabricate due to the simple structure of the DMI. In this
work, the principle of operation of an integrated DMI with differential outputs is presented which
allows the unambiguous phase shift detection with a single wavelength measurement, rather than
using a wavelength sweep and evaluating the optical output power spectrum. Fluctuating optical
input power or varying attenuation due to different analyte concentrations can be compensated
by observing the sum of the optical powers at the differential outputs. DMIs with two differential
single-mode outputs are fabricated in a 250 nm SOI platform, and corresponding measurements are
shown to explain the principle of operation in detail. A comparison of DMIs with the conventional
Mach-Zehnder interferometer using the same technology concludes this work.

Keywords: integrated photonics; lab-on-a-chip; dual-mode; interferometer

1. Introduction
The increasing demand for point of care diagnostics and highly sensitive monitoring in today’s
process technology has led to huge efforts in miniaturizing optical sensors. Additionally, the attractive
preliminary studies of integrated Raman sensors [1], chip-integrated interferometers that are sensitive
to refractive index changes, are promising candidates for diverse applications. Surface functionalization
on top of waveguides acts as the recognition element for specific substances [2]. By implementing
photodiodes and read-out electronics on the same SOI chip, a compact sensing platform with a large
number of interferometers in parallel can be built, which is a key element for a lab on a chip. Compared
with the state-of-the-art sensitivity of such non-resonant interferometers [3], the so-called dual-mode
or bimodal interferometer (DMI) represents a fairly unknown but very promising design. The DMI
uses two optical paths in a single waveguide by utilizing two different optical modes. These two
modes of a single waveguide are confined differently, resulting in different penetration depths of the
evanescent fields in the cladding or environment. Because it requires only one waveguide, this design
is much more compact than its counterpart, the Mach-Zehnder interferometer (MZI). In addition,
further isolation layers to passivate the reference waveguide are avoided. Moreover, both modes

Biosensors 2017, 7, 37; doi:10.3390/bios7030037 www.mdpi.com/journal/biosensors

Biosensors 2017, 7, 37 2 of 10

in a DMI are at nearly the same temperature, due to their close proximity; they are not thermally
separated as in common MZI designs. In this work, operating principles for an extended DMI design
are presented, which allow the determination of the phase relation of the two modes with a single
wavelength
Biosensors 2017,measurement
Biosensors 2017, 7, 37
7, 37 [4]. The paper is structured as follows: A short 22 of summary of DMI designs is
of 10
10
Biosensors 2017, 7, 37 2 of 10
Biosensors 2017, 7, 37 2 of 10
given to illustrate
avoided. Moreover,
Moreover, both the advantages
both modes
modes inin aa DMI and
DMI are
are at drawbacks
at nearly
nearly the
the same of existing
same temperature, designs.
temperature, duedue to
to their The
close theory of basic and single
their close
avoided.
avoided. Moreover, both modes in a DMI are at nearly the same temperature, due to their close
wavelength
avoided.
proximity; operation
proximity;Moreover,
they
they are not
are not is explained
boththermally
modes separated
inseparated
a DMI are in detail
asatin common
nearly inMZI
the Section
MZI
same designs.3.In
InThe
temperature,thisdueperformance
work, operating
to their close parameters of the DMI
proximity;
principles for they
for anare not thermally
thermally
extended separated
DMI design
design
as
as in
in common
are presented,
presented,common
whichMZI
designs.
allowdesigns. In this
this work,
work,
the determination
determination
operating
operating
of the
the phase
proximity;
principles they an are not thermally
extended DMI separated
are as in common
which MZI
allow designs.
the In this work,of operating
phase
design, such
principles
relation of
principles
for
of for
as
theanan insertion
twoextended DMI
modes with
extended
loss (IL)
design
with adesign
DMI a single
and
are
singleare
extinction
presented,
wavelength which ratio
allow
measurement
presented,measurement
which allow the
(ER),
the are analyzed
determination
[4]. determination
The paper
paper is of the in
phase
is structured
structured
of the phase as
Section 4. A comparison to
relation
relation of the
the two
two modes
modes with a single wavelength
wavelength measurement [4].
[4]. The
The paper is structured as
as
MZIs using
follows:
relation
follows: of
A thetwo
A short
short
the same
summary
modes
summary technology
of DMI
with
of DMI designs
a single (Section
is given
wavelength
designs is to5)
given to and athe
illustrate
measurement
illustrate summary
the advantages
[4]. concludes
The paper
advantages and
and drawbacks
is drawbacks
structured thisof work.
as
of
follows: A short The
existing designs.
designs. summary
theoryof ofDMI
basicdesigns is given
and single
single to illustrate
wavelength operationthe is
advantages
explained and drawbacks
in detail
detail of
in Section
Section
follows:
existing A short The
summary
theory of
of DMI
basic designs
and is given to illustrate
wavelength operationthe is
advantages
explained and
in drawbacks
in of
existing
3. The designs. Theparameters
performance theory of basicof andDMI
the single wavelength
design, such operation
as insertion is loss
explained
(IL) in detail
and in Section
extinction ratio
existing designs. Theparameters
theory of basic andDMI
single wavelength operation isloss
explained in detail in Section
2. Designs
3.
3. The
Theare of Dual-Mode
performance
performance Interferometers
of the design, such as insertion (IL) and extinction ratio
(ER),
3. Theare
(ER), analyzed in
performance inparameters
Section 4.
parameters
ofcomparison
4. A the DMI design,
to MZIs
of the DMI design,
such as insertion
using
such the sameloss
as insertion
(IL) and(Section
technology
loss
extinction
(IL) and extinction
ratioa
5) and
ratio
(ER),
summaryare analyzed
analyzed
concludes
Section
inthis
Section
work.4. AA comparison
comparison to to MZIs
MZIs using
using the
the same
same technology
technology (Section
(Section 5) 5) and
and aa
(ER),
summaryare analyzed
concludes inthis
Section
work. 4. A comparison to MZIs using the same technology (Section 5) and a
Various DMI layouts
summary concludes this work. can be found in scientific publications. While this overview is focused on
summary concludes this work.
DMI2.2. Designs
designs utilizing Interferometers
Designs of Dual-Mode
Dual-Mode only the fundamental and second-order quasi-transverse electric (TE) modes,
2. Designs of of Dual-Mode Interferometers
Interferometers
2. Designs ofDMI
Dual-Mode
layouts Interferometers
there are further
Various
Various DMI designs
layouts canwhich
can be founduse
be found eitherpublications.
in scientific
in scientific higher-order
publications. While modes
While [5],isisor
this overview
this overview two on
focused
focused different polarizations in the
on
DMIVarious
designs DMI
Various DMI
layouts
utilizing onlycan
layouts thebe
can be
found in scientific
fundamental publications.
and second-order
second-order
found in scientific publications.
While this overview
quasi-transverse electricis
While this overview
focused
(TE)
is(TE)
on
modes,
focused on
dual-mode
DMI
DMI designs
theredesigns
section
utilizing
utilizing
are further
further
[6].
designs
only
only The
the basic
fundamental
the fundamental
which use either
DMIand types
and second-order
either higher-order
higher-order
utilizing the
quasi-transverse
modesquasi-transverse
[5], or
or two
two lowest-order
electric
electric
two different
different
modes,
(TE) modes,
polarizations in
TE modes are depicted in
DMI
there designs
are utilizing
designsonly the fundamental
which use and second-order
modes quasi-transverse
[5], electric (TE) modes,
polarizations in
there
Table
the1. are further
dual-mode designs
section which
[6].which use
The basic
basic either
DMI higher-order
types modes
utilizing the
the [5],
two [5], or two
lowest-order different
TE modespolarizations
modes are depicted
depictedin
there
the are further
dual-mode designs
section [6]. The use DMI
either higher-order
types utilizing modes
two or two different
lowest-order TE polarizations
are in
the
in dual-mode
Table 1. section [6]. The basic DMI types utilizing the two lowest-order TE modes are depicted
theTable
in dual-mode
1. section [6]. The basic DMI types utilizing the two lowest-order TE modes are depicted
in Table 1.
inTable
Table 1.1. A comparison of different TE dual-mode interferometer designs is shown: The schematic
Table 1.
Table 1. A
A comparison
comparison of
of different
different TE
TE dual-mode
dual-mode interferometer
interferometer designs
designs is
is shown:
shown: The
The schematic
schematic
Table 1. A comparison of different TE dual-mode interferometer designs is shown: The schematic
buildups
Table
Table
1. A
buildups
1. A
buildups
ofcomparison
of
of
the
the designs
designs
comparison
the designs
of
of
are
are depicted
aredifferent
depicted
different
depicted
TE
ondual-mode
TE
on the left.
dual-mode
the left.
on the left.assets
interferometer
left. Corresponding
Corresponding
interferometer
Corresponding
Corresponding
designs
assets and is
designs
and
shown: The
drawbacks
is shown: The
drawbacks ofassets
of the
the
and drawbacks of the design
schematic
design
schematic
design
buildups of the designs are depicted on the assets and drawbacks of the design
buildups ofonthethe
designs are depicted on the large
left. Corresponding assets and drawbacks of the design
areare
shown
are
are shown
buildups
shownofon
shown on onthe
the the
designs
the right.
right.
right.
right. Notes:
are Notes:
0
depicted
Notes:
Notes: on the0large
indicates
00 indicates
indicates indicates
left.
large additionallarge
additional
Corresponding
additional space additional
requirements;
assets
space
space and drawbacks
requirements;
requirements; ++ space
+ of therequirements;
indicates
indicates
indicates minor
design
minor
minor + indicates minor
are shown space
additional on therequirements;
right. Notes: 0++indicates
indicateslarge
largeadditional space requirements;
on-chip arrays
arrays + indicates
possible; *** indicates
indicates minor
theoretical
are shown space
additional
additional on therequirements;
space right. Notes: 0++
requirements; ++indicates
indicates
indicateslarge
large
largeadditional
on-chip
on-chip space requirements;
arrays possible;
possible; + indicates
indicates minor
theoretical
theoretical
additional
additional
values for
space
space
for balanced
balanced
requirements;
requirements;
mode excitation.++
excitation.
++
indicates indicates
large large
on-chip on-chip
arrays possible; arrays
* possible;
indicates theoretical* indicates theoretical values
additional
values space requirements;
mode ++ indicates large on-chip arrays possible; * indicates theoretical
forvalues
values
for balanced
balanced mode
for balanced
mode
mode
excitation.
excitation.
excitation.
Single
Single
Reference/ DMI Excess Loss
Loss of
of Single
Reference/
Reference/ DMI
DMI Excess
Excess Loss of Single
Reference/ DMI Excess Loss of Wavelength
Single
Wavelength
Year
Reference/ Array
DMI Mode Loss of Excess
Conversion
Excess Wavelength
Loss of
Year
Year Array
Array
Reference/Year Mode
Mode Conversion
Conversion Wavelength
Year Array DMI Array
Mode Conversion Operation
Wavelength
Operation
Mode Conversion
Operation
Single Wavelength Operation
Year Array Mode Conversion Operation
Operation
Type111
Type
Type
Type 1 [7]/2009
[7]/2009
[7]/2009
++
[7]/2009 ++
++ 0.5 dB
++0.5
0.5 dB *
dB ** 
0.5 dB 
* 8
(side
(side
(side view)
view)
Type 1
view) [7]/2009 ++ 0.5 dB * 
(side view)
(side view) [7]/2009 ++ 0.5 dB * 
(side view)
Type CCD
Type
Type
Type 2222 CCD
CCD [8]/2011 + unknown 
Type 2 Array
CCD
Array [8]/2011
[8]/2011
[8]/2011 ++ +unknown
unknown unknown

 X
(side
(side
(side
(side
view)
view)
Type
view)
view)2 Array
CCD [8]/2011 + unknown 
(side view)
Array [8]/2011 + unknown 
(side view)
Array
Type 3
Type333
Type
Type [9]/2014 ++ 0.5 dB
dB * 
Type
(top 3
view) [9]/2014
[9]/2014
[9]/2014 ++
++ ++0.5
0.5 dB ** 
0.5 dB 
* 8
(top
(top
(top view)
Type
view)
view)3 [9]/2014 ++ 0.5 dB * 
(top view) [9]/2014 ++ 0.5 dB * 
(top view)
Type
Type
Type
Type 4444 [10]/2014 ++ < 0.22
0.22 dB
dB * *
Type
(side 4
view)
[10]/2014
[10]/2014
[10]/2014 ++
++ ++
<< 0.22 dB ** 
< 0.22 dB
 8
(side
(side
(sideview)
Type
view)
view)4 [10]/2014 ++ < 0.22 dB * 
(side view) [10]/2014 ++ < 0.22 dB * 
(side view) Fiber
Fiber
Type5555
Type
Type
Type Fiber
Type 5 Fiber [11]/2015
[11]/2015 0
00 0 <<< 4.2
4.2 dB
dB 

(top view)
Type 5 Fiber [11]/2015
[11]/2015 dB 
< 4.2 dB 8
(top
(top
(top view)
view)
view) [11]/2015 0 < 4.2 dB 
(top view) [11]/2015 0 < 4.2 dB 
(top view)
Type 666
Type 0.25 dB
0.25 dB *
Type [12]/2016
[12]/2016 ++
++ 0.25 dB ** 
Type
Type
(top
(top 666
view)
view) [12]/2016 ++ 0.25
<< 0.5dB
0.5 dB**
dB 0.25 dB
*
Type
(top view)
(topview)
view) [12]/2016 ++
[12]/2016
[12]/2016 ++ ++0.25
< 0.5
< 0.5 dB
dB
dB
 8
(top
(top view) < 0.5dB dB < 0.5 dB
Type 777
Type
Type
[4]/2017
[4]/2017
[4]/2017 ++
0.55
0.55 dB
0.55 dB *** 
Type
(top 7
view) [4]/2017
this work ++
++ 0.55
< 2 dB
dB ** 

(top
(top
Typeview)
Type 77
view)
(top view)
this work
[4]/2017
this work
this work
++
++
<< 22 dB
0.55 dB 0.55 dB
*
(topview)
view)
[4]/2017
this work
this work ++ << 22 dB
dB
X
(top < 2 dB
The main
The main difference
difference between
between the the designs
designs is is the conversion
conversion mechanism
mechanism from from aa single
single mode
mode at at the
inputThe main
to the
the difference
desired between
two modes
modes inthe
thedesigns
two-mode is the
the conversion
section the mechanism
of the interferometer,fromanda single mode In
vice versa.
versa. at the
the
input The
to main difference
desired two between in the
the designs
two-mode is the conversion
section of mechanism
interferometer,from a single
and vice modeIn at the
the
input
first DMI to the
DMI desired
type, two modes
the mode
mode splitterin andthecombiner
two-modeare section of the
realized interferometer,
as vertical
vertical waveguide andsteps.
vice versa.
Type 22In the
uses
input
first
The to the desired
type,
main the two
difference modes
splitter in
and
betweenthe two-mode
combiner thearearesection of
realized
designs the
as interferometer,
isvertical waveguide
the conversion and vice
steps. versa.
Type
mechanism In the from a single mode at the
uses
first DMI type, the mode splitter and combiner realized as waveguide steps. Type 2 uses
aaafirst
vertical
vertical step
DMI step atthe
type,at themode
the inputsplitter
input and aa Young
and Young interferometer
and combiner configuration,
are realized
interferometer including
as verticalincluding
configuration, waveguide a free
a free space
steps. section,
Typesection,
space 2 uses
input at
tooutput.
vertical
aatvertical
the
the
the desired
step
step
at the
An external
at the two
input
input
andmodes
CCD aarray
Young
and aarray
Young in the
detects two-mode
interferometer
the interference
interference
interferometer section
configuration,
pattern.
configuration, In typeof
including the
a free
3, the
including interferometer,
space
mode
a free
section,
splitter
space and
section, and vice versa. In the
at the output.
combiner output.
are
An
An external
external
realized by
CCD
CCD array detects
asymmetric detects the
thewaveguide
lateral
pattern.
interferencesteps.
pattern.
No
In type
type 3,
3, the
Inadditional mode
mode splitter
theetching splitter
process
and
andis
firstcombiner
atDMI type,
the
combiner output.
are An the mode
external
realized
are realized
by CCD splitterlateral
array
asymmetric
by asymmetric
detectsandwaveguide
lateralthe combiner
interference
waveguide areNorealized
pattern.
steps.
steps.
No In type 3,
additional
additional
asetching
the vertical
mode
etching splitter
process
process
waveguide
andis
is
steps. Type 2 uses
required
combinerhere.
required here. In the
the design
are realized
In design of type
type 4,
by asymmetric
of 4, aalateral
special
special long-period
waveguide
long-period grating
steps. No is
grating is used to
to convert
additional
used convert the
etchingthe modes,
process
modes, is
a vertical
required step
here.
enabling aahere.
required at
In
theoreticalthe
the input
design of
excess loss
In the excess
design loss and
type 4,
of mode
of type mode a Young
special
conversion
4, a special interferometer
long-period grating
of less than
long-period than is
0.22 dB.
grating configuration,
used to convert
The optical
is used
the
coupling
to convert including
modes,
to the
the modes, the a free space section,
enabling
enabling a theoretical
theoretical excess loss of
ofand
mode conversion
conversion of
of less
less than 0.22
0.22 dB.
dB. The
The optical
optical coupling
coupling to
to 1,
the2
at thesingle
single
output.
single
enabling mode
mode An
waveguides
a theoretical
waveguides
mode waveguides
external
at input
excess
at
at
input
loss of
input
CCD
and output
mode
output
andstep
output
array
conversion
is detects
is accomplished
accomplished the
of less than
is accomplished
interference
through
0.22
through further
dB. The optical
further pattern.
gratings. For type
coupling
gratings. For typeto 1,In
the2 type 3, the mode splitter
and
single444 DMIs,
DMIs, the
mode the depth of
waveguides of the
the vertical
at input andstep
outputin the waveguide or
is accomplished or through further
of the grating
through grating gratings.
grooves,
further
For type 1, is
respectively,
gratings. For type 1,is2
2
andandandcombiner
a critical
criticalDMIs, depth
theare
depth
parameter,
realized
and of has
the tovertical
by
vertical step
be controlled
controlled
in
in the
asymmetricthe waveguide
waveguide
carefully
lateral
during
of
of the
or the the grating grooves,
waveguide
fabrication grooves, respectively,
steps. No additional
process. respectively, is etching process is
aand 4 DMIs, the depth
parameter, and ofhas
the tovertical
be step in the waveguide
carefully during orthe
of the grating grooves,
fabrication process. respectively, is
a criticalhere.
required parameter,
In andthehas to be controlled
design of type carefully
4, a during
special the fabrication process.
a critical parameter, and has to be controlled carefully during the long-period
fabrication process. grating is used to convert the modes,
enabling a theoretical excess loss of mode conversion of less than 0.22 dB. The optical coupling to
the single mode waveguides at input and output is accomplished through further gratings. For type
1, 2 and 4 DMIs, the depth of the vertical step in the waveguide or of the grating grooves, respectively,
is a critical parameter, and has to be controlled carefully during the fabrication process.
The design of type 5 uses laterally displaced fibers on common grating couplers. This allows the
excitation of two modes with arbitrary power ratio, defined by the displacement of the fiber in respect
to the waveguide axis. This setup is very flexible in terms of mode excitation; however, it requires
proper and stable alignment of the fibers. For large arrays of interferometers in labs on a chip, the Young
Biosensors 2017, 7, 37 3 of 10

The design of type 5 uses laterally displaced fibers on common grating couplers. This allows the
excitation of two modes with arbitrary power ratio, defined by the displacement of the fiber in respect
Biosensors 2017, 7, 37 axis. This setup is very flexible in terms of mode excitation; however, it requires
to the waveguide 3 of 10
proper and stable alignment of the fibers. For large arrays of interferometers in labs on a chip, the
Young configuration of type 2 and the configuration with displaced fibers (type 5) are less convenient,
configuration of type 2 and the configuration with displaced fibers (type 5) are less convenient, due to
due to the additional space and adjustment requirements. In type 6 and 7, the mode splitters and
the additional space and adjustment requirements. In type 6 and 7, the mode splitters and combiners
combiners are separate parts that have fixed dimensions. Tapers are used to adapt to different
are separate parts that have fixed dimensions. Tapers are used to adapt to different waveguide widths.
waveguide widths. This makes the design of the dual-mode section independent from other
This makes the design of the dual-mode section independent from other geometries, such as input
geometries, such as input waveguide width, resulting in a modular DMI design. Moreover, the
waveguide width, resulting in a modular DMI design. Moreover, the tolerance of the splitter and
tolerance of the splitter and combiner to imperfections of the fabrication process is enhanced
combiner to imperfections of the fabrication process is enhanced compared to the type 3 DMI due
compared to the type 3 DMI due to the enlarged splitter and combiner dimensions, and no critical
to the enlarged splitter and combiner dimensions, and no critical vertical etching step is required.
vertical etching step is required. The type 7 DMI is an extended version of type 6, with a mode
The type 7 DMI is an extended version of type 6, with a mode combiner with two outputs. A more
combiner with two outputs. A more detailed view is depicted in Figure 1a. The individual output
detailed view is depicted in Figure 1a. The individual output powers of the two outputs are inverse to
powers of the two outputs are inverse to each other, due to the resulting superposition of the two
each other, due to the resulting superposition of the two phase-shifted modes (see Figure 1b). The sum
phase-shifted modes (see Figure 1b). The sum of the two powers is always constant, and equal to the
of the two powers is always constant, and equal to the input power, assuming an ideal transmission
input power, assuming an ideal transmission with no losses.
with no losses.

Figure 1. Buildup
Figure 1. Buildup ofof the
the DMI
DMI with
with two
two differential
differential DMI
DMI outputs
outputs (type
(type 7).
7). The
The schematic
schematic view
view (a)
(a) and
and
the
the field
field profiles
profiles (b)
(b) of
of the
the combiner
combinersection
sectionare
areshown.
shown.

The type
The type7 7DMIDMI hashas a slightly
a slightly higher
higher excessexcess
loss ofloss of theconversion
the mode mode conversion in the compared
in the combiner combiner
compared
to to types
types 4 and 6, but4 with
and 6,the
buttwowith the two differential
differential outputs
outputs it is it is
possible topossible
determineto determine the phase
the phase condition
condition of the interference in a range of π [4] with only one measurement at a single
of the interference in a range of π [4] with only one measurement at a single wavelength, rather than wavelength,
rather thanthe
evaluating evaluating the transmission
transmission spectrum in spectrum in a broad
a broad range. Thus,range. Thus,laser
no tunable no tunable
sourceslaser sources are
are required to
requiredthe
operate to DMI
operatein athe DMIsystem.
sensor in a sensor system.

3. Operating Principle of the Dual-Mode

3. Dual-Mode Interferometer
Interferometer with
with Differential
Differential Outputs
Outputs
For describing
For describingthe theresulting
resulting sensitivity
sensitivity of aoftype
a type
7 DMI,7 DMI, a distinction
a distinction can becan
madebebetween
made between
surface
surface sensitivity
sensitivity and homogenous
and homogenous bulk sensitivity
bulk sensitivity as in [5]. as
Theinlatter
[5]. The
canlatter can betorelated
be related to homogenous
homogenous cladding
claddingand
changes, changes,
is used and
in is used
this workin to
this work to
explain theexplain the principle.
operating operating Theprinciple. The homogenous
homogenous bulk
bulk sensitivity
Ssensitivity
bulk is S
definedbulk is
as defined
the as
change the
of change
the phaseof the phase
difference ∆φ
difference
over theΔφ over
change the
of change
the of the
refractive refractive
index of the
index of the
cladding, andcladding, and can be
can be expressed byexpressed by

 ∂(=∆φ)  = 2πLsens  ∂(nm1 − nm2= )  2πLηsens ,

   
(1)
Sbulk =   = = ηbulk , (1)
∂nc  λ0  ∂nc  λ0
where is the change of the refractive index in the sensing region with the length Lsens. The
effective
where ∂ncrefractive
is the changeindices of the
of the fundamental
refractive index inand thesecond-order
sensing region modes are length
with the nm1 andLsens
nm2., The
respectively.
effective
The phaseindices
refractive difference
of thebetween the two
fundamental andmodes changesmodes
second-order with theare refractive
nm1 and nm2index of the cladding
, respectively. due
The phase
to differentbetween
difference modal sensitivities
the two modes resulting
changesfrom their
with themodal profiles.
refractive indexThis fundamental
of the cladding due behavior is also
to different
the same
modal for other DMIs
sensitivities [5], and
resulting fromthe sensitivity
their of a waveguide
modal profiles. cross-section
This fundamental can be characterized
behavior is also the same by
the other
for intrinsic
DMIsbulk[5],sensitivity to changesof
and the sensitivity inathe cladding cross-section
waveguide layer ηbulk, which can becancharacterized
be larger than by50%.
the
However,
intrinsic depending
bulk sensitivity ontothe requested
changes in theΔφ, the optical
cladding layersignals in thecan
ηbulk , which dual-mode
be largerwaveguide of a type
than 50%. However,
7 DMI can be
depending on coupled
the requested ∆φ, 1the
to output oroptical
outputsignals
2 (see Figure 1b).
in the dual-mode waveguide of a type 7 DMI can
When to
be coupled only one output
output port (as
1 or output in DMIs
2 (see Figureof1b).
type 1 or types 3–6) is available, the determination of
phase difference
When only oneΔφoutput
must be done
port (asbyin evaluating
DMIs of type the1transmission
or types 3–6)spectrum
is available,in athe
certain wavelength
determination of
phase difference ∆φ must be done by evaluating the transmission spectrum in a certain wavelength
range. This is illustrated in Figure 2 with an exemplary detection of Pb(II) ions from Pb(NO3 )2 in water
(see Appendix A). The transmission minimum moves towards a longer wavelength due to the phase
Biosensors 2017,
Biosensors 2017, 7,
7, 37
37 44 of
of 10
10

Biosensors 2017, 7, 37 4 of 10
range. This
range. This is
is illustrated
illustrated inin Figure
Figure 22 with
with anan exemplary
exemplary detection
detection ofof Pb(II)
Pb(II) ions
ions from
from Pb(NO
Pb(NO33))22 in
in water
water
(see Appendix A). The transmission minimum moves towards a longer wavelength
(see Appendix A). The transmission minimum moves towards a longer wavelength due to the phase due to the phase
shift induced
shift induced by by the
the increased
increased concentration
concentration of of Pb(II)
Pb(II) ions. At the
ions. At the same
same time,
time, the
the spectrum
spectrum is is shifted
shifted
upwards to higher transmitted power levels, i.e., the total losses also depend on the
upwards to higher transmitted power levels, i.e., the total losses also depend on the concentration of concentration of
Pb(II) ions. With a measurement of the transmitted power at only a single wavelength
Pb(II) ions. With a measurement of the transmitted power at only a single wavelength there is neither there is neither
the possibility
the possibility to distinguish between
to distinguish between phase
phase shift induced change
shift induced change in transmission and
in transmission loss dependent
and loss dependent
change in transmission nor the possibility of
of perceiving
perceiving input
input power
power
change in transmission nor the possibility of perceiving input power fluctuations. fluctuations.
fluctuations.

Figure 2.
Figure 2. Measurements
Measurements
Measurements ofof
of aaa type
type
type 555 DMI
DMI with
DMI with two
with two different
two different concentrations
different concentrations of Pb(II)
concentrations of Pb(II) ions
Pb(II) ions in
in water.
in water. The
water. The
The
optical input
optical input power
power is
is 00 dBm.
dBm.

The evaluation
The evaluation of
of the
the two
two differential
differential outputs of a type 77 DMI
DMI results in
in a measured transmission
transmission
two differential outputs of a type
type 7 DMI results
results in aa measured
measured transmission
spectrum as
spectrum as shown
shown in
in Figure
Figure 3a.
3a. The
The antiphase
antiphase transmission
transmission characteristic
characteristic is
is clearly
clearly visible,
visible, and
and the
the
sum of the two output powers (dotted line) is constant.
sum of the two output powers (dotted line) is constant.

Figure 3.
Figure 3. (a)
(a) Transmission
Transmission spectra of
of two differential DMI outputs (type
(type 7)
7) and
and the
the corresponding
(a) Transmission spectra
spectra of two differential DMI outputs
outputs corresponding
sum. The
sum. The dual-mode
dual-mode waveguide
waveguide length
length and
and width
width are 200 μm and 625 nm, respectively; (b) Resulting
The dual-mode waveguide width are 200 µm and 625 nm, respectively; (b) Resulting
are 200 μm and 625 nm, respectively; Resulting
phase relation.
phase relation.

Assuming an ideal
Assuming ideal modulation transfer
transfer curve with ∞, which is
with ER == ∞, is frequently used
used for MZIs,
MZIs,
Assuming anan ideal modulation
modulation transfer curve
curve with ER
ER = ∞, which
which is frequently
frequently used for
for MZIs,
the
the phase difference between the two modes is calculated from the two output powers P1 and P2 by
the phase
phase difference between the
difference between the two
two modes
modes is
is calculated
calculated from
from the
the two
two output
output powers
powers PP11 and
and P
P22 by
by
Δϕ ≈
Δϕ ≈ 22 cos
coss .. (2)
(2)
−1 P1
∆φ ≈ 2 cos . (2)
The result
result calculated
calculated from
from the
the shown
shown formula
formula is P1 + P2 in Figure 3b. The calculation is done
is depicted
depicted
The in Figure 3b. The calculation is done
over an
over an interval
interval from
from minimum
minimum P at 1530
P11 at 1530 nm
nm to
to maximum
maximum P P11 at
at 1541
1541 nm,
nm, where
where thethe phase
phase relation
relation
The
∆φ ranges result
ranges from calculated
from 180°
180° to
to 0°. from
0°. The the shown
The well-known formula
well-known problem is
problem of depicted
of estimatingin Figure
estimating aa phase 3b. The
phase shift calculation
shift larger
larger than
than ππiscan
done
can be
∆φ
over an interval from minimum P1a at maximum P1 versus
1530 nm to measurement at 1541time
nm, or
where theplacing
phase relationbe
circumvented
circumvented [13]
[13] by either
by◦ to
either using
using transmission
a transmission measurement versus time or else
else by
by placing several
∆φ ranges from
interferometers 180
with 0◦ . The
different well-known
sensing problem
lengths of estimating
in parallel,
parallel, which cancanabe
phase
be shift
done very larger
easilythan πseveral
with can be
DMIs.
interferometers with different sensing lengths in which done very
circumvented [13] by either using a transmission measurement versus time or else by placing severaleasily with DMIs.
interferometers
4. Performance with
Performance of different
of Type
Type sensing lengths in parallel, which can be done very easily with DMIs.
4. 77 Dual-Mode
Dual-Mode Interferometers
Interferometers
Apart from
fromof
4. Performance
Apart the sensitivity
Type
the of aa DMI,
7 Dual-Mode
sensitivity of DMI, there are
are several
several important
Interferometers
there important operational
operational parameters,
parameters, such
such as
as
IL and
IL and ER. To
ER. To determine
determine these parameters,
these parameters, type 7
typeare DMIs
7 DMIs were fabricated in the 250 nm SOI platform
Apart from the sensitivity of a DMI, there severalwere fabricated
important in the 250
operational nm SOI platform
parameters, such as
IL and ER. To determine these parameters, type 7 DMIs were fabricated in the 250 nm SOI platform
Biosensors 2017, 7, 37 5 of 10
Biosensors 2017, 7, 37 5 of 10
Biosensors 2017, 7, 37 5 of 10

at IMS
IMS CHIPS Stuttgart.
Stuttgart. The transmission
transmission spectra ofof the two
two outputs of
of a DMI with
with a 200 µm
μm long
at IMSCHIPS
at CHIPS Stuttgart.TheThe transmissionspectra
spectra ofthe
the twooutputs
outputs ofaaDMI
DMI withaa200
200 μmlong
long
dual-mode waveguide are depicted in Figure 4.
dual-modewaveguide
dual-mode waveguideare aredepicted
depictedininFigure
Figure4.4.

Figure 4. Transmission spectra of two differential DMI outputs (type 7) with straight output tapers
Figure4.4.Transmission
Figure Transmission spectra
spectra of of
twotwo differential
differential DMIDMI outputs
outputs (type(type 7) with
7) with straight
straight outputoutput
taperstapers
and
and the transmission spectrum of the reference. The dual-mode waveguide length and width are
and
the the transmission
transmission spectrum
spectrum of the reference.
of the reference. The dual-mode
The dual-mode waveguidewaveguide length
length and and
width arewidth
200 µmare
200 μm and 625 nm, respectively. The transmission of a reference structure is shown, which is used
200625
and μmnm,
andrespectively.
625 nm, respectively. The transmission
The transmission of a reference
of a reference structurestructure
is shown,iswhich
shown,is which is used
used for the
for the determination of the IL.
for the determination
determination of the IL.of the IL.

The resulting IL of 1.9 dB at a wavelength of 1538 nm consists of two times the excess loss of the
Theresulting
The resultingILILofof1.9
1.9dBdBatataawavelength
wavelengthofof1538 1538nmnmconsists
consistsofoftwo
twotimes
timesthetheexcess
excessloss
lossofofthe
the
mode converters and the waveguide and taper losses. Assuming small waveguide and taper losses,
mode converters and the waveguide and taper losses. Assuming small waveguide
mode converters and the waveguide and taper losses. Assuming small waveguide and taper losses, and taper losses,
the measurement is in good accordance to the simulation of the excess loss of mode conversion in
themeasurement
the measurementis is in in good
good accordance
accordance to the
to the simulation
simulation of excess
of the the excess loss
loss of of mode
mode conversion
conversion in
in [12],
[12], which results in 0.9 dB at 1550 nm.
[12], which results in 0.9 dB
which results in 0.9 dB at 1550 nm. at 1550 nm.
In this measurement, the ER of the DMI is limited mainly due to the dynamic range of the power
InInthis
thismeasurement,
measurement,the theER
ERofofthe
theDMI
DMIisislimited
limitedmainly
mainlydueduetotothe
thedynamic
dynamicrangerangeofofthe
thepower
power
meter (Agilent 81625A). In general, the ER of the DMI sensor depends on the achievable ratio of mode
meter(Agilent
meter (Agilent81625A).
81625A).InIngeneral,
general,the
theERERofofthe
theDMI
DMIsensor
sensordepends
dependson onthe
theachievable
achievableratio
ratioofofmode
mode
powers P1/P2 at the combiner [12]. Due to different mode profiles in the sensing region, usually the
powers P 1/P2 at the combiner [12]. Due to different mode profiles in the sensing region, usually the
powers P1 /P2 at the combiner [12]. Due to different mode profiles in the sensing region, usually the
mode losses differ, which consequently decreases the ER (see Figure 5a). This effect is clearly visible
modelosses
mode lossesdiffer,
differ, which
which consequently
consequently decreases
decreases thethe
ERER (see
(see Figure
Figure 5a).5a).
ThisThis effect
effect is clearly
is clearly visible
visible in
in the measured transmission spectra of two MZIs with different arm length differences (see Figure
in the measured transmission spectra of two MZIs with different arm length
the measured transmission spectra of two MZIs with different arm length differences (see Figure 5b).differences (see Figure
5b). The additional waveguide loss in one MZI arm decreases the ER.
5b).additional
The The additional waveguide
waveguide loss inloss
oneinMZI
onearmMZIdecreases
arm decreases
the ER.the ER.

Figure 5. (a) Extinction ratio versus signal loss difference ∆α for a 1 mm long DMI; (b) Transmission
Figure 5. (a) Extinction ratio versus signal loss difference Δα for a 1 mm long DMI; (b) Transmission
Figureof
spectra 5. two
(a) Extinction
exemplaryratio
MZIs versus signal loss difference
with corresponding Δα for
arm length a 1 mm long
differences. TheDMI; (b) Transmission
decreased extinction
spectra of two exemplary MZIs with corresponding arm length differences. The decreased extinction
spectra
ratio can of
betwousedexemplary MZIs with
for the calculation of corresponding
the waveguidearmloss.length differences. The decreased extinction
ratio can be used for the calculation of the waveguide loss.
ratio can be used for the calculation of the waveguide loss.
Therefore,
Therefore, deeper
deeper insight
insight in in the
the mode
mode loss
loss mechanism
mechanism and and the
the estimation
estimation of of the
the order
order ofof
Therefore, deeper insight in the mode loss mechanism and the estimation of the order of
magnitude
magnitude of especiallythe
of especially thehigher
higherorder
ordermode
mode losses
losses helps
helps in designing
in designing an attractive
an attractive DMIDMIwithwith
high
magnitude of especially the higher order mode losses helps in designing an attractive DMI with high
high ER low
ER and and IL.
lowSeveral
IL. Several
works works
treattreat the fundamental
the fundamental mode mode
losseslosses
in Siinwaveguides,
Si waveguides, but there
but there is
is less
ER and low IL. Several works treat the fundamental mode losses in Si waveguides, but there is less
less
work work
on theon the experimental
experimental determination
determination of the
of the higher-order
higher-order mode mode
loss,loss, which
which is also
is also important
important for
work on the experimental determination of the higher-order mode loss, which is also important for
for other systems, such as multimodal on-chip data
other systems, such as multimodal on-chip data links [14]. links [14].
other systems, such as multimodal on-chip data links [14].
The loss ααm2 of
The loss of the
the second-order
second-order mode
mode in in aa DMI
DMI isis composed
composed of of common
common scattering
scattering andand
The loss α of the second-order mode in a DMI is composed of common scattering and
absorption losses ααnorm , , and
absorption losses and the
the power
power fraction
fraction leaking
leaking to to the
the fundamental
fundamental mode,mode, expressed
expressed by by
absorption losses α , and the power fraction leaking to the fundamental mode, expressed by
ααm2→ . To estimate the higher-order mode loss there are several options. The
m1 . To estimate the higher-order mode loss there are several options. The pure excitation of thepure excitation of
α →→ . To estimate the higher-order mode loss there are several options. The pure excitation of the
the interested higher-order mode in combination with the commonly used
interested higher-order mode in combination with the commonly used cut-back method is an easy- cut-back method is an
interested higher-order mode in combination with the commonly used cut-back method is an easy-
easy-to-understand
to-understand way.way. For Forthisthis purpose,
purpose, a DMI
a DMI with
with twotwo single-modeinputs
single-mode inputscancanbebe used,
used, acting
acting asas aa
to-understand way. For this purpose, a DMI with two single-mode inputs can be used, acting as a
kind of 2 × 2 multimode interferometer (MMI). Like in a tunable power splitter, the phase condition
kind of 2 × 2 multimode interferometer (MMI). Like in a tunable power splitter, the phase condition
Biosensors 2017, 7, 37 6 of 10

Biosensors
kind of 2 2017, 7, 37
× 2 multimode interferometer (MMI). Like in a tunable power splitter, the phase condition 6 of 10

of the input signals controls the power ratio. With a suitable phase relation at the input, almost pure
of the input
excitation signals
of the controls
higher the power
order mode ratio.
can be With a suitable phase relation at the input, almost pure
achieved.
excitation
In this of the higher
work, we useorder
DMIs mode
withcan be achieved.
different strip waveguide lengths L and DMI mode splitters,
In this work, we use DMIs with different
which are also used as mode combiners, with a theoreticalstrip waveguide lengths
input power L and
ratio of 1.DMI mode is
If αm2 splitters,
small
→m1
which are also used as mode combiners, with a theoretical input power ratio of 1. If α → is small
enough, the measured ER can be used to calculate the difference between the higher-order mode loss
enough, the measured ER can be used to calculate the difference between the higher-order mode loss
and the fundamental mode loss ∆α by
and the fundamental mode loss Δα by
√ 
20 log10 √ ER√ +1

|∆α| =| |α|
αm2=− = |=
|α αm1−| α √ −1
ER
dB.
dB. (3)
(3)
L
Thiscalculation
This calculation isis done
done for
forDMIs
DMIswith
withdifferent
differentdual-mode
dual-mode waveguide
waveguide widths
widthsW, and ERs ERs
W, and were
extracted using the minima and maxima of the optical transmission spectrum close
were extracted using the minima and maxima of the optical transmission spectrum close to 1550 nm. to 1550 nm. The
resulting loss differences are depicted over wavelength in Figure 6.
The resulting loss differences are depicted over wavelength in Figure 6.

Figure6.6.Loss
Figure Lossdifference
differenceversus
versuswavelength
wavelengthfor
fortwo
twofabricated
fabricatedDMIs
DMIswith
witha awaveguide
waveguidewidth
widthofof
420
420nm
nmand
and575
575nm.
nm.

Notethat
Note thatthe
thewaveguide
waveguidethickness
thicknessisisalways
always250250nm
nmandandthe
thewaveguides
waveguidesare arecladded
claddedby bya a1 1µmμm
thickSiO
thick SiO layer.Utilizing
2 2layer. Utilizingmeasurements
measurementsininthe thesame
sametechnology
technologyresulting
resultingfrom
fromthethecut-back
cut-backmethod
method
for a width W of 400 nm and
for a width W of 400 nm and for the for the fundamental TE mode results in α
mode results in αm1 equal to 3.3 dB/cmat
is equal to 3.3 dB/cm
at1550
1550nm.
nm.AsAs a consequence,
a consequence, α αm2forfor
WW = 420 nmnm
= 420 cancan
be estimated
be estimated to be
tosmaller
be smallerthanthan14 dB/cm
14 dB/cmin the
in the analyzed wavelength range. Taking the smallest loss difference into account and assuming>
analyzed wavelength range. Taking the smallest loss difference into account and assuming α
α α results in α ≈ 4.5 dB/cm. The variation of the additional loss over wavelength may be closely
m2 > αm1 results in αm2 ≈ 4.5 dB/cm. The variation of the additional loss over wavelength may be
associated
closely with the
associated withinfluence of stitching
the influence errorserrors
of stitching caused by the
caused bye-beam lithography
the e-beam lithography fabrication
fabricationstep.
In addition, the wavelength dependency of the mode-to-mode
step. In addition, the wavelength dependency of the mode-to-mode coupling coupling strength should bebe
should
investigated in further works. In addition to the IL and ER, the thermal dependency
investigated in further works. In addition to the IL and ER, the thermal dependency is an important is an important
characteristicfor
characteristic forsensor
sensorapplications.
applications.ForFortype
type77DMIs,
DMIs,thethethermal
thermaldependency
dependencyisismainly mainlycaused
causedby by
the phase relation of both modes in the dual-mode waveguide, and can be minimized
the phase relation of both modes in the dual-mode waveguide, and can be minimized with a proper with a proper
waveguidegeometry
waveguide geometryasasdiscussed
discussedinin[12].
[12].

5.5.Comparison
Comparisonbetween
betweenDMI
DMIand
andMZI
MZI
ToTocompare
comparethe theperformance
performanceofofthe
theDMI
DMIwith
withthe
theperformance
performanceofofMZIs,
MZIs,both
bothinterferometer
interferometer
types
typeswere
werefabricated
fabricatedusing
usingaa250
250nm
nmSOI
SOItechnology
technologywith
withpartially
partiallyuncovered
uncoveredSiSiwaveguide
waveguidearms.
arms.
The devices are shown in Figure
The devices are shown in Figure 7. 7.

Figure 7. 7.
Figure Schematic view
Schematic of of
view fabricated devices
fabricated forfor
devices thethe
comparison of of
comparison thethe
DMIs with
DMIs thethe
with MZIs.
MZIs.

The in-coupling and out-coupling of the optical signals to single-mode fibers (SMF-28) is done
by using identical aperiodic grating couplers. With the help of two different sensor region lengths
Lsens for each interferometer type, a determination of the additional loss caused by only the sensing
Biosensors 2017, 7, 37 7 of 10

The in-coupling and out-coupling of the optical signals to single-mode fibers (SMF-28) is done
by using identical aperiodic grating couplers. With the help of two different sensor region lengths
Lsens for each interferometer type, a determination of the additional loss caused by only the sensing
region becomes possible. An increased sensor region length leads to a larger sensitivity of the device,
although usually there is a compromise between loss and sensitivity of a designed interferometer.
The decrease in loss given in dB and gain in sensitivity increase linearly with the sensor region length.
Therefore, the product of the sensitivity and the reciprocal additional loss is calculated to make the
comparison as fair as possible, and length-independent. The resulting figure of merit (FOM) is the
bulk sensitivity per loss, which can be expressed by

2π∆Lsens ηbulk
Sbulk /α = (4)
λ0 ∆IL

where ∆Lsens is the difference of the sensor region waveguide length and ∆IL is the difference in IL of
the two interferometers. This parameter and the other results are listed in Table 2 for DMIs and MZIs.

Table 2. Resulting performance values for DMIs and MZIs close to 1550 nm. The intrinsic bulk
sensitivity is simulated with the FIMMWAVE finite difference waveguide mode solver for a refractive
index change of 0.01 and for the DMI as difference of the two intrinsic mode sensitivities.

Intrinsic Bulk Sensitivity

Device Waveguide Width Sensor Region Length Measured IL Measured ER
Bulk Sensitivity per Loss
500 µm 1 dB >30 dB
MZI 250 nm
5000 µm 11 dB
79%
>25 dB 1441 dB−1

500 µm 2.5 dB ≈20 dB

DMI 575 nm
6400 µm 5.2 dB
44%
≈10 dB 3237 dB−1

Even if the IL for short MZIs is better than for short DMIs, this situation is reversed for longer
devices, due to the different additional sensor region losses. This fact is reflected by the much higher
sensitivity per loss FOM of the DMI. Following the calculated values in Table 2, the fabricated long
DMIs are predominant, but the MZIs achieve a higher ER. Higher ER and lower losses are also achieved
by the DMI with 420 nm waveguide width, but the influence of temperature on the phase difference is
increased. Please note that these ILs depend strongly on the technology used, and this comparison is
not valid for all technologies.

6. Discussion and Conclusions

In this work, the operating principles of silicon integrated dual-mode interferometers for a
stable single-wavelength operation are presented. The influence of an undetermined or fluctuating
optical input power, as well as varying absorption in the sensor region, can be compensated by
exploiting power measurements at a single-wavelength using power sum and ratio of two differential
outputs. Using the latest designs, detection with high extinction ratio and low insertion loss is possible.
The phase condition ambiguity can be circumvented by two or more interferometers with different
sensor region lengths in parallel. In this work, the operating principle is discussed in detail, and the
detection of Pb(II) ions in water is shown, which demonstrates a sensor application of the DMI. Further,
procedures are presented to estimate the losses of higher-order modes in optical waveguides. In the
technology utilized here the bulk sensitivity per loss given in dB−1 in the sensor region of dual-mode
interferometers is more than twice as high as in MZIs. This makes the DMI attractive especially when
very long sensing waveguides are required.

Acknowledgments: We thank T. Föhn for the fruitful discussions of DMI designs and simulations. Further, we are
thankful for the support of M. Félix Rosa, V. Albernaz, M. Bach and M. Kaschel.
Author Contributions: All authors contributed extensively to the study presented in this manuscript.
Niklas Hoppe is a student with a focus on dual-mode sensors, and wrote the paper. He has performed the
investigations of the mode losses and the DMI designs. Pascal Scheck, Philipp Diersing and Lotte Rathgeber
Biosensors 2017, 7, 37 8 of 10

Biosensors
performed 2017,
the7, 37 8 of 10
measurements used in chapters 4 and 5. Philipp Diersing was also engaged in the theoretical
evaluation of type 7 DMIs. Rami Sweidan performed the experiments described in Figure 2, under supervision and
with the help
Wolfgang of Pascal
Vogel Scheck Berroth
and Manfred and Alexander Southan.
participated Benjamin
in the Riegger
conceptual fabricated
design of this the used
paper, particles.
and did theWolfgang
internal
Vogel and Manfred Berroth participated in the conceptual design of this paper, and did the internal proofreading.
proofreading.
Conflicts of Interest: The authors declare no conflict of interest.
Conflicts of Interest: The authors declare no conflict of interest.

Appendix
Appendix A
P(VPS-co-EBA) nanoparticles for coating of the sensor surface were prepared by
P(VPS-co-EBA) nanoparticles for coating of the sensor surface were prepared by miniemulsion
miniemulsion polymerization using the monomer vinylphosphonic acid (VPS) and the cross-linker
polymerization using the monomer vinylphosphonic acid (VPS) and the cross-linker N,N′-
N,N 0 -ethylenebisacrylamide (EBA), according to Niedergall et al. [15]. For the coating of the sensor
ethylenebisacrylamide (EBA), according to Niedergall et al. [15]. For the coating of the sensor surface,
surface, the particles were redispersed in water, where the phosphonic acid groups present in the
the particles were redispersed in water, where the phosphonic acid groups present in the particles
particles were partly deprotonated to negatively charged phosphonates. The Si waveguides were
were partly deprotonated to negatively charged phosphonates. The Si waveguides were fabricated
fabricated using 250 nm SOI technology at IMS CHIPS Stuttgart using e-beam lithography. Figure A1
using 250 nm SOI technology at IMS CHIPS Stuttgart using e-beam lithography. Figure A1 illustrates
illustrates the fabrication process for the particle layer. First, the chip of interest was cleaned with
the fabrication process for the particle layer. First, the chip of interest was cleaned with acetone,
acetone, isopropanol and distilled water to remove organic impurities on the surface (Figure A1a,
isopropanol and distilled water to remove organic impurities on the surface (Figure A1a, top left).
top left). Next, the chip surface was activated to increase the concentration of hydroxyl groups.
Next, the chip surface was activated to increase the concentration of hydroxyl groups. This was
This was accomplished by exposing the chip to a mixture of hydrogen peroxide solution (30%) and
accomplished by exposing the chip to a mixture of hydrogen peroxide solution (30%) and ammonium
ammonium hydroxide solution (25%) in a volume ratio of 2:3 (v/v) at 70 ◦ C for 40 min (Figure A1a,
hydroxide solution (25%) in a volume ratio of 2:3 (v/v) at 70 °C for 40 min (Figure A1a, top right).
top right). After that, the chip was submerged overnight in a solution of (3-aminopropyl)triethoxysilane
After that, the chip was submerged overnight in a solution of (3-aminopropyl)triethoxysilane in
in ethanol, followed by a washing step in ethanol and drying (Figure A1a, bottom left). Thus, in contact
ethanol, followed by a washing step in ethanol and drying (Figure A1a, bottom left). Thus, in contact
with water, positively charged ammonium groups were present on the surface. Finally, the chip was
with water, positively charged ammonium groups were present on the surface. Finally, the chip was
immersed in the suspension containing the P(VPS-co-EBA) nanoparticles, and the particles adhered
immersed in the suspension containing the P(VPS-co-EBA) nanoparticles, and the particles adhered
to the surface, probably due to electrostatic interactions of the oppositely charged ammonium and
to the surface, probably due to electrostatic interactions of the oppositely charged ammonium and
phosphonate groups. Thus, the sensor surface was coated with the particles via dip coating (Figure A1a,
phosphonate groups. Thus, the sensor surface was coated with the particles via dip coating (Figure
bottom right), and was then ready for measurements (Figure A1b).
A1a, bottom right), and was then ready for measurements (Figure A1b).

(a) (b)

Figure
FigureA1.
A1. Fabrication
Fabricationof
ofaa P(VPS-co-EBA)
P(VPS-co-EBA)particle
particle covered
covered dual-mode
dual-mode waveguide.
waveguide. (a)
(a)Fabrication
Fabrication
steps; (b) Resulting layer stack.
steps; (b) Resulting layer stack.

The used particles already showed an enhanced Pb(II) ion adsorption property in other works
The used particles already showed an enhanced Pb(II) ion adsorption property in other works [15],
[15], which can be ascribed to the generally strong interaction of Pb(II) ions with phosphonate and
which can be ascribed to the generally strong interaction of Pb(II) ions with phosphonate and phosphate
phosphate groups as documented by the very low solubility of lead(II) phosphate in water. In our
groups as documented by the very low solubility of lead(II) phosphate in water. In our studies, the
studies, the sensor response and resulting phase shift, which is depicted in Figure A2, is enhanced in
sensor response and resulting phase shift, which is depicted in Figure A2, is enhanced in a specific
a specific working range by the particle layer compared to the same DMI without functional cladding.
working range by the particle layer compared to the same DMI without functional cladding.
Despite this enhancement, the repetition of the measurement with the same particles is
Despite this enhancement, the repetition of the measurement with the same particles is challenging,
challenging, and its purification will be the subject of further research.
and its purification will be the subject of further research.
Biosensors 2017, 7, 37 9 of 10
Biosensors 2017, 7, 37 9 of 10

Figure A2. Measurement of the resulting phase condition with fixed fibers resulting from a
Figure A2. Measurement of the resulting phase condition with fixed fibers resulting from a particle-
particle-covered DMI and a not-covered DMI by varying concentration of Pb(II) ions in water.
covered DMI and a not-covered DMI by varying concentration of Pb(II) ions in water. The evaluation
The evaluation of the phase condition is done using the minima in the transmission spectra close
of the phase condition is done using the minima in the transmission spectra close to 1540 nm. The
to 1540 nm. The temperature influence is reduced by recording the solution temperature and using
temperature influence is reduced by recording the solution temperature and using a temperature
a temperature coefficient of 0.1249 rad/K for the resulting corrected phase shift. The temperature
coefficient of 0.1249 rad/K for the resulting corrected phase shift. The temperature coefficient can be
coefficient can be decreased with an optimized waveguide geometry, as shown in [12].
decreased with an optimized waveguide geometry, as shown in [12].

References
References
1. Dhakal,
Dhakal, A.;A.;Wuytens,
Wuytens, P.; P.;
Peyskens,
Peyskens,F.; Jans, K.; LeK.;
F.; Jans, Thomas, N.; Baets,
Le Thomas, N.;R.Baets,
Nanophotonic WaveguideWaveguide
R. Nanophotonic Enhanced
Raman
Enhanced Spectroscopy of Biological
Raman Spectroscopy of Submonolayers. ACS Photonics
Biological Submonolayers. ACS2016, 3, 2141–2149.
Photonics [CrossRef]
2016, 3, 2141−2149.
2. Makarona, E.; Petrou, P.; P.; Kakabakos,
Kakabakos, S.; S.; Misiakos,
Misiakos, K.;K.; Raptis,
Raptis, I.I. Point-of-Need
Point-of-Need bioanalytics
bioanalytics based
based on planar
interferometry. Biotechnol.
optical interferometry. Biotechnol. Adv. 2016, 34,
Adv. 2016, 34, 209–233.
209–233. [CrossRef] [PubMed]
3. Duval, D.; González-Guerrero, A.; Dante, Dante, S.; Osmond,
Osmond, J.; Monge, Monge, R.; Fernández,
Fernández, L.; L.; Zinoviev,
Zinoviev, K.;
Domínguez, C.; Lechuga, M. Nanophotonic lab-on-a-chip platforms including novel bimodal interferometers,
microfluidics
microfluidics and and grating couplers. Lab Chip
grating couplers. 2012, 12,
Chip 2012, 12, 1987–1994.
1987–1994. [CrossRef] [PubMed]
4. Hoppe,
Hoppe, N.;N.; Diersing,
Diersing, P.;P.;Föhn,
Föhn,T.;T.;Kaschel,
Kaschel,M.; M.;Polder,
Polder,T.;T.;
Vogel,
Vogel, W.;W.;
Rathgeber,
Rathgeber, L.; L.;
Félix Rosa,
Félix Rosa,M.;M.;
Berroth, M.
Berroth,
Integrated
M. IntegratedDual-Mode
Dual-ModeInterferometer with Differential
Interferometer Single-Mode
with Differential Outputs.Outputs.
Single-Mode In Proceedings of the European
In Proceedings of the
Conference on Integrated
European Conference Optics (ECIO),
on Integrated OpticsEindhoven, The Netherlands,
(ECIO), Eindhoven, 3–5 April 2017.
The Netherlands, 3–5 April 2017.
5.
5. Ramirez,
Ramirez, J.; Lechuga, L.; Gabrielli, L.; Hernandez-Figueroa, H. Study of a low-cost trimodal
J.; Lechuga, L.; Gabrielli, L.; Hernandez-Figueroa, H. Study of a low-cost trimodal polymer
polymer
waveguide for interferometric optical biosensors. Opt. Express 2015,
waveguide for interferometric optical biosensors. Opt. Express 2015, 23, 11985–11994. 23, 11985–11994. [CrossRef] [PubMed]
6.
6. Chao,
Chao, T.;
T.; Davis,
Davis, S.;S.; Rommel,
Rommel, S.; S.; Farca,
Farca, G.;G.; Luey,
Luey, B.;B.; Martin,
Martin, A.; A.; Anderson,
Anderson, M. M. Compact
Compact LiquidLiquid Crystal
Crystal
Waveguide
Waveguide Based Based Fourier
FourierTransform
TransformSpectrometer
Spectrometer forfor In-Situ
In-Situ andand Remote
Remote GasGas and Chemical
and Chemical Sensing.
Sensing. Proc.
Proc. SPIE 2008, 6977, 69770P. [CrossRef]
SPIE 2008, 6977, 69770P, doi:10.1117/12.838260.
7.
7. Levy,
Levy,R.;R.;Ruschin,
Ruschin,S.S.Design
Design ofofa Single-Channel
a Single-Channel Modal
Modal Interferometer
Interferometer Waveguide
Waveguide Sensor. IEEE IEEE
Sensor. Sens. Sens.
J. 2009,
J.
9, 146–153.
2009, [CrossRef]
9, 146–153.
8.
8. Zinoviev,
Zinoviev, K.; K.; González-Guerrero,
González-Guerrero, A.; A.; Domínguez,
Domínguez, C.; Lechuga, M.
C.; Lechuga, M. Integrated
Integrated Bimodal
Bimodal Waveguide
Waveguide
Interferometric Biosensor for Label-Free Analysis. J. Light. Technol. 2011,
Interferometric Biosensor for Label-Free Analysis. J. Light. Technol. 2011, 29, 1926–1930. 29, 1926–1930. [CrossRef]
9.
9. Liu,
Liu, Q.;
Q.; Kim,
Kim, K.;K.; Gu,
Gu, Z.;
Z.; Kee,
Kee, J.;
J.; Park,
Park, M.M. Single-channel
Single-channel Mach-Zehnder
Mach-Zehnder interferometric
interferometric biochemical
biochemical sensor
sensor
based on two-lateral-mode spiral waveguide. Opt. Express 2014,
based on two-lateral-mode spiral waveguide. Opt. Express 2014, 22, 27910–27920. 22, 27910–27920. [CrossRef] [PubMed]
10.
10. Bruck,
Bruck, R.;
R.; Hainberger,
Hainberger, R. R. Sensitivity
Sensitivity and and design
design of of grating-assisted
grating-assisted bimodal
bimodal interferometers
interferometers for for integrated
integrated
optical biosensing. Opt. Express 2014, 22,
optical biosensing. Opt. Express 2014, 22, 32344–32352.32344–32352. [CrossRef] [PubMed]
11.
11. Hoppe,
Hoppe, N.;N.; Föhn,
Föhn, T.;
T.; Félix
Félix Rosa,
Rosa, M.;
M.P.;P.;Vogel,
Vogel,W.;W.;Sfar
SfarZaoui,
Zaoui, W.;W.; Kaschel,
Kaschel, M.; M.; Butschke,
Butschke, J.; Letzkus, F.;
J.; Letzkus, F.;
Berroth, M. Integrated Dual-Mode Waveguide Interferometer.
Berroth, M. Integrated Dual-Mode Waveguide Interferometer. NUSOD 2015, 155–156, NUSOD 2015, 155–156. [CrossRef]
12. Hoppe, N.; Föhn, T.; Diersing, P.; Scheck, P.; Vogel, W.; Félix Rosa, M.; Kaschel, M.; Bach, M.; Berroth, M.
doi:10.1109/NUSOD.2015.7292869.
12. Design
Hoppe, N.;an
of Integrated
Föhn, Dual-Mode
T.; Diersing, Interferometer
P.; Scheck, P.; Vogel, W.; onFélix
250 Rosa,
nm Silicon-on-Insulator.
M.; Kaschel, M.; Bach, IEEE J. Sel. Top.
M.; Berroth, M.
Quantum Electron. 2017, 23, 444–451. [CrossRef]
Design of an Integrated Dual-Mode Interferometer on 250 nm Silicon-on-Insulator. IEEE J. Sel. Top.
13. Dante,
Quantum S.;Electron.
Duval, D.; 2017, Sepúlveda,
23, 444–451. B.; González-Guerrero, A.; Sendra, J.; Lechuga, M. All-optical phase
13. modulation for integrated
Dante, S.; Duval, interferometric
D.; Sepúlveda, biosensors. Opt. Express
B.; González-Guerrero, 2012, 20,
A.; Sendra, J.; 7195–7205.
Lechuga, M. [CrossRef] [PubMed]
All-optical phase
modulation for integrated interferometric biosensors. Opt. Express 2012, 20, 7195–7205.
Biosensors 2017, 7, 37 10 of 10

14. Zhang, Z.; Hu, X.; Wang, J. On-chip optical mode exchange using tapered directional coupler. Sci. Rep. 2015,
5, 1–7. [CrossRef] [PubMed]
15. Niedergall, K.; Kopp, D.; Besch, S.; Schiestel, T. Mixed-Matrix Membrane Adsorbers for the Selective Binding
of Metal Ions from Diluted Solutions. Chem. Ing. Tech. 2016, 88, 437–446. [CrossRef]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).