d7.25 - progress activity report #7 · progress activity report #7 covered ... - fabrication of...

CARBON BASED SMART SYSTEM FOR WIRELESS APPLICATION

Start Date : 01/09/12 Project n° 318352 Duration : 48 months Topic addressed : Very advanced nanoelectronic comp onents: design, engineering, technology and manufacturability

WORK PACKAGE 7 : Project management

DELIVERABLE D7.25

Progress Activity Report #7 Covered period: T0+36 – T0+45

Due date : T0+45 Submission date : T0+48

Lead contractor for this deliverable: TRT

Dissemination level : PU – Public

D7.25 - Progress Activity Report #7 (T0+36 – T0+45)

2/34

2

WORK PACKAGE 7: Project management

PARTNERS ORGANISATION APPROVAL

Name Function Date Signature

Prepared by: S.Xavier R&D Engineer 05/10/16

Approved by: Afshin Ziaei Research Program Manager 05/10/16

DISTRIBUTION LIST

QUANTITY ORGANIZATION NAMES

1 ex Thales Research and Technology TRT Afshin ZIAEI

1 ex Chalmers University of Technology CHALMERS Johan LIU

1 ex Foundation for Research & Technology - Hellas FORTH George KONSTANDINIS

1 ex Laboratoire d’Architecture et d’Analyse des Systèmes

CNRS-LAAS Patrick PONS

1 ex Université Pierre et Marie Curie UPMC Charlotte TRIPON-CANSELIET

1 ex National Research and Development Institute for Microtechnologies

IMT Mircea DRAGOMAN

1 ex Graphene Industries GI Peter BLAKE

1 ex Thales Systèmes Aéroportés TSA Yves MANCUSO

1 ex SHT Smart High-Tech AB SHT Yifeng FU

1 ex Universita politecnica delle Marche UNIVPM Luca PIERANTONI

1 ex Linköping University LiU Rositsa YAKIMOVA

1 ex Fundacio Privada Institute Catala de Nanotecnologia

ICN Clivia SOTOMAYOR

1 ex Tyndall-UCC Tyndall Mircea MODREANU


3/34

3

CHANGE RECORD SHEET

REVISION LETTER DATE PAGE NUMBER DESCRIPTION

v0 07/2015 10 Initial version

v1 28/09/2016 28 All partners contributions

v2 06/10/2016 35 Final version


4/34

4

CONTENTS

1 PROJECT OBJECTIVES AND MAJOR ACHIEVEMENTS DURING RE PORTING PERIOD (T0+36 – T0+45) 7

1.1 PROJECT OBJECTIVES DURING THE REPORTING PERIOD (T0+36-T0+45) 7

1.2 MAJOR ACHIEVEMENTS DURING THE REPORTING PERIOD (T0+36-T0+45) 8

1.2.1 WP2 : Design and simulation activities 8

1.2.2 WP3 : Fabrication activities 8

1.2.3 WP4 : Test activities 9

1.2.4 WP6 : Dissemination and exploitation activities 9

1.2.5 WP7 : Management activities 9

2 MANAGEMENT ACTIVITIES (WP7) 9

2.1 MEETING 9

3 PROGRESS IN THE ACTIVE THECNICAL WORKPACKAGE 2 10

3.1 WP2 : DESIGN AND SIMULATION ACTIVITIES 10

3.1.1 WP OBJECTIVES 10

3.1.2 PROGRESS TOWARD OBJECTIVES : STATUS OF ACTIVES TASKS 10 3.1.2.1 Task 2.3 : Design of the CNT filter/oscillator (UNIVPM/IMT) 10 3.1.2.1 Task 2.4 : Design the CNT based antenna (UMPC/UNIVPM) 12 3.1.2.2 Task 2.5 : Design and simulation of RF graphene devices (UNIVPM/IMT) 14


4.1 WP3 : FABRICATION ACTIVITIES 15

4.2 WP OBJECTIVES 15

4.2.1 PROGRESS TOWARD OBJECTIVES : STATUS OF ACTIVES TASKS 16 4.2.1.1 Task.3.1 CNT and graphene growth technology 16 4.2.1.2 Task.3.6 CNT based antenna fabrication 19 4.2.1.3 Task.3.7 Development of carbon nanotube interconnects (SHT/CHALMERS) 19 4.2.1.4 Task.3.8 Development of RF graphene devices 21


5.1 WP4 : TEST ACTIVITIES 23


5.3 PROGRESS TOWARD OBJECTIVES : STATUS OF ACTIVES TASKS 24

5.3.1 Task.4.2 CNT RF switch 24

5.3.2 Task.4.4 CNT and Graphene based antenna 26


6.1 WP5 : PROJECT DEMONSTRATORS AND SYSTEMS INTEGRATION 27


6.1 PROGRESS TOWARD OBJECTIVES : STATUS OF ACTIVES TASKS 27

6.1.1 Task.5.1 Design of demonstrator and T5.3 Fabrication and tests of demonstrators 27

7 DISSEMINATION AND EXPLOITATION ACTIVITIES (WP6) 33

7.1 PUBLICATIONS & CONFERENCE 33


5/34

5

FIGURES Figure 1 : New NANO-RF Timetable....................................................................................................8

Figure 2 : CST MWS design (with main dimensions) of the new compact CNT-based tunable MW band-pass filter with: (a) wire inductors; (b) meander inductors. ..........................................................................11

Figure 3: EM simulation results (performed by means of CST MWS) for the new compact CNT-based tunable MW band-pass filter with: (a) wire inductors; (b) meander inductors. ....................................................12

Figure 4 : AWR simulated return loss and transmission, without and with CNTs, for the new compact CNT-based tunable MW band-pass filter with meander inductors and CNTs capacitance of 0.2 pF. The downshift of the filtering frequency is evident....................................................................................................12

Figure 5 : experimental VA MWCNT bundle impedance (real and imaginary part) extracted from five different

antenna devices under TE mode excitation in a 5-15 GHz frequency band ............................................13

Figure 6 : Example of co-simulation (circuit-based/3D electromagnetic) results of inductive VA MWCNT

antennas in TE mode, in reflection (left: S11 parameter) and input impedance (right). ...........................13

Figure 7 : X-band graphene antenna integrated with CNT TSVs: (a) schematic; (b) 3D radiation characteristic at 10 GHz (directivity); (c) comparison of |S11| parameter for the initial antenna with graphene patch (black trace) and for the antenna with additional TSV and backside transition (red trace) ...................................14

Figure 8 : Integration of antenna with graphene patch and CNT TSV on a PCB: (a) schematic; (b) comparison of |S11| parameter of the antenna and the antenna integrated with the PCB .............................................15

Figure 9 : a) Step height vs. buffer layer and ML coverage, b) ML graphene coverage vs. Ar pressure. .....................................................................................................................................17

Figure 10 : a) Buffer layer and ML coverage vs. temperature, b) ML and bilayer coverage vs. growth time .............................................................................................................................................17

Figure 11 : Increasing the size of SiC samples from 7×7 mm2 to 20×20 mm2 with large area continues ML coverage. .......................................................................................................................................18

Figure 12: CNT growth process profile ....................................................................................................19

Figure 13 : Schematic of process flowchart.............................................................................................20

Figure 14 : (a) MWCNT bundles were transferred into the via and second densification was performed to shrink the root part of the bundle. (b) The close-up of the SEM photos of figure 5(a). .......................................20

Figure 15 : (a) Illustration of the four-probe measurement. (b) DC I-V response for CNT TSV. ....................21

Figure 16 : Statistical coverage of single layer graphene as inferred from MicroRaman spectroscopy for sample G798 supplied by LiU ................................................................................................23

Figure 17 : Statistical coverage of single layer graphene as inferred from MicroRaman spectroscopy for sample G799 supplied by LiU ................................................................................................23

Figure 18 : test set description...........................................................................................................24

Figure 19 : picture of the probes and resistor Nota the RF probes are not used ...............................25

Figure 20 : Identification of the capacitance ......................................................................................25

Figure 21 : Observation of two actuations on the same DC switch device ........................................26

Figure 22 : : Example of experimental reflection coefficient (magnitude and phase) and input impedance

(real and imaginary part) of a VA MWCNT-based antenna under TE mode excitation (In red: with CNT

bundle – In blue: without CNT bundle). ...........................................................................................27


6/34

6

TABLES


7/34

7

1 PROJECT OBJECTIVES AND MAJOR ACHIEVEMENTS DURING REPORTING PERIOD (T0+36 – T0+45)

1.1 PROJECT OBJECTIVES DURING THE REPORTING PERIOD (T0+36-T0+45)

According to the timetable (see Figure 1), the project objectives over the reporting period are:

• Within the framework of WP 2 ‘Design and simulation activities’ : according to the first review

meeting, the work for WP2 continues in parallel of WP3 during the second year of the project

• Within the framework of WP 3 ‘Fabrication activities’ to: all the CNT and graphene based sub-modules that have been designed in the work package 2. The principal objective of this work package is to fabricate the test structures and also final components based on the results of WP2

• Within the framework of WP 4 ‘Test activities’: the different manufactured CNT components will be tested, consisting of the following test:

CNT FET and graphene electrical characterization: DC to RF o CNT RF switch tests o CNT based filter/oscillator tests CNT based antenna tests o RF graphene device (LNA ,mixer, detector) o Calibration of basic equivalent circuit models (parameter extraction based on

experimental data)

• Within the framework of WP 5 ‘Project demonstrators and Systems Integration‘: o fabrication and test CNT demonstrator, o fabrication and test Graphene demonstrator,

• Within the framework of WP 6 ‘Dissemination and exploitation activities : dissemination information and results of the project within the partners and youside the project, as well as proposing exploitation transfer plans and managing the Intellectual

List of publication and conference Explotation activities

• Within the framework of WP 7 ‘Project Management’ : to establish durable basis for the

project management and monitoring all along the project duration, through the following actions:

Regular project and technical meetings;

As described in the following sections, all these objectives have progressed between T0+36 and T0+45.


8/34

8

Figure 1 : New NANO-RF Timetable

1.2 MAJOR ACHIEVEMENTS DURING THE REPORTING PERIOD (T0+36-T0+45)

1.2.1 WP2 : Design and simulation activities In the reporting period (T0+36-T0+45), different devices of the project have been redesigned:

- New design of graphene antenna - Design of passive RF component

1.2.2 WP3 : Fabrication activities Regarding the development of the technology for the CNT and graphene growth, some achievements have been obtained:

- Improvement of the quality of CVD graphene after growth condition modification

Several CNT and graphene samples are also prepared and delivered to the consortium. During the period, some devices have been also fabricated:

- New run of fabrication for CNT and Graphene FET - Fabrication of test sample for evaluate the RF performance concerning the TSV CNTs - New CNT antenna - Optimization of the dielectric layer (HfO2) deposition for CNT and Graphene FET


9/34

9

1.2.3 WP4 : Test activities The different manufactured graphene components fabricated in WP3 are measured:

Characterization of CNT filter electrical characterization of dielectric layer (HfO2) necessary for CNT and Graphene FET

1.2.4 WP5 : Project demonstrators and Systems Integration Fabrication and test a nano T/R module based on the technologies developed in this project:

Fabrication and measure of CNT T/R module Fabrication and measure of graphene receiver T/R module

1.2.5 WP6 : Dissemination and exploitation activiti es There are several publication and conference from the consortium.

1.2.6 WP7 : Management activities During the period several meeting were organized:

- 1 project meetings, where all the partners are represented

2 MANAGEMENT ACTIVITIES (WP7)

WP leader Involved Partners

Duration Deliverables Milestones

Active Tasks Status

TRT

TRT, CHALMERS,

FORTH, LAAS, IMT, TSA,

SHT, UNIVPM, ICN, Tyndall

T0 – T0+45 D7.1 to D7.21

M7.1 - On-Going

2.1 MEETING In order to ensure a correct progress and a high coherence of the collaborative project, during the reporting period, numerous meetings and teleconference were organized: 3 project meetings, where all the partners are represented

o 40-Month Meeting, held in Barcelona (ICN) on 12-13 January 2016; o 42-Month Meeting, held in Heraklion (FORTH) on 02-03 March 2016; o 44-Month Meeting, held in Bucharest (IMT) on 18-19 May 2016;

2 exectuvive meetings, where WP leader are represented held in Brussels 10 February 2016

and 30 June 2016

1 teleconference with WP leader: o Teleconference meeting on 21 July 2016;

The agendas, minutes and presentations made for these meetings are downloadable on the private area of the website.


10/34

10

3 PROGRESS IN THE ACTIVE THECNICAL WORKPACKAGE 2

3.1 WP2 : DESIGN AND SIMULATION ACTIVITIES



Active Tasks Status

IMT

TRT, UPMC, IMT, TSA,

UNIVPM, LiU, ICN

T0 +3 – T0+41 D2.1 to D2.5

M2.1 T2.1 to T2.5 Postpone

All the deliverable for WP2 are submitted at T0+13.

3.1.1 WP OBJECTIVES In this WP, we will design and simulate all the CNTs and graphene based sub-components forming the nano T/R module to be demonstrated in WP5 comprising filters/oscillators, switches, mixers, LNAs and PAs and finally an antenna. The main technical objectives are:

Design and modeling of CNT and graphene based FET and then design and modeling of LNA, PA and mixer based on either CNT or graphene

Design and modeling of CNT switch Design and modeling of CNT filter/oscillator and graphene mixer Design and modeling of CNT antenna Design and modeling of LNA based graphene Design and modeling of a graphene loaded antenna.

3.1.2 PROGRESS TOWARD OBJECTIVES : STATUS OF ACTIVE S TASKS

3.1.2.1 Task 2.3 : Design of the CNT filter/oscillator (UNIVPM/IMT) We report the design and results of the simulations carried out with CST Microwave Studio® (CST MWS) for the optimization of the CNT-based tunable microwave (MW) band-pass filter. The layout is very compact and exploits:

1. three wires (diameter of 25 µm, length of 1 mm) in parallel for the inductors on the coplanar waveguide (CPW) signal (100 µm-wide), or three meander inductors;

2. Three CNT-based varactors (same as first fabricated filter). Fig. 1a shows the new filter with wire-based inductors, whereas Fig. 1b displays the filter with meander inductors.


11/34

11

The two configurations are 4.2-mm wide and 8.4-mm long. This substantial miniaturization provides small MW metal (gold) losses; furthermore, the CNT growth process is also improved with respect to the first prototype. CPW dimensions are already in agreement with standard measurement probe tips. Namely, the gap/signal/gap (G/S/G) dimensions are 50/100/50 µm, respectively. The technical details of the filters are: molybdenum in the IDTs area for CNT growth has a thickness of 50 nm; gold for CPW signal and ground deposition has a thickness of 1 µm, to prevent from skin depth effects; the inductance of each wire is about 5.121 nH, so that the inductance of 3 wires in parallel is about 1.707 nH (it could be possible to use up to 5 wires to tune the inductance value); the inductance of each meander is about 0.92 nH. As in the case of the first prototype, the biasing technique can be realized via the CPW by simply applying a “+” to the signal and a “-“ to the ground, respectively. Fig. 2a shows the electromagnetic simulations of the filter with wire inductors (modeled as thin perfect electric conductor – PEC – components) in terms of return loss |S11| and transmission |S21|, whereas Fig. 2b provides the simulation results for the filter with meander inductors. In both cases, we considered only the presence of the IDTs without CNTs. Each IDT exhibits an estimated capacitance value of 0.387 pF, so that the presence of the CNTs is expected to modify the overall varactor capacitance (hence, tuning the filtering frequency) according to the applied DC bias voltage. Table I summarizes and compares the performance of the two solutions.

Table 1 : main performance of the two proposed CNT-based tunable MW band-pass filters. Performance Layout with wires Layout with meanders

Filtering frequency 9.98 GHz 8.63 GHz

|S11|/|S22| -10.73/-30.45 dB -46.78/-11.07 dB

Insertion loss 1.68 dB 1.59 dB

Bandwidth (-3 dB) 120 MHz 519 MHz

Q-factor 617 489

From Table I, it is apparent how the layout with wires offers the best performance in terms of bandwidth and Q-factor. Finally, Fig. 3 shows the simulations (for the filter with meander inductors) performed with AWR to verify how |S11| and |S21| change by putting in parallel to the each IDT scattering matrix a lumped element (capacitor) emulating the CNTs effect. The significant thing is that AWR provides a fast method to test if changing the overall varactor capacitance affects the performance of the filter. In detail, by assuming a capacitance of 0.2 pF for each CNT matrix, we observed a downshift of the filtering frequency of about 500 MHz.

(a) (b)

Figure 2 : CST MWS design (with main dimensions) of the new compact CNT-based tunable MW band-pass filter with: (a) wire inductors; (b) meander inductors.


12/34

12

The fabrication steps are the same as reported last year and will be not reproduced here again. The masks production and devices fabrication will be performed at Thales France, whereas for the wire inductors we expect to solder the wires at IMT Bucharest, in order to verify the overall effect of the wires themselves on the filtering performance.

(a) (b)

Figure 3 : EM simulation results (performed by means of CST MWS) for the new compact CNT-based tunable MW band-pass filter with: (a) wire inductors; (b) meander inductors.

Figure 4 : AWR simulated return loss and transmission, without and with CNTs, for the new compact CNT-based tunable MW band-pass filter with meander inductors and CNTs capacitance of 0.2 pF. The downshift of the filtering frequency is evident.

3.1.2.1 Task 2.4 : Design the CNT based antenna (UMPC/UNIVPM) For each device, after a de-embedding technique procedure, frequency-dependant CNT bundle impedance extraction in TE and TM mode excitation confirm accessible additional design matching solutions at a frequency of 10 GHz, by compensative inductive (TE mode) or capacitive (TM mode) elements (Figure 5), while maintaining a 50 Ohms impedance (real part). A design choice of a serial inductive compensation technique has been validated for technological simplification. After several inductive structures studies in association with technological process, spiral configurations have been selected and led to the layout of 13 antenna designs. Circuit-based and 3D electromagnetic co-simulations have revealed achievable impedance matching and circuit resonance thanks to CNT bundle impedance (see Figure 6).


13/34

13

Figure 5 : experimental VA MWCNT bundle impedance (real and imaginary part) extracted from five different

antenna devices under TE mode excitation in a 5-15 GHz frequency band

Figure 6 : Example of co-simulation (circuit-based/3D electromagnetic) results of inductive VA MWCNT antennas in

TE mode, in reflection (left: S11 parameter) and input impedance (right).


14/34

14

3.1.2.2 Task 2.5 : Design and simulation of RF graphene devices (UNIVPM/IMT)

X band slot antenna with graphene patch and carbon nanotube (CNT) through CNT vias (TSVs) This work is performed to demonstrate the integration of graphene antenna with TSV CNT to demonstrate the integration of the antenna into a MMIC which is a thorny issue in many cases. The X band slot antenna loaded with a 200x200 µm2 square graphene patch described in D4.4. was further adapted for flip-chip integration using CNT TSVs. The CNT TSVs have a diameter of 200 µm and can be processed in a 250 µm thick silicon substrate. The initial design of the antenna considered on-wafer probing and bond wire integration and had the gap-signal-gap widths of the input coplanar waveguide (CPW) of 50-100-50 microns, with a ~50 Ω line impedance for a 525 µm thick high-resistivity (ρ~5000 Ω·cm) silicon wafer. To accommodate the CNT TSVs the input CPW was widened to 150-300-150 microns. The thinning of the substrate also meant a redesign of the antenna altogether, with the final size of the chip of 15.5 x 12.6 mm2. Three CNT TSVs are placed on the input CPW transmission line and connect the front side CPW with the backside. The 3D expanded model of the antenna with graphene patch and CNT TSVs is shown in Fig.x.1.(a). The 3D radiation characteristic is shown in Fig.x.1.(b), with a simulated directivity of 4.4 dBi at 10 GHz and two main lobes. A comparison between the reflection losses of the initial antenna design (black trace) and the antenna with CNT TSVs (red trace) is shown in Fig.x.1.(c).

(a) (b)

(c)

Figure 7 : X-band graphene antenna integrated with CNT TSVs: (a) schematic; (b) 3D radiation characteristic at 10 GHz (directivity); (c) comparison of |S11| parameter for the initial antenna with graphene patch (black trace) and for the antenna with additional TSV and backside transition (red trace)

The antenna can then be bonded directly using silver epoxy on a PCB. The PCB substrate chosen for integration was RO4003 (Rogers), which is a low loss material suitable for X band operation (εr=3.55; tanδ=0.0027 @10 GHz; substrate thickness 0.406 mm; copper cladding 17 µm). A window will be cut under the antenna so as not to interfere with the radiation characteristic. The coplanar waveguide on the PCB is computed for 50 Ohm characteristic impedance and is grounded using two


15/34

15

rows of vias on each side of the signal line. These vias ensure limited leakage. At the end of the PCB a standard SMA connector can be mounted, as described in Figure 8.(a). A 3D electromagnetic model of the antenna with CNT TSVs, integrated on the RO4003 PCB was developed and simulated. A comparison between the reflection losses of the antenna with CNT TSVs and the integrated antenna is shown in Figure 8(b). There is a slight influence which should not impede the proper operation of the antenna.

(a) (b) Figure 8 : Integration of antenna with graphene patch and CNT TSV on a PCB: (a) schematic; (b) comparison of |S11| parameter of the antenna and the antenna integrated with the PCB


4.1 WP3 : FABRICATION ACTIVITIES



Active Tasks Status

FORTH

TRT, CHALMERS,

FORTH, LAAS, UPMC, IMT, GI, SHT, LiU, ICN, Tyndall

T0 +3 – T0+42 D3.1 to D3.8

M3.1 T3.1 to T3.10 On Going

4.2 WP OBJECTIVES In this work package we will fabricate all the CNT and graphene based sub-modules that have been designed in the previous work package. The principal objective of this work package is to fabricate the test structures and also final components based on the results of WP2. The manufacturing process in this WP will be optimised and the products will be delivered to WP4. The technical objectives for this work package are:


16/34

16

- To develop CNT growth technology to achieve the desired structure following results of the design activities. CNT growth must be compatible with all the substrate technology to achieve the desired RF components as specified in Work-package 1.

- To develop graphene growth techniques either based on exfoliation for proof of concept RF graphene devices and at the wafer scale for graphene circuits to be used in the sub-modules

- To set-up a pilot line for manufacturing CNT for microwave applications.

- To fully characterize the CNT that have been grown by thermal CVD or plasma enhanced CVD . The characterization of graphene via exfoliation or epitaxial growth on SiC. The goal will be to verify that physical, structural, etc. properties will be compatible with the RF functions we want to achieve.

- To fabricate the RF submodules designed in the previous work package: CNT FET (LNA, PA and mixer), the RF switch, the RF filter/oscillator and the antenna, LNA based on graphene, graphene antenna , graphene mixer and graphene detector

- To develop various technologies that will allow for the integration of the sub-modules produced within this work package to be integrated on a single Si carrier wafer.

- To supply other work packages with CNT and graphene for characterization, modeling, simulation and demonstration and to optimize CNT and graphene according to their feedbacks.

4.2.1 PROGRESS TOWARD OBJECTIVES : STATUS OF ACTIVE S TASKS

4.2.1.1 Task.3.1 CNT and graphene growth technology

Graphene growth There are several ways to obtain graphene material and we started to study the different graphene synthesis methods.

SiC decomposition (LiU) Based on LiU work aims in the Nano RF project, we have developed our growth conditions to increase the size of graphene/SiC from 7×7mm2 to 15×15 mm2 and 20x20 mm2 by having full control on the epitaxial graphene growth process on SiC and obtain cm scale continuous coverage of ML (monolayer) graphene. To reach the aims during the reported period a series of graphene samples were grown on the Si face of 4H SI SiC in an inductively heated furnace at a temperature ranging from 1700-1950°C, in argon ambient with a p ressure range from 750 - 950 mbar, in Si rich ambient, and at different growth times. As the buffer layer is a precursor for graphene formation and can strongly influence its quality; depending on the buffer layer the integrity of the graphene may contain defects. Graphene formation was analyzed in respect to step bunching and surface decomposition energy differences created by the SiC basal plane stacking sequence on SiC polytypes. We showed that buffer layer halts step bunching process (Fig 1a) on 4H and 6H-SiC, which means surface energy becomes uniform all over the substrate surface after coverage by a buffer layer and subsequently resulting in uniform and continuous ML coverage. In Fig. 1a 100% means full coverage with buffer layer and the excess value is ML coverage.


17/34

17

The results from graphene samples grown at different argon ambient pressure show that there is an optimal argon pressure yielding a large coverage of ML graphene (Fig 1b). We demonstrate that the same optimal argon pressure holds for graphene growth at different temperatures and both Si and C-rich conditions at which a max monolayer graphene coverage takes place (Fig. 1b).

Figure 9 : a) Step height vs. buffer layer and ML coverage, b) ML graphene coverage vs. Ar pressure.

The temperature dependence of the buffer layer and ML graphene coverage is sublinear, which means the formation process is surface kinetics limited (Fig. 2a). The time dependence of ML and bilayer graphene growth shows (Fig. 2b) that graphene spreads faster on 4H-SiC substrates from the start of the growth, and ML coverage increases approximately linearly with increasing growth time. After ML completion the growth time does not have a pronounced effect on enlargement of the bilayer graphene area for both 4H and 6H-SiC polytypes.

Figure 10 : a) Buffer layer and ML coverage vs. te mperature, b) ML and bilayer coverage vs. growth time

Based on the above results we were able to develop our growth conditions to increase the size of SiC samples from 7×7 mm2 to 20×20 mm2 (Figure 11). We have elaborated a growth protocol for graphene on 15x15mm2 and 20×20 mm2 SiC substrates on which ~99% ML area can be obtained Figure 11, we have also increased ML graphene reproducibility up to 98 % for 7×7mm2 samples.


18/34

18

Figure 11 : Increasing the size of SiC samples from 7×7 mm2 to 20×20 mm2 with large area continues ML coverage.

Furthermore, we have shown that formation of nitride interface is a method to modify the electronic properties of graphene on SiC, which can be applied for intercalation and chemical doping. Our modeling results, using first principles calculations, demonstrate that nitrogen intercalation is a promising way to access charge neutral graphene on a semiconducting surface.

Graphite ex-foliation (GI) During this period (T+36M → T+45M) there were no requests from consortium partners for monocrystalline graphene flakes. So instead of preparing additional graphene samples, we allocated time to enhancing our system for manipulating graphene and other 2D crystals within an inert atmosphere. The system consists of a fully motorised microscope and a pair of manipulators within an argon glovebox, as shown in the two photographs below:

The atmosphere is under continual purification to maintain water and oxygen concentrations below 0.1 ppm. Both water and oxygen can contaminate or otherwise degrade a wide range of 2D crystals, particularly those that are chemically unstable in air such as black phorphorus (2D semiconductor) and niobium diselenide (NbSe2, 2D superconductor). Our system can be used to transfer 2D crystals on top of each other to create van der Waals heterostructures [1] – materials with novel properties.


19/34

19

Some of the enhancements made during this period:

• Addition of linear encoders to sample XY stage to improve movement accuracy. Movement resolution is now better than 1 µm.

• Addition of dedicated hardware interfaces for all motor axes – increases throughput and reduces user training time.

• Upgrade of microscope camera from 2MP CCD to 16MP CMOS imaging sensor - increases to both resolution and signal-to-noise ratio.

• Rationalisation of glovebox service feed-throughs while maintaining all remote control features. The number of feed-throughs was reduced from 8 to 4 – decreasing costs and installation time.

• Design changes to custom components to make machining and system assembly easier. [1] 2D materials and van der Waals heterostructures, K. S. Novoselov. A. Mishchenko, A. Carvalho, A. H. Castro Neto, Science (2016), http://science.sciencemag.org/content/353/6298/aac9439

4.2.1.2 Task.3.6 CNT based antenna fabrication The growth was performed in an Aixtron Black Magic Cold-wall CVD system. Chips were placed on a 2’ graphite heater, with a thermocouple for temperature monitoring. A typical profile of the chamber condition can be seen in Figure 12. First, the chamber was pumped down to less than 0.1 mbar, and 837 sccm flow of H2 was subsequently inputted. The graphite heater was heated up to a temperature of 500°C at a rate of 300nm/min and mai ntained at that temperature for 3 min in order to reduce the catalyst. The temperature was subsequently rapidly increased to a growth temperature 700°C. At the same time, acetylene was introduced a t a rate of 240 sccm as carbon source, and maintained for the 10 min duration of the growth. After the growth the chamber was cooled down to below 200°C in N2 atmosphere (1000 sccm flow rate).

Figure 12: CNT growth process profile

4.2.1.3 Task.3.7 Development of carbon nanotube interconnects (SHT/CHALMERS)

The proposed process is illustrated as figure 1 has shown. The CVD-grown MWCNT bundles (Figure 13 (a)) were partially densified on the upper half part. Therefore a funnel-shape CNT bundle (Figure 13 (b)) was formed. In parallel, the via was etched (Figure 13 (c)) on the chips by deep reactive ion etching (DRIE), and a layer of thermal release tape was attached on the front surface of the via chip as the transfer medium (Figure 13 (d)). The upper part densified CNT bundles were aligned to the via and launched down to the tape by flip chip bonder (Figure 13 (e)). The CNT bundles with root


20/34

20

part undensified were kept in the via (Figure 13 (f)) due to the adhesive strength between the CNT bundles and the tape when the donor silicon is lift up and removed. The second round densification was performed (fig. 1(g)), where after polymer was filled into the gap between chip and CNT bundles (Figure 13 (h)). The overburden of polymer was flattened (Figure 13(i)) by chemical mechanical polishing (CMP). Hereafter the thermal release tape was released (Figure 13 (g)).

Figure 13 : Schematic of process flowchart

The principle of densification on CNT bundles by solvent is that capillary forces stemming from the exposure to solution will squeeze together the sparsely located CNTs, thereby increasing the density of CNTs in the bundle by reducing its cross-sectional area. To improve the controllability of the CNT bundles shape, a vapor phase densification method was introduced in some papers, where solvent vapor is condensed onto the surface of the CNT bundles, instead of immersing CNT bundles into the solvent. The pristine CNT bundles are placed upside down over a bath of 1:1 mixture of acetone and water, heated to 100 °C. Since the densificatio n rate depends on the condensation rate on the CNT bundles, i.e. on the vapor pressure of acetone, the vapor pressure has to be controlled. This was done by confirming variables affecting the vapor pressure and fixing them, so that all parameters except time is kept constant. The following parameters were confirmed to affect the rate of densification, and corresponding are the method to keep them stable and uniform.

Figure 14 : (a) MWCNT bundles were transferred into the via and second densification was performed to shrink the root part of the bundle. (b) The close-up of the SEM photos of figure

5(a). By comparing the different degrees of densification, it can be clearly seen CNT bundles of 20 s exposure to the vapor solvent is best configuration for the scenario of double-densification transfer. Because on the one hand, the diameter of upper top part of CNT bundles was shrunken to 20 ~ 30


21/34

21

µm, which make it possible to insert the CNT bundles to the smaller via; On the other hand, along with the upper half densified part, the undensified lower half part of the CNT bundles make the whole CNT bundles funnel shape structure, which is quite robust to withstand external force interference. In this point, 20 s densification is selected for the following transfer into the via. When inserting of upper half densified CNT bundles into the via, as long as the upper top part of the CNT bundles slide into the via, the undensified root part of the CNT bundles would be squeezed inward mechanically by contacting the wall of the via. With the same setup of the first round densification, the porous root part of the CNT bundles were densified as shown in Figure 14(a) and Figure 14 (b). After the second densification, the gap between the CNT bundles and the wall of the via needs to be filled with curable epoxy. The purpose of polymer filling is to help fix the CNT bundles after curing and form a robust structure for the following CMP planarization process of back surface of substrate. This epoxy layer can also work as the insulation layer, which could reduce capacitive coupling and increase the reliability of TSV compared to conventional silicon oxide.

Figure 15 : (a) Illustration of the four-probe measurement. (b) DC I-V response for CNT TSV.

In order to measure the electrical conductivity of the CNT interconnects, 20 nm Ti and 300 nm Au were sputtered onto the back surface of the sample. The configuration of measurement by four probes method was illustrated as Figure 15(a), and three CNT TSVs were utilized for extracting one resistance of CNT TSV. The electrical resistance was found to be between 0.7 Ω and 1.3 Ω. If the whole CNT bundle was assumed to be a cylinder with average diameter around 30 µm and thickness of the 280 µm, the resistivity of the CNT bundles is calculated to be 2 ~ 3 mΩ cm.

4.2.1.4 Task.3.8 Development of RF graphene devices Dielectrics (Tyndall, ICN, FORTH) During the period covered by this report (i.e.T0+36-T0+45), Tyndall-UCC effort was directed to the activities from WP3, WP4 and WP7. In particular during this period, Tyndall-UCC has continued his work towards developing a room temperature e-beam process for the growth of high-k metal oxides (HfO2) suitable as gate oxides for graphene (GFET) and CNT FET devices that will be developed in Task 3.3 and Task 3.8.


22/34

22

During reporting period, Tyndall-UCC work have concentrated on the characterisation of the high-k dielectrics growth impact on the CVD single layer graphene corresponding to growth batches NANO RF 38 to NANO RF 41 (see Table 1 for details). Following the request of partner FORTH a new high-k metal oxides/high–k dielectric stack process at a low ALD growth temperature (200oC) was developed in Tyndall The new high-k metal oxides/ high–k dielectric stack have been grown on MIM structures using as bottom buffer HfO2 E-beam layers grown in FORTH (NANO-RF 38 and 39 runs in Table 1) and in Tyndall (corresponding NANO-RF 40 and 41 runs in Table 1). On the resulting MIM structures the electrical characterisation have perform by partner IMT and following their assessment it was confirmed that the dielectric is of good quality with low leakage current. Tyndall has proceeded to the growth of the new high-k metal oxides/ high–k dielectric stack on GFET (NANO-RF 42 run) and CNET FET (NANO-RF 43) devices with the new developed ALD process. Batch samples NANO-RF 42 and NANO-RF 43 have been send to FORTH for the completion of the GFET and CNT FET processing. Table 1. Summary of Tyndall-UCC high-k dielectrics growth activity during the reporting period (i.e.T0+36-T0+45)

During the reporting period Tyndall has continued the MicroRaman characterisation of the epitaxial graphene samples grown on SiC by partner LiU. The main results for MicroRaman characterisation of two epitaxial graphene samples grown on SiC (G798 and G799) samples supplied by LiU are presented in Figure 16 and Figure 17.


23/34

23

30 35 40 45 50 550

10

20

30

40

50

60

89% SLGassuming SLG

2D HWHM<50cm-1

Cou

nts

2D HWHM (cm-1)

G798 SLG on SiC

Figure 16 : Statistical coverage of single layer gr aphene as inferred from MicroRaman spectroscopy for

sample G798 supplied by LiU

35 40 45 50 55 600

20

40

60

80

100 G799 SLG on SiC

Cou

nts

2D HWHM (cm-1)

97% SLGassuming SLG

2D HWHM<50cm-1

Figure 17 : Statistical coverage of single layer g raphene as inferred from MicroRaman spectroscopy for sample G799 supplied by LiU

The MicroRaman characterization results outline the excellent quality of sample G799 (97% single layer graphene coverage) grown by partner LiU.


5.1 WP4 : TEST ACTIVITIES



Active Tasks Status

LAAS

TRT, CHALMERS,

FORTH, LAAS, UPMC,IMT, GI, SHT,UNIVPM, LIU,TYNDALL

T0 +15 – T0+45 D4.1 to D4.7

MS4 T4.1 to T4.6 On Going


24/34

24

5.2 WP OBJECTIVES This work package is dedicated to the electrical characterization (DC to RF) of all the devices fabricated in WP3. The results obtained will be compared with the results obtained through simulations in WP2 and basic equivalent circuit models will be calibrated following parameter extraction from static (DC) to high-frequency (HF) measurements. A strong interaction is planned with WPs 2 and 3, in order to take into account all the fabrication available parameters as well to feedback into fabrication and application design the needed optimization

5.3 PROGRESS TOWARD OBJECTIVES : STATUS OF ACTIVES TASKS

5.3.1 Task.4.2 CNT RF switch A test set was built to measure the switching behavior of the Carbon-Nano-Tube (CNT). The CNT do not stands large currents; a large resistor is needed to limit the DC current. We made the test with a 50MegOhm. Figure 18 present a schematic of the experimental set-up

Figure 18 : test set description

A function generator provides the chosen wave shape and it is amplified up to tens of volts. An oscilloscope is used to measure the waveforms. The scope has a very large memory, so we can store the whole waveforms and analyzed them later. The waveforms can also be read by the spice simulator, so we can fit the circuits values and behavior. It is required to limit the parasitic capacitance in the test implementation and especially the capacitances after the 50MegΩ and on the measurement side. The DC current will be limited by the 50MegΩ and the scope10MegΩ. When switching, the current is not limited except by the CNT resistance. The available energy is set by the parasitic capacitance between the 50MΩ and the probe to ground (i.e the baseplate on which the sample is placed). On switching, this energy will flow very quickly to the measurement probe and scope probe capacitances. Figure 19 shows a picture of the set up used for the DC measurement with probes, 50Mohm resistance and the scope probe


25/34

25

Probes

50MΩ

Scope

Probe

Figure 19 : picture of the probes and resistor Nota the RF probes are not used The critical point for the measurement on the CNT DC switch is the parasitic capacitance which can cause the destruction of CNTs. The parasitic capacitance before the CNT switch is identified with preliminary measurements. The scope probe is typically 10pf and 10MegΩ. Rough order of magnitude of the current may be: if we have 25V and a CNT resistance of 25kΩ, this is 1mA peak and it decreases to 0.5mA in ~0.12µs. (this may lead to the CNT destruction). Figure 20 show a schematic of the different capacitance

Figure 20 : Identification of the capacitance Several measurements have been done using this experimental set-up and two successive actuations have been observed. The results are summarized on Figure 21


26/34

26

Figure 21 : Observation of two actuations on the sa me DC switch device A 25 V, we observe a peak corresponding to a first actuation between 2 CNTs of middle. SEM observations show a reduction of the CNTs height and validate the actuation. Same measure on the same device is performed and a second activation is observed at 45V. The difference can be explained by the fact that CNTs are shorter and therefore requires a higher voltage for activation. As before, destruction CNT is observed (one is shorter and one moved out of the electrode) This has been reproduced across devices and the same behavior is observed.

5.3.2 Task.4.4 CNT and Graphene based antenna Technological process improvements have been done by SHT partner in order to enhance aspect ratio of CNT bundles from 6 to 7 from available previous antenna designs with direct CPW RF feeding in TE or TM excitation mode, with optional series or shunt capacitive coupling. S-parameters measurements of 54 devices (27 identical versions with and without CNT bundle) , and in two mode RF excitation, have been performed ) in a dark probe test experimental setup environment after repeatability verification by three iterative measurement sessions with associated SOLT calibration procedure in 0.05-67 GHz frequency band (2001 points). Experimental results demonstrate device impedance matching with a TE or TM mode excitation (from reflection coefficient) (47 Devices) at a microwave frequency around 60 GHz, an associated antenna complex impedance resonance @ 60 GHz leading to radiation expectation (21 Devices), and a resonance frequency shifting imposed by CNT bundle presence (21 Devices).


27/34

27

Figure 22 : : Example of experimental reflection coefficient (magnitude and phase) and input impedance (real and

imaginary part) of a VA MWCNT-based antenna under TE mode excitation (In red: with CNT bundle – In blue: without

CNT bundle).


6.1 WP5 : PROJECT DEMONSTRATORS AND SYSTEMS INTEGRA TION



Active Tasks Status

ICN

TRT, CHALMERS,

FORTH, LAAS, UPMC,IMT, GI,TSA,LiU

T0 +32 – T0+48 D5.1 to D5.4 MS5, MS6

T5.1 to T5.3 On Going

6.2 WP OBJECTIVES The objective of this WP is to fabricate and test in a simulated environment a nano T/R module (smart system), representative of a complete wireless system, but with a limited number of Nano T/R module elements (20 or 30), based on the technologies developed in this project, to validate the feasibility of a low cost assembling and production process, and identify potential problems and issues to develop optimised T\R module at a later stage outside the project. Another objective is to assess the optimisation of the full process (technology production and assembling) to devise and evaluate an industrial plan for a follow-on.

6.1 PROGRESS TOWARD OBJECTIVES : STATUS OF ACTIVES TASKS

6.1.1 Task.5.1 Design of demonstrator and T5.3 Fabr ication and tests of demonstrators

• Low noise reception module based on graphene devices :


28/34

28

Graphene demonstrator specifications

The graphene demonstrator schematic and mockup of the MCM wire bonding implementation is given in the figure below. This includes on chip-passives for module (LNA, detector) lPCB evel matching passives networks and I/O interconnections

This solution provides several advantages with respect to the NANORF project:

Quick and versatile platform to test NANO-RF buildi ng blocks Integration of chip coming from different technolog y Flexibility in passive and interconnect definition (on-chip and/or off-chip) Independent testing of each module (validation and deembedding)


29/34

29

Chip Count: Assembly type

- Antenna chip : G loaded slot antenna CPW input - LNA: GaAs LNA TGA 2512 (Triquint) MS input - Bias (LNA) capacitor: ATC600 series (100pF) SMD mount - Detector chip: G-YBJ CPW input

Note that the originally planned bias T network ()at the antenna and detector input was eliminated in order to simplify the design. The tenability properties of the antennas and of the detector can be demonstratetd with dedicated PCB specifically manufactured for each part. The chosen PCB was the RO4003: Laminate which thickness is 406µm with 35µm thick copper metallization on both sides. The circuit was desoigned on a GND-CPW of size (trace/gap) 723/254 µm and by adding grounding vias of 400µm diameter. The SMA connector are edged mounted.

Grounded CPW dimensions and edge mounted SMA Bias networks and LNA was provided by COTS elements and SMD assembled and wire bonding respectively. The NANORF project graphene loaded dipole antenna and the Grapnene Y-branch Junction (YBJ) based detector ware cutted and fitted I dedicated PCB opening and wire bondned to the PCB. Noteworthy is the presence of a metal (brass) chuck below the PCB such to provide mechanical robustness and thermal cooling to the circuit. This is even more important for the LNA which was positioned on a custom micromachined ridge made out of the metal chuck


30/34

30

Figure: Schematic of the PCB mounting and the underlying metal carrier.

Picture of the actual PCB board and the different elmements assembled therein. Finally in order carry out the independednt test of eack block, (antenna, LNA, and detetctor), single PCB part have been realized allowing SMA independedent testing.


31/34

31

Figure: Independent test board for the demonstrator building blocks. In a similar manner the CNT demonstrator has been assembled with a combination of available part issued from the project and completed with COTS elements.

• Tx/Rx module based on CNT devices :


32/34

32

Specification fo the CNT demonstrator

- Figure: Schematic of the CNT demonstrator

Figure: Mockup fo the CNT demonstrator

Chip Count: Assembly type

- Antenna chip : CNT monopole on CPW access line CPW input - LNA: GaAs LNA TGA 2512 (Triquint) MS input - Bias (LNA) COTS capacitor and resistors SMD mount - SPDT (COTS) :MACOM MA4SW series CPW input - PA (COTS) TGA2627 Triquint SMD type


33/34

33

Figure: layout and compenent view of the PCB currently under fabrication.

7 DISSEMINATION AND EXPLOITATION ACTIVITIES (WP6)



Active Tasks Status

TRT TRT, ICN T0 +0 – T0+45 D6.1 to D6.6 - On Going

7.1 PUBLICATIONS & CONFERENCE

• Publications

o M. Dragoman, A. Dinescu, D. Dragoman,Room temperature on-wafer ballistic graphene field-effect-transistor with oblique double-gate Journal of Applied Physics 119, 244305 (2016); doi: 10.1063/1.4954639

o M.Dragoman, Carbon-based nanodevices routes: inks, flakes, spaghetti, wafers-invitated papers 8th International Conference on Advanced Materials, ROCAM 2015.

o M.Dragoman, Receiving microwave signals with graphene-invited paper , WOCSDICE ,Smolenice, 2015.

o D. Mencarelli, S. Bellucci, A. Sindona, L. Pierantoni, Spatial dispersion effects upon local excitation of extrinsic plasmons in a graphene micro-disk Article in Journal of Physics D, Applied Physics, October 2015.

o D. Mencarelli, L. Pierantoni, M. Stocchi, and S. Bellucci, Efficient and versatile graphene-based multilayers for EM field absorption Applied Physics Letters 109, 093103, 2016.


34/34

34

o Sun, S., et al, Vertically aligned CNT-Cu nano-composite material for stacked through-silicon-via interconnects, Nanotechnology, 27 (33), 335705, 2016.

o Sindona, A., Pisarra, M., Mencarelli, D., Pierantoni, L., Bellucci, S., Plasmon modes in extrinsic graphene, Ab initio simulations vs semi-classical models, NATO Science for Peace and Security Series B: Physics and Biophysics, pp. 125-144, 2016.

o Xin Jin1, James. C. M. Hwang1, Davide Mencarelli2, Luca Pierantoni, Marco Farina2 Calculating tip-sample capacitance and charge density of a near-field scanning microwave microscope, accepted to IEEE Magazine

o N.M. Caffrey, R. Armiento, R. Yakimova and I.A. Abrikosov, PHYSICAL REVIEW, B 92, 081409(R) (2015)

• Conference o A.Dinescu , M.Dragoman, A.Avram, Scanning electron microscopy for nanoscale

characterization and patterning of graphene devices, 16th Nanoscience and Nanotechnology Conf. , Frascatti, 2015-invited paper.

o M. Dragoman, M. Aldrigo and A. Radoi, Switching microwaves with 2D materials, MEMSWAVE Barccelona, Spain, 2015.

o M. Aldrigo, A. Stefanescu, M. Dragoman and D. Vasilache, Enhancement of capacitive RF MEMS switches reliability based on a carbon nanotubes array embedded in the dielectric, MEMSWAVE Barccelona, Spain, 2015.

o M.Dragoman, The sinuous path of electromagnetic waves in 2D materials inks, flakes,islands and flatlands, invited paper , 11 International Conference on Optics, Micro-nanophotonics IV ROMOPTO 2015, Romanian Academy, p. Bucharest, 2015.

o M.Dragoman, 2D materials nanotechnologies between great expectations and lost illusions, invited paper, 3rd International Conference on Nanotechnologies and Biomedical Engineering,, p.47 , Chisinau, Moldova (2015).

o A.C. Obreja, S.Iordanescu, R.Gavrila, A.Dinescu, F. Comanescu, A.Matei, M.danila, M.dragoman, H. Iovu, Flexible films based on graphene/polymer nanocomposite with improved electromagnetic interference shielding, 38 th IEEE CAS Conference, Sinaia, Romania,p.49-52, 2015.

o M. Aldrigo, M.Dragoman, L.Pierantoni, D.Mencarelli, and G. Deligeorgis, Back-gate bias of a graphene antenna via a smart background metallization, 38th CAS Conference , Sinaia, Romania, p. 131-134, 2015.

o Mencarelli, D., Pierantoni, L., Rigorous simulation of ballistic graphene-based transistor, IEEE MTT-S International Microwave Symposium Digest, 2016-August, 7540140.

d7.25 - progress activity report #7 · progress activity report #7 covered ... - fabrication of...

Documents