Wafer-scale Ge freestanding membranes for lightweight and flexible optoelectronics

Abstract

Semiconductor-based freestanding membranes (FSM) have recently emerged as a highly promising area of advanced materials research. Their unique properties, such as lightweight and flexibility, make them attractive for a wide range of disruptive device applications. However, the production of high-quality, single-crystalline FSM, especially from elemental materials such as germanium (Ge), remains a significant challenge. In this work, we report on the formation of easily detachable wafer-scale Ge FSM on porous Ge (PGe) substrate. The proposed method relies on low-temperature Ge epitaxy, allowing to preserve the porous structure's integrity during the FSM formation, and easy substrate preparation for multiple reuses. Analysis of the surface morphology as a function of the deposited Ge thickness reveals that the FSM formation occurs in two distinct regimes. During the initial epitaxial regime, the Ge growth is governed by 3D nucleation on the PGe top surface. The nanoscale islands size increase, and consequent coalescence are found to increase the surface roughness up to a critical thickness, allowing full coalescence of islands into a 2D epilayer. The analysis of the membrane's surface morphology for various thicknesses shows continuous improvement, achieving sub-nanometer surface roughness. Moreover, we demonstrate that the FSM formation process is applicable regardless of the PGe porosity and thickness, while offering facile and sustainable substrate reconditioning for multiple FSM generation from the same substrate. Our findings open new opportunities to produce lightweight and flexible, high-performance optoelectronics based on Ge FSM, while ensuring the reduction of both cost and critical materials consumption.

Full text

Wafer-scale Ge freestanding membranes for lightweight and flexible
optoelectronics
Tade
a
s Hanu
s
a
,
b
,
*
, Bouraoui Ilahi
a
,
b
, Alexandre Chapotot
a
,
b
, Hubert Pelletier
a
,
b
,
Jinyoun Cho
c
, Kristof Dessein
c
, Abderraouf Boucherif
a
,
b
a
Institut Interdisciplinaire d’Innovation Technologique (3IT), Universit
e de Sherbrooke, 3000 Boulevard de l’Universit
e, Sherbrooke, J1K 0A5, QC, Canada
b
Laboratoire Nanotechnologies Nanosyst
emes (LN2), CNRS IRL-3463 Institut Interdisciplinaire d’Innovation Technologique (3IT), Universit
e de Sherbrooke,
3000 Boulevard de l’Universit
e, Sherbrooke, J1K 0A5, QC, Canada
c
Umicore Electro-Optic Materials, Watertorenstraat 33, 2250, Olen, Belgium
article info
Article history:
Received 2 March 2023
Received in revised form
18 April 2023
Accepted 22 April 2023
Available online xxx
Keywords:
Germanium
Porous substrate
Freestanding membranes
Epitaxial growth
Layer transfer
Substrate re-use
abstract
Semiconductor-based freestanding membranes (FSM) have recently emerged as a highly promising area
of advanced materials research. Their unique properties, such as lightweight and flexibility, make them
attractive for a wide range of disruptive device applications. However, the production of high-quality,
single-crystalline FSM, especially from elemental materials such as germanium (Ge), remains a signifi-
cant challenge. In this work, we report on the formation of easily detachable wafer-scale Ge FSM on
porous Ge (PGe) substrate. The proposed method relies on low-temperature Ge epitaxy, allowing to
preserve the porous structure's integrity during the FSM formation, and an easy substrate preparation for
multiple reuses. Analysis of the surface morphology as a function of the deposited Ge thickness reveals
that the FSM formation occurs in two distinct regimes. During the initial epitaxial regime, the Ge growth
is governed by 3D nucleation on the PGe top surface. The nanoscale islands size increase, and consequent
coalescence are found to increase the surface roughness up to a critical thickness, allowing full coales-
cence of islands into a 2D epilayer. The analysis of the membrane's surface morphology for various
thicknesses shows continuous improvement, achieving sub nanometer surface roughness. Moreover, we
demonstrate that the FSM formation process is applicable regardless the PGe porosity and thickness,
while offering facile and sustainable substrate reconditioning for multiple FSM generation from the same
substrate. Our findings open new opportunities to produce lightweight and flexible, high-performance
optoelectronics based on Ge FSM, while ensuring reduction of both cost and critical materials
consumption.
©2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
1. Introduction
In recent years, high-quality freestanding membranes (FSM)
made of functional materials have become central to the rapidly
expanding frontiers of nanoscience and technology [1e6]. Group
IV, III-V and III-N semiconducting membranes in particular have
high potential for applications such as stretchable on-skin elec-
tronic [7,8], vertically stacked LEDs [9], and flexible photodetectors
[10]. Indeed, FSM offer an extra degree of freedom for
implementations that cannot be obtained by conventional tech-
niques such as heterointegration of dissimilar materials with high
lattice mismatch in crystalline structures [11]. In addition to being
lightweight and flexible, FSM allow for various materials to be
stacked on top of each other, enabling easy coupling of physical
properties between dissimilar materials [12,13]. Furthermore, the
use of FSM provides significant cost savings for the device pro-
duction, especially for materials with orders of magnitude higher
prices than that of silicon, when compared to bulky wafers. In this
context, germanium (Ge) FSM particularly attract a lot of attention
for their applications in high-performance optoelectronics and
high-speed telecommunication devices such as wave guides
[14,15], THz transmission [16 ], photodetectors [17e19] and lasers
[20,21] as well as for their biocompatibility [22,23]. However, the
fabrication of high-quality Ge FSM is still a challenging task.
*Corresponding author. Institut Interdisciplinaire d’Innovation Technologique
(3IT), Universit
e de Sherbrooke, 3000 Boulevard de l’Universit
e, Sherbrooke, J1K
0A5, QC, Canada.
E-mail addresses: tadeas.hanus@usherbrooke.ca (T. Hanu
s), abderraouf.
boucherif@usherbrooke.ca (A. Boucherif).
Contents lists available at ScienceDirect
Materials Today Advances
journal homepage: www.journals.elsevier.com/materials-today-advances/
https://doi.org/10.1016/j.mtadv.2023.100373
2590-0498/©2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Materials Today Advances 18 (2023) 100373

For instance, remote epitaxy has shown tremendous potential
for fabrication of III-N and III-V semiconductor compounds FSM
[24e26], and for other materials such as complex oxides [27], pe-
rovskites [28], metals [29]. Nevertheless, this technique is based on
ionic interaction between the epilayer and the underling substrate
trough the graphene interface, prohibiting its application to
nonpolar materials such as Ge. To this date, a variety of lift-off
techniques allowing the production of single-crystalline Ge FSM
and the reuse of wafers have already been reported, namely
epitaxial lift-off (ELO) [30,31], mechanical spalling [32], smart cut
method [33], growth on nanopatterned graphene [34],
Germanium-on-nothing (GON) [35,36] or porous lift-off [37e39].
Despite achieving significant advancements, the widespread
adoption of Ge FSM is still hindered by various obstacles including
process complexity, high cost and substrates damage and/or
contamination issues. Epitaxial growth on porous Ge (PGe) sub-
strate has specially demonstrated high potential to produce light-
weight solar cells [36,38]. The device detachment is achieved
through the mechanically weak interface, formed by nano-to
microscale-sized pillars [36,38,39]. After FSM detachment, the
substrate surface contains broken pillars with various sizes whose
reconditioning for multiple reuses requires either conventional
chemical mechanical polishing (CMP) treatment [40] or wet
chemical etching over several microns [41]. Despite the significant
improvement compared to the use of conventional wafers, these
techniques still leave room for further reduction of both cost and Ge
consumption during the reconditioning process.
Indeed, PGe substrates offer a wide range of morphologies and
physical properties [42e45], that can be directly used in the Ge FSM
fabrication by epitaxial growth on an unreconstructed porous
structure. Compared to the homoepitaxial growth on conventional
substrate, where optimum 2D layer-by-layer growth can be easily
reached, several challenges occur for nanostructured substrates. In
fact, nanopores mediated epitaxial material diffusion [46] and
porous reconstruction induced 3D micro-structuring [47] are
among the encountered difficulties. These obstacles need to be
overcome to ensure the epitaxial growth of FSM on PGe substrate.
So far, successful epitaxy on PGe substrates is mainly based on high
temperature annealing steps either before [35e38] or during the
material deposition growth [39], triggering the thermal recon-
struction of the porous layer. For instance, PGe with sponge like
morphology show a strong temperature dependance [48], inducing
the formation of large pillars [39] structures, while losing its
intrinsic PGe properties and complicating the reconditioning pro-
cess [41].
In this work, we demonstrate easily detachable wafer-scale Ge
FSM formation on PGe substrate by low temperature growth. The
method, based on the preservation of the porous structure's
integrity, is applicable regardless the PGe porosity and thickness
while ensuring easy substrate preparation for multiple reuses.
2. Results and discussions
The PGe substrates, used for epitaxial growth, are prepared by
bipolar electrochemical etching (BEE) on P-type Ge substrate, in
HF-based electrolyte. The process, which was reported in our pre-
vious work [49], enables fine tuning of the PGe thickness and
porosity, providing on-demand properties while ensuring a low
surface roughness and a substrate oriented crystalline nature,
making them a viable option for epitaxial growth. The produced
PGe thickness and porosity demonstrate an overall variation
standing below 2% across the wafer. Fig. 1a depicts an optical image
of typical 100 mm Ge substrate, with homogeneous porous layer on
top, produced by BEE. To shed light on the initial stages of the Ge
epitaxy on PGe nanostructures, we first consider 230 nm thick
uniform PGe layers (Fig. 1c) over 100 mm Ge wafer with interme-
diate porosity around 54%, calculated from critical angle, measured
by X-ray reflectivity (XRR) [49](Fig. 1b), and having a surface RMS
(Root mean square) roughness below 2 nm (Fig. S1). All the
investigated samples have been grown at 300

C in Chemical beam
epitaxy reactor (CBE) equipped with solid source Ge. The growth
rate has been maintained constant throughout this study at 0.5
m
m/
h. Additionally, since our objective is toperform epitaxial growth of
Ge on PGe structure, the growth temperature needs to be suffi-
ciently low to avoid PGe reorganization as discussed further below.
To evaluate the stability of the porous structure at the growth
temperature, the PGe substrate has been first in-situ annealed at
300

C for 30 min and characterized by XRR, scanning electron
microscopy (SEM) and atomic force microscopy (AFM). Fig. 1b
shows XRR measurement of the porous Ge layer before and after
annealing, with a negligible shift of the PGe layer critical angle
(
q
PGe
), from 0.414

to 0.410

(less than 1% increase of porosity),
signifying that the porosity remains unchanged. Furthermore,
Fig. 1cef depict cross-sectional and top-view SEM images, showing
identical in-plane and in-depth porous morphologies before and
after the annealing step. The surface RMS roughness does not un-
dergo any major changes either, with only slight increase from
1.1 nm to 1.6 nm (Fig. S1). Accordingly, Ge epitaxial growth can be
performed on PGe substrate while preserving the porous structure
integrity. For growth temperatures of 350

C and above, morpho-
logical transformation of the PGe has been found to occur, making it
unsuitable for the present work.
To study the initial growth stages and understand the Ge
nucleation on PGe substrates, several samples have been prepared
with deposited Ge nominal thicknesses ranging from 5 nm to 1
m
m.
The morphological evolution of the epilayers was systematically
evaluated as a function of the nominal thickness of Ge on the PGe
structure and characterized via SEM and AFM. Typical cross-
sectional and top-view SEM images taken from samples with 5,
30, 60 and 100 nm deposited Ge thickness are shown by Fig. 2a-
d and Fig. 2e-h respectively (Additional data are provided in
Fig. S2). The results reveal that the nucleation occurs on the top
surface of the pore walls, forming nanoscale three-dimensional
(3D) islands (Fig. 2: a and e). The combined low substrate tem-
perature and small pores' size (below 10 nm) is likely to limit the
adatoms diffusion into the porous structure [50] in favor of top
surface nucleation. As more material is added, the 3D island's size
increase, eventually coalescing to form a 2D Ge membrane (Fig. 2:b
and f). Although, the densely packed nucleation enables continuous
Ge membrane formation, the corresponding surface morphology
remains rough as can be seen in Fig. 2 (c and g). The observed
behavior is also confirmed by AFM morphological investigations
(Fig. 3 and S3). Ind eed, the analysis of the RMS roughness indicates
that the roughness rises first, with increasing Ge thickness up to
60 nm and then drops rapidly towards surfaces with sub nano-
meter roughness for membranes thicknesses beyond 750 nm.
Additionally, during the Ge nucleation on PGe structure, the first
islanding regime starting with 3D nucleation, considerably impacts
the surface morphology through islands size increase and conse-
quent seed islands coalescence, leading to the observed RMS
roughness increase. As shown by Fig. S3, the coalescence of the
nanosized islands occurs for deposited Ge thicknesses higher than
10 nm giving rise to the appearance of pits. The islanding phase has
been reported in case of low temperature growth of Ge onporous Si
substrate [51]. However, in the latter case, the lattice mismatched
strain and the differences in the thermal expansion coefficients
have been shown to induce grain boundary formations that pro-
hibits the full islands coalescence into homogeneously dense epi-
layer. Moreover, similar behavior has already been reported at
micrometer scale, where persistent separated microcrystals occur
T. Hanu
s, B. Ilahi, A. Chapotot et al. Materials Today Advances 18 (2023) 100373
2

for Ge growth on Si micropillars [50], while good quality suspended
Ge layer can be achieved by high temperature epitaxy/annealing on
Ge micropillars [35]. Furthermore, the formation of good quality
dense epilayer has also been recently reported by GaN nucleation
on porous GaN buffer on sapphire substrate [52], suggesting that
the homoepitaxial growth is a key for obtention of fully coalesced
Fig. 2. Initial growth stages of Ge on the PGe substrate a-d) cross-sectional and e-h) top view SEM micrograph of Ge layer at 5/30/60/100 nm of grown nominal thickness.
Fig. 1. Evaluation of the PGe layer quality before and after the annealing at 30 0 C during 30 min. (a) An optical image of a typical uniform PGe layer on 100 mm wafer. (b) XRR
measures of PGe layer before and after annealing at 300 C, the critical angles of both PGe layer(
q
PGe
) and Ge substrate(
q
Ge
) are indicated by dotted lines. (c)/(d) and (e)/(f) depict top
view and cross-sectional SEM micrographs of PGe layer before (PGe) and after annealing at 300 C, (APGe) respectively.
T. Hanu
s, B. Ilahi, A. Chapotot et al. Materials Today Advances 18 (2023) 100373
3

layers on porous substrates at low temperature.
While the islands coalescence starts in an early nucleation stage,
promoting initial pit formation, some deep pits crossing the
membrane are still present up to a deposited Ge thickness of 50 nm
(Fig. S2e). Also considering the measured RMS roughness peak
around 60 nm of deposited Ge (Fig. 3g), this specific thickness ap-
pears as a critical one for the membrane formation process by
homoepitaxial nucleation on PGe structure. Accordingly, we define
the critical thickness (Tc), as the minimum thickness required to
ensure the full islands' coalescence into continuously dense
membrane, without deep pits crossing the entire membrane. Once
Tc is reached, the fully densified membrane can be further thick-
ened, while improving the surface morphology as shown in
Fig. 3(gei). The pits start merging (Fig. 3aef) and their depth gets
continuously reduced as more Ge material is deposited (Fig. 3h),
while the surface pit density decreases until the obtention of
perfectly flat surface seen in Fig. 3i. As the thicknesses rises beyond
Tc, the growth becomes basically dominated by 2D layer by layer
mode. Consequently, the RMS roughness decreases exponentially
testifying an improved membrane's surface morphology. After the
deposition of a 750 nm thick Ge layer, significant drop in the sub-
micron scale pits' depth and increase of their diameter occur
leading to their merging and flattening towards a smooth surface.
Consequently, the large pits turn into surface ripples-like
morphology making them strongly anisotropic and hardly quanti-
fiable (Fig. 3e and h). Despite the persistence of nanoscale sized pits
residues (Fig. 3e), the surface roughness is already well below 1 nm,
testifying an excellent morphology suitable for further epitaxial
growth. Indeed, as shown by Fig. 3d, imperfection free completely
smooth surface with RMS roughness of 0.3 nm is obtained for a
1
m
m thick membrane. The critical thickness for fully coalesced
layer may vary depending on the PGe layer's properties, as the high
porosity structures obviously requires more material to ensure the
transition from 3D nucleation to 2D growth mode. These results
demonstrate that once the deposited thickness exceeds Tc, and the
membrane is fully densified, its surface morphology improves with
more deposited material. All corresponding growth stages on
porous substrate are schematically illustrated in Fig. 4.
To assess the feasibility of the proposed FSM growth process for
different PGe thickness and porosity, 1
m
m thick Ge membrane has
been grown under the same conditions on 1
m
m thick PGe layer
with approximately 70% porosity. This represents an extreme case
of thick and high porosity PGe substrate. Interestingly, fully
densified Ge FSM has been successively fabricated at 100 mm
wafer-scale (Fig. 5a), while the high porosity layer remains unre-
constructed (as shown later in the inset of Fig. 7a). The FSM show
Fig. 3. (a)e(f) 5 5
m
m
2
AFM scans of the Ge membrane for various Ge epilayers thicknesses. (g) Evolution of the surface RMS roughness during all the growth stages, where the
Red dashed line indicates the complete coalescence of the layer and the blue one corresponds to the initial surface roughness of the PGe layer. (h) Surface pits' depth and size
evolution as a function of the membrane thickness. (i) Pits surface density as a function of the membrane thickness.
T. Hanu
s, B. Ilahi, A. Chapotot et al. Materials Today Advances 18 (2023) 100373
4

RMS roughness of 1.2 nm, as illustrated by the AFM scan in Fig. 5.
Indeed, the surface still show ripple-like morphology. Referring to
the AFM analysis of the FSM thickness evolution for 54% porosity,
such a surface state is likely to characterize fully coalesced pits with
depth variation below 5 nm. This surface condition makes the
layers comparable to membranes grown on 54% PGe substrate with
thickness between 500 and 750 nm. This phenomenon is actually
predictable as the higher porosity generates sparser nucleation
sites and eventually necessitate more material to form a fully
coalesced layer and annihilate all the pits on the surface. The sur-
face morphology can be further improved by thickening of the
membrane following the expected flattening trend.
To evaluate the crystalline quality of the membranes grown
either on 70% or 54% PGe substrates, a comprehensive X-ray
diffraction (XRD) analysis was conducted. The out-of-plane 2ɵ
scans in Fig. 6a reveal unique well-defined Ge (400) and (200) [53]
peaks that correspond to the (001) substrate orientation, indicating
the monocrystalline nature of the FSM grown on both 54% and 70%
PGe substrates. This is possible due to the substrate-oriented
crystallites of the PGe layer [49], which transfer their orientation
to the membrane during epitaxial growth. Furthermore, the in-
plane configuration was used to ensure a low penetration depth
of the beam, probing only the Ge membrane, thus discriminating its
signal from that of the substrate. Interestingly, both in-plane pole
figures of Ge (220), shown by Fig. 6b, depict 4 well defined sharp
peaks, with fourfold symmetry, corresponding to the cubic crystal
structure of Ge, undoubtedly confirming the single-crystal quality
of the Ge FSMs independently of the PGe's porosity used as the
substrate.
The presence of unreconstructed PGe structure underneath the
Ge SFM constitutes a well-adapted separation layer with nano-
structured interface allowing for membrane lift-off. Indeed, the
cross-sectional SEM image (Fig. 7a) shows the easy fracture of the
70% porous interface between the Ge membrane and the bulk
substrate. For demonstration purpose, we show a successful full
100 mm wafer Ge FSM release by simple adhesive polymer tape and
transfer to a transparent, flexible plastic holder (Fig. 7 b).
Furthermore, the unreconstructed PGe separation layer also
offers a unique opportunity for substrate reuse, as the bulk sub-
strate material remains largely intact, with only PGe's residuals on
the top surface as shown by Fig. 7c. Compared to sub-micron
[38,41] to micron scale [35,36] pillars formed by high-
temperature annealing methods, in our case, the PGe crystallites
are only few nm in size [54] and have high specific surface
compared to the volume of the Ge material (in respect to Ge bulk
material or large pillars). This allows to completely oxidize the
majority of the PGe structure in H
2
O
2
solution, and then dissolve it
in HF with minimal bulk Ge material etching, compared to wet
etching methods. This treatment results in a clean surface with
RMS roughness ~0.72 nm (Fig. 7dee), which is suitable for addi-
tional cycles of porosification/epitaxial growth/membrane release,
as previously demonstrated [41]. This further highlights the ad-
vantages of using unreconstructed porous interface. The estimated
Ge consumption is approximately 1
m
m per cycle, since the bulk
material stays intact during the substrate cleaning. This process has
the potential to produce multiple Ge membranes from a single
substrate with minimal material loss. Accordingly, a 175
m
m thick
Ge wafer could be reused around 30 times before reaching the
Fig. 4. Schematic illustration of initial growth steps starting with PGe substrate, 3D island growth, coalescence of islands into Ge membrane, and thickening of the Ge membrane by
2D layer-by-layer growth.
Fig. 5. (a) Optical image of ~1
m
m thick Ge membrane grown on top of 1
m
m thick PGe layer with high porosity. (b) AFM scan of Ge membrane's surface grown on high porosity
substrate.
T. Hanu
s, B. Ilahi, A. Chapotot et al. Materials Today Advances 18 (2023) 100373
5

thickness of 145
m
m (Thinnest commercially available 100 mm Ge
wafers) and being recycled. Giving the rarity and cost of Ge [55],
this method has a potential to significantly reduce the cost of Ge-
based devices, while offering all the advantages of FSM.
3. Conclusion
In summary, we demonstrate the growth of monocrystalline Ge
membranes at 300

C on PGe substrates, while leaving the porous
structure of the substrate unchanged during the growth. The initial
nucleation stages on porous structure have been experimentally
investigated showing two growth regimes. Initially the growth is
dominated by 3D nucleation on top of the pores and their coales-
cence. Once the Ge membrane reaches the critical thickness of
coalescence, a dense membrane is formed, and the growth be-
comes governed by 2D layer-by-layer growth regime. At this stage,
the remaining pits at the surface are being annihilated during the
thickening of the membrane and good surface quality, with an RMS
roughness below 1 nm, can be reached. The XRD analysis demon-
strate the monocrystalline quality of the grown Ge membranes for
all samples independently of their porosity and thickness. Our re-
sults show that PGe layers can be used to fabricate detachable
Fig. 6. (a) 2
q
out-of-plane XRD scan of the Ge membrane grown on 54% and 70% PGe substrates and of the Ge bulk substrate as a reference, with logarithmic scale on y-axis. (b)
Ge(220) In-Plane pole figure of the Ge membranes prepared on PGe substrates with 54% and 70% porosity.
Fig. 7. (a) Cross-sectional SEM micrograph of the Ge FSM on the weak porous interface illustrating the fracture of the nanostructured interface and the detachment of the
membrane. The inset show a zoom on the unreconstructed high porosity layer underneath the membrane. (b) Optical image of 100 mm Ge membrane transferredtoflexible
substrate using adhesive tape. (c) Optical microscope image of the PGe remnants on the substrate after the detachment (d) and (e) Optical microscope image and AFM scan,
respectively, of the Ge substrate after the cleaning.
T. Hanu
s, B. Ilahi, A. Chapotot et al. Materials Today Advances 18 (2023) 100373
6

wafer-scale Ge FSM. Moreover, the nanometric crystallite size and
high specific surface of the PGe remnants on the substrate surface,
allow an easy cleaning process by oxidation and reuse of the sub-
strate for production of multiple Ge FSM. Furthermore, our finding
paves also the way to the fabrication of wafer scale FSM from low
temperature grown small bandgap materials for mid-IR optoelec-
tronics such as Ge(Si)Sn.
4. Methods
4.1. Sample preparation
PGe layers were prepared by optimized BEE process [49] of Ga-
doped, 100 mm (100) Ge wafers with 6

off-axis miscut towards
(111) orientation and 8e30 m
U
cm in resistivity, provided by
Umicore. The BEE was performed in a custom made 100 mm
porosification cell [49], using SP-50 BioLogic generator. Prior to this
process, Ge wafers were treated with HF (49%) solution for 5 min to
dissolve any native oxides present on the surface, rinsed with
EtOH(99%) and dried under N
2
flow. Sample were then introduced
into the porosification cell with HF(49%):EtOH(99%) (4:1, V:V)
electrolyte and etching and passivation pulses with 1 s duration
and 1 s rest time at the end of each cycle were applied. The medium
porosity (~54%) layers were formed using 1 mA cm
2
symmetric
etching/passivation current density. To produce high porosity
layers (~70%), the etching current density was increased to
2mAcm
2
. At the end of BEE process substrates with PGe structure
were rinsed with EtOH (99%), dried under N
2
flow and introduced
into the loading chamber of the CBE reactor.
Ge growth was carried-out in VG Semicon VG90H CBE reactor,
with a load-lock, transfer module maintained at ~6.10
9
Torr, and
thermocouple as a mean of monitoring the temperature during the
growth. The solid source of Ge, with a K-Cell temperature at
1250

C, was used to growth Ge with nominal growth rate of
500 nm h
1
. Samples were introduced to the growth chambre
directly at 300

C. Then various nominal thickness of Ge
(5e1000 nm) was deposited on PGe substrates at 300

C, and
chambre pressure ~6.10
6
Torr.
After the detachment of the membrane with adhesive tape, the
retrieved substrate was immersed in concentrated H
2
O
2
(30%) so-
lution for 1 min to fully oxidize the remains of the PGe structure.
This is followed by deoxidation in concentrated HF (49%) prior to
the reporosification.
4.2. Characterization
The top-view and cross-section of PGe layers and Ge/PGe
structures were observed with a Zeiss LEO 1540 XB scanning
electron microscope at 4.3 mm of working distance and 20 keV of
acceleration voltage, to evaluate the thickness of deposited material
and any morphological changes of the structure. The surface
morphology of the membranes was evaluated using Veeco
Dimension 3100, atomic force microscopy system, in tapping mode
with SSS-NCHR silicon probe and scan resolution 512 512 pixel.
The collected AFM profile data on various wafers locations were
also used to evaluate the pits' size and depth evolution on the FSM's
surface for various thicknesses. The structural properties of PGe and
epitaxial layers were investigated using Rigaku smartlab HRXRD
system with Cu K
a
X-ray source, Ge (220)x2 monochromator on the
incident beam, and HYPIX-3000 hybrid pixel array 2D detector. The
XRR was used to determine the critical angle of PGe layers, which is
directly linked to the porosity [49]. The out-of-plane and in-plane
XRD configurations were used to identify crystalline quality of
the Ge membranes.
Author contributions
The manuscript was written through contributions of all au-
thors. All authors have given approval to the final version of the
manuscript. Tade
a
s Hanu
s: Conceptualization, Methodology,
Investigation, Data curation, Original draft preparation, Review and
Editing, Visualization. Bouraoui Ilahi: Conceptualization, Method-
ology, Supervision, Validation, Review and Editing. Alexandre
Chapotot: Investigation, Review and Editing. Hubert Pelletier:
Investigation, Validation, Review and Editing. Jinyoun Cho: Vali-
dation, Review and Editing. Kristof Dessein: Validation, Review and
Editing. Abderraouf Boucherif: Supervision, Validation, Review and
Editing.
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgement
We thank Thierno Mamoudou Diallo and Rapha€
el Dawant for
scientific discussions, Guillaume Bertrand, Philippe-Olivier Provost
and all the technical staff of 3IT for the technical support. We thank
Umicore, Saint-Augustin Canada Electric (Stace), Innovation en

energie 
electrique (Innov
E
E), the Natural Sciences and Engineering
Research Council of Canada (NSERC), Fonds de recherche du Qu
ebec
(FRQNT), Mitacs, for the financial support. Abderraouf Boucherif is
grateful for a Discovery grant supporting this work.
LN2 is a joint International Research Laboratory (IRL 3463)
funded and co-operated in Canada by Universit
e de Sherbrooke
(UdeS) and in France by CNRS as well as ECL, INSA Lyon, and Uni-
versit
e Grenoble Alpes (UGA). It is also supported by the Fonds de
Recherche du Qu
ebec Nature et Technologie (FRQNT).
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.mtadv.2023.100373.
References
[1] K. Myny, Nat Electron 1 (2018) 30e39.
[2] Y. Lee, J.W. Chung, G.H. Lee, H. Kang, J.-Y. Kim, C. Bae, H. Yoo, S. Jeong, H. Cho,
S.-G. Kang, J.Y. Jung, D.-W. Lee, S. Gam, S.G. Hahm, Y. Kuzumoto, S.J. Kim,
Z. Bao, Y. Hong, Y. Yun, S. Kim, Sci. Adv. 7 (2021), eabg9180.
[3] P. Wang, M. Hu, H. Wang, Z. Chen, Y. Feng, J. Wang, W. Ling, Y. Huang, Adv. Sci.
7 (2020), 2001116.
[4] S. Choi, C. Kang, C.-W. Byun, H. Cho, B.-H. Kwon, J.-H. Han, J.-H. Yang, J.-
W. Shin, C.-S. Hwang, N.S. Cho, K.M. Lee, H.-O. Kim, E. Kim, S. Yoo, H. Lee, Nat.
Commun. 11 (2020) 2732.
[5] M.M. Shulaker, G. Hills, R.S. Park, R.T. Howe, K. Saraswat, H.-S.P. Wong,
S. Mitra, Nature 547 (2017) 74e78.
[6] G. Qin, Y.H. Jung, H. Zhang, N. Jiang, P. Ma, S. Stetson, M. Racanelli, Z. Ma,
Advanced Electronic Materials 8 (2022), 2101350.
[7] D.-H. Kim, J.-H. Ahn, W.M. Choi, H.-S. Kim, T.-H. Kim, J. Song, Y.Y. Huang, Z. Liu,
C. Lu, J.A. Rogers, Science 320 (2008) 507e511.
[8] Y. Kim, J.M. Suh, J. Shin, Y. Liu, H. Yeon, K. Qiao, H.S. Kum, C. Kim, H.E. Lee,
C. Choi, H. Kim, D. Lee, J. Lee, J.-H. Kang, B.-I. Park, S. Kang, J. Kim, S. Kim,
J.A. Perozek, K. Wang, Y. Park, K. Kishen, L. Kong, T. Palacios, J. Park, M.-C. Park,
H. Kim, Y.S. Lee, K. Lee, S.-H. Bae, W. Kong, J. Han, J. Kim, Science 377 (2022)
859e864.
[9] J. Shin, H. Kim, S. Sundaram, J. Jeong, B.-I. Park, C.S. Chang, J. Choi, T. Kim,
M. Saravanapavanantham, K. Lu, S. Kim, J.M. Suh, K.S. Kim, M.-K. Song, Y. Liu,
K. Qiao, J.H. Kim, Y. Kim, J.-H. Kang, J. Kim, D. Lee, J. Lee, J.S. Kim, H.E. Lee,
T. Hanu
s, B. Ilahi, A. Chapotot et al. Materials Today Advances 18 (2023) 100373
7

H. Yeon, H.S. Kum, S.-H. Bae, V. Bulovic, K.J. Yu, K. Lee, K. Chung, Y.J. Hong,
A. Ougazzaden, J. Kim, Nature 614 (2023) 81e87.
[10] Y.-C. Tai, S. An, P.-R. Huang, Y.-T. Jheng, K.-C. Lee, H.-H. Cheng, M. Kim, G.-
E. Chang, Nanoscale (2023), https://doi.org/10.1039/D2NR07107J.
[11] B. Kunert, Y. Mols, M. Baryshniskova, N. Waldron, A. Schulze, R. Langer,
Semicond. Sci. Technol. 33 (2018), 093002.
[12] H. Jin, H. Jin, L. Chen, L. Chen, J. Li, J. Li, X. An, Y.P. Wu, L. Zhu, H. Yi, K.H. Li,
K.H. Li, Opt. Lett., OL 45 (2020) 6671e6674.
[13] J. Chun, K.J. Lee, Y.-C. Leem, W.-M. Kang, T. Jeong, J.H. Baek, H.J. Lee, B.-J. Kim,
S.-J. Park, ACS Appl. Mater. Interfaces 6 (2014) 19482e19487.
[14] J. Lim, J. Shim, D.-M. Geum, S. Kim, IEEE Photon. J. 14 (2022) 1e9.
[15] J. Kang, Z. Cheng, W. Zhou, T.-H. Xiao, K.-L. Gopalakrisna, M. Takenaka,
H.K. Tsang, K. Goda, Opt. Lett., OL 42 (2017) 2094e2097.
[16] C. Liu, W. Ma, M. Chen, W. Ren, D. Sun, Nat. Commun. 10 (2019) 4873.
[17] M.R.M. Atalla, S. Assali, A. Attiaoui, C. Lemieux-Leduc, A. Kumar, S. Abdi,
O. Moutanabbir, Adv. Funct. Mater. 31 (2021), 2006329.
[18] N. Cho, M. Kim, Z. Ma, in: 10th International Conference on Group IV Pho-
tonics, 2013, pp. 77e78.
[19] S. Wu, H. Zhou, Q. Chen, L. Zhang, K.H. Lee, S. Bao, W. Fan, C.S. Tan, Appl. Phys.
Lett. 119 (2021), 191106.
[20] D. Nam, A.M. Roy, K.C.Y. Huang, M.L. Brongersma, K.C. Saraswat, CLEO:2011 -
Laser Applications to Photonic Applications, 2011, Optica Publishing Group,
2011, p. JTuI85. Paper JTuI85.
[21] F.T. Armand Pilon, A. Lyasota, Y.-M. Niquet, V. Reboud, V. Calvo, N. Pauc,
J. Widiez, C. Bonzon, J.M. Hartmann, A. Chelnokov, J. Faist, H. Sigg, Nat.
Commun. 10 (2019) 2724.
[22] A.A. La Mattina, S. Mariani, G. Barillaro, Adv. Sci. 7 (2020), 1902872.
[23] H. Zhao, Z. Xue, X. Wu, Z. Wei, Q. Guo, M. Xu, C. Qu, C. You, Y. Mei, M. Zhang,
Z. Di, Q. Guo, Npj Flex Electron 6 (2022) 1e10.
[24] Y. Kim, S.S. Cruz, K. Lee, B.O. Alawode, C. Choi, Y. Song, J.M. Johnson,
C. Heidelberger, W. Kong, S. Choi, K. Qiao, I. Almansouri, E.A. Fitzgerald,
J. Kong, A.M. Kolpak, J. Hwang, J. Kim, Nature 544 (2017) 340e343.
[25] W. Kong, H. Li, K. Qiao, Y. Kim, K. Lee, Y. Nie, D. Lee, T. Osadchy, R.J. Molnar,
D.K. Gaskill, R.L. Myers-Ward, K.M. Daniels, Y. Zhang, S. Sundram, Y. Yu, S. Bae,
S. Rajan, Y. Shao-Horn, K. Cho, A. Ougazzaden, J.C. Grossman, J. Kim, Nat.
Mater. 17 (2018) 999e1004.
[26] S.-H. Bae, K. Lu, Y. Han, S. Kim, K. Qiao, C. Choi, Y. Nie, H. Kim, H.S. Kum,
P. Chen, W. Kong, B.-S. Kang, C. Kim, J. Lee, Y. Baek, J. Shim, J. Park, M. Joo,
D.A. Muller, K. Lee, J. Kim, Nat. Nanotechnol. 15 (2020) 272e276.
[27] H.S. Kum, H. Lee, S. Kim, S. Lindemann, W. Kong, K. Qiao, P. Chen, J. Irwin,
J.H. Lee, S. Xie, S. Subramanian, J. Shim, S.-H. Bae, C. Choi, L. Ranno, S. Seo,
S. Lee, J. Bauer, H. Li, K. Lee, J.A. Robinson, C.A. Ross, D.G. Schlom,
M.S. Rzchowski, C.-B. Eom, J. Kim, Nature 578 (2020) 75e81.
[28] J. Jiang, X. Sun, X. Chen, B. Wang, Z. Chen, Y. Hu, Y. Guo, L. Zhang, Y. Ma, L. Gao,
F. Zheng, L. Jin, M. Chen, Z. Ma, Y. Zhou, N.P. Padture, K. Beach, H. Terrones,
Y. Shi, D. Gall, T.-M. Lu, E. Wertz, J. Feng, J. Shi, Nat. Commun. 10 (2019) 4145.
[29] Z. Lu, X. Sun, W. Xie, A. Littlejohn, G.-C. Wang, S. Zhang, M.A. Washington, T.-
M. Lu, Nanotechnology 29 (2018), 445702.
[30] J. Yoon, S. Jo, I.S. Chun, I. Jung, H.-S. Kim, M. Meitl, E. Menard, X. Li, J.J. Coleman,
U. Paik, J.A. Rogers, Nature 465 (2010) 329e333.
[31] T. Maeda, W.-H. Chang, T. Irisawa, H. Ishii, H. Hattori, V. Poborchii,
Y. Kurashima, H. Takagi, N. Uchida, Appl. Phys. Lett. 109 (2016), 262104.
[32] J.S. Mangum, A.D. Rice, J. Chen, J. Chenenko, E.W.K. Wong, A.K. Braun,
S. Johnston, H. Guthrey, J.F. Geisz, A.J. Ptak, C.E. Packard, Adv. Energy Mater. 12
(2022), 2201332.
[33] T. Akatsu, C. Deguet, L. Sanchez, F. Allibert, D. Rouchon, T. Signamarcheix,
C. Richtarch, A. Boussagol, V. Loup, F. Mazen, J.-M. Hartmann, Y. Campidelli,
L. Clavelier, F. Letertre, N. Kernevez, C. Mazure, Mater. Sci. Semicond. Process.
9 (2006) 444e448.
[34] H. Kim, S. Lee, J. Shin, M. Zhu, M. Akl, K. Lu, N.M. Han, Y. Baek, C.S. Chang,
J.M. Suh, K.S. Kim, B.-I. Park, Y. Zhang, C. Choi, H. Shin, H. Yu, Y. Meng, S.-I. Kim,
S. Seo, K. Lee, H.S. Kum, J.-H. Lee, J.-H. Ahn, S.-H. Bae, J. Hwang, Y. Shi, J. Kim,
Nat. Nanotechnol. 17 (2022) 1054e1059.
[35] S. Park, J. Simon, K.L. Schulte, A.J. Ptak, J.-S. Wi, D.L. Young, J. Oh, Joule 3 (2019)
1782e1793.
[36] V. Depauw, C. Porret, M. Moelants, E. Vecchio, K. Kennes, H. Han, R. Loo, J. Cho,
G. Courtois, R. Kurstjens, K. Dessein, V. Orejuela, C. Sanchez-Perez, I. Rey-
Stolle, I. García, Prog Photovolt Res Appl. (2022) 1e14, https://doi.org/
10.1002/pip.363414.
[37] A. Boucherif, G. Beaudin, V. Aimez, R. Ar
es, Appl. Phys. Lett. 102 (2013),
011915.
[38] E. Winter, W. Schreiber, P. Schygulla, P.L. Souza, S. Janz, D. Lackner,
J. Ohlmann, J. Cryst. Growth 602 (2023), 126980.
[39] N. Paupy, Z.O. Elhmaidi, A. Chapotot, T. Hanu
s, J. Arias-Zapata, B. Ilahi,
A. Heintz, A.B.P. Mbeunmi, R. Arvinte, M.R. Aziziyan, V. Daniel, G. Hamon,
J. Chr
etien, F. Zouaghi, A. Ayari, L. Mouchel, J. Henriques, L. Demoulin,
T.M. Diallo, P.-O. Provost, H. Pelletier, M. Volatier, R. Kurstjens, J. Cho,
G. Courtois, K. Dessein, S. Arcand, C. Dubuc, A. Jaouad, N. Quaegebeur,
R. Gosselin, D. Machon, R. Ar
es, M. Darnon, A. Boucherif, Preprint (2022),
03144.
[40] J.S. Ward, T. Remo, K. Horowitz, M. Woodhouse, B. Sopori, K. VanSant,
P. Basore, Prog. Photovoltaics Res. Appl. 24 (2016) 1284e1292.
[41] A. Chapotot, B. Ilahi, J. Arias-Zapata, T. Hanu
s, A. Ayari, G. Hamon, J. Cho,
K. Dessein, M. Darnon, A. Boucherif, Surfaces and Interfaces (2023) (To Be
Published) (n.d.).
[42] S. Tutashkonko, A. Boucherif, T. Nychyporuk, A. Kaminski-Cachopo, R. Ar
es,
M. Lemiti, V. Aimez, Electrochim. Acta 88 (2013) 256e262.
[43] S. Tutashkonko, S. Alekseev, T. Nychyporuk, Electrochim. Acta 180 (2015)
545e554.
[44] A. Dupuy, M.R. Aziziyan, D. Machon, R. Ar
es, A. Boucherif, Electrochim. Acta
378 (2021), 137935.
[45] Y.A. Bioud, A. Boucherif, A. Belarouci, E. Paradis, S. Fafard, V. Aimez, D. Drouin,
R. Ar
es, Electrochim. Acta 232 (2017) 422e430.
[46] A. Heintz, B. Ilahi, A. Pofelski, G. Botton, G. Patriarche, A. Barzaghi, S. Fafard,
R. Ar
es, G. Isella, A. Boucherif, Nat. Commun. 13 (2022) 6624.
[47] A.B.P. Mbeunmi, M. El-Gahouchi, R. Arvinte, A. Jaouad, R. Cheriton, M. Wilkins,
C.E. Valdivia, K. Hinzer, S. Fafard, V. Aimez, R. Ar
es, A. Boucherif, Sol. Energy
Mater. Sol. Cell. 217 (2020), 110641.
[48] M.N. Beattie, Y.A. Bioud, D.G. Hobson, A. Boucherif, C.E. Valdivia, D. Drouin,
R. Ar
es, K. Hinzer, Nanotechnology 29 (2018), 215701.
[49] T. Hanu
s, J. Arias-Zapata, B. Ilahi, P.-O. Provost, J. Cho, K. Dessein, A. Boucherif,
Advanced Materials Interfaces (2023) 2202495.
[50] C.V. Falub, H. von K€
anel, F. Isa, R. Bergamaschini, A. Marzegalli, D. Chrastina,
G. Isella, E. Müller, P. Niedermann, L. Miglio, Science 335 (2012) 1330e1334.
[51] W.H. Thompson, Z. Yamani, H.M. Nayfeh, M.-A. Hasan, J.E. Greene,
M.H. Nayfeh, MRS Online Proc. Libr. 452 (1996) 255e260.
[52] L. Liu, X. Zhang, S. Wang, G. Wang, J. Yu, X. Hu, Q. Xu, X. Xu, L. Zhang, Crys-
tEngComm 24 (2022) 1840e1848.
[53] P. Zaumseil, J. Appl. Crystallogr. 48 (2015) 528e532.
[54] A. Boucherif, A. Korinek, V. Aimez, R. Ar
es, AIP Adv. 4 (2014), 107128.
[55] G. Van Hoof, M. Schurmans, B. Robertz, J.-F. M
enard, K. Dessein, J. Sustain.
Metall. 6 (2020) 333e343.
T. Hanu
s, B. Ilahi, A. Chapotot et al. Materials Today Advances 18 (2023) 100373
8