Hollow Core Fibers:

Past, Present & Future

Thomas Bradley, University of Southampton
 Hollow Core Optical Fibres:
Past, Present & Future
Thomas D Bradley, Gregory Jasion, Hesham Sakr, John Hayes, Kerrianne Harrington, Eric Numkam
Fokoua, Ian A Davidson, Austin Taranta, Seyd Mohammad Abokamis, Yong Chen, N V Wheeler, Marco
Petrovich, David J Richardson and Francesco Poletti
Micro-structured Optical Fibre Group, Optoelectronics Research Centre, University of Southampton

10/11/2020
OSA Webinar: Fibre Optics Technical Group
Optoelectronics Research
Centre – Fibre Fabrication
• 4x Fibre Draw Towers
• 2x MCVD Lathe + 1x OVD System
• 1x Glass Working Lathe
• Dedicated ISO5 Preform Preparation Area
Micro-structured Optical Fibre
Group
• Simulations
• Electromagnetics
• Fluid Dynamics
• Fibre Fabrication
• State of the Art Cleanroom Facilities
• Characterisation
• Spatial & Spectral Imaging
• Low Coherence Interferometry
• Data Transmission
• State of the art Laser Labs
• Applications
• High Power Laser Delivery
• Gas Sensing
• Telecommunications
• Nonlinear Optics
Hollow Core Fibres – the
research field
• CPPM, University of Bath, UK Nonlinear Optics
• Xlim Research institute, France
• Fibre Optics Research Centre, Russia Molecular & Atomic Spectroscopy
• CREOL, University of Central Florida, USA
• Optoelectronics Research Centre, University of Southampton, Telecommunications
UK
• Max Planck Institute for the Science of Light, Erlangen,
Germany Laser Power Delivery
• PhLam, University of Lille, France
• OFS Research Labs, USA Fibre Design & Fabrication
• Beijing University of Technology, China
• Danish Technical University, Denmark Fibre Optic Sensors
• Ecole Polytechnique Fédérale de Lausanne, Switzerland
• ….
Overview
• Hollow Core Optical Fibres
• What?
• Why?
• How?
• Hollow Core Photonic Bandgap Fibres
• Guidance Mechanism
• State of the Art
• Hollow Core Anti-Resonant Fibre
• Path way to NANF
• Nested Anti-Resonant Nodeless Fibres
• Progress
• State of the art
• Conclusions
• Applications of Hollow Core Fibres
 Hollow Core Optical Fibres –
The Differences
Single Mode Fibre Multi Mode Fibre Hollow Core Fibre

Cladding

Cladding

Cladding

Cladding
Cladding
Cladding
Core

Core

Core
Refractive
Index Profile
 Hollow Core Fibres – what?

Ohmic and macrobend loss Absorption loss Surface scattering loss Strong Resonant Lossesl

All few/multi moded

Hollow core fibres – why?

Vast topological range of fibres proposed:

• Low nonlinearity • Wide enough
• Low latency bandwidth
• Low dispersion • High damage threshold

Develop a waveguiding technology with all the

advantages of flexible all-glass fibres without the
penalties of light propagation in glass

8
 Hollow Core Fibres – how?
2 Stage Stack & Draw

Glass Tube (OD ~ 20

– 30 mm)

Capillary (OD ~ 1 –
2 mm)
~ 100 nm scale cladding
Stacking features

Jacketing (OD ~ 20 mm)

Caning (OD ~ 3 – 5 mm)

Fibre Drawing (OD ~ 150 µm)
Benabid, Phil. Trans. R. Soc. A, 2006, 364

9
Virtual Draw Model
Aim: to predict the final geometry of fibre given a
preform geometry and draw conditions.
Geometry parameters:
• Capillary outer / inner diameters.
• Cane and jacket tube inner/outer diameters.
• Target fibre outer and inner diameters.
Draw parameters:
• Feed / draw speeds
• Furnace temperature and profile
• Material data: viscosity and surface tension
• Core and Capillary pressures

Jasion et al. FIO 2018, Washington DC 10

Virtual Draw Model
1.A simple capillary draw model1 derived from
the Navier-Stokes equations is used for the
outer jacket tube.
2.The inner capillary tubes use the same
model but are bound to the jacket tube.
3.The inner capillary solution has 2 parts:
• axial draw down determined by the
jacket tube solution.
• The lateral dynamics are solved by the
Fitt model using gas pressures,
surface tension and viscous stress.

1 Fitt et al, J. of Eng. Mathematics 43, no. 2-4, 201-227, (2002).

2 Jasion et al, Opts Exp 27, 15, 20567-20582, (2019)

11
0%
0

10 Draw Dynamics
20

30
During the draw the microstructure
geometry is governed by forces:
40
• applied gas pressure.
50
• surface tension.
Z%

60 • viscous stress.
70
Gas pressure dominates at the start
(expansion), and surface tension
80
dominates at the end (contraction).
90

100
-1 0 1
draw diameter ratio 12
 0%
0

10 Draw Dynamics
20

30

50% 40

50
Z%

60

70

100%
80

90

100
-1 0 1
draw diameter ratio 13
Overview
• Hollow Core Optical Fibres
• What?
• Why?
• How?
• Hollow Core Photonic Bandgap Fibres
• Guidance Mechanism
• State of the Art
• Hollow Core Anti-Resonant Fibre
• Path way to NANF
• Nested Anti-Resonant Nodeless Fibres
• Progress
• State of the art
• Conclusions
• Applications of Hollow Core Fibres
 Key HC-PBGF properties Hollow core
Obtained by removing elements at the
Periodic lattice of centre
rods/holes of the lattice
Determines optical
bandgap
in the longitudinal “Air guided modes”
direction At wavelengths within the
Optical Bandgap
Low overlap with glass surround (0.1-0.2%)
Transmission

∆λ

λ0
Stack First Stage Draw “Cane”
 Low loss & wide bandwidth at 1.55 µm
40
Wheeler et al. OFC PDP 2012
35 Mangan et al. OFC PDP 2004
30

Loss (dB/km)
25
20
15
10

• Low loss (3.5 dB/km) combined 5 • SOTA loss (1.7 dB/km), surface
with wide bandwidth (160 nm) 0 mode limited bandwidth (20 nm)
1400 1450 1500 1550 1600 1650
Wavelength (nm)
• Achieved through careful use of pressure in
fabrication to control surface modes
• Eight-fold BW improvement with just ~2x
higher loss than current SOTA
Wheeler et al., OFC PDP 2012
 The promise of HC-PBGF
100

Lowest Loss @2µm

SOTA (wide BW design)
Lowest loss ≈ 1.8dB/km
Loss (dB/km)

10
(1.7 dB/km)
Model Prediction:
0.2dB/km

1
SMF

0.1
1000 1250 1500 1750 2000 2250 2500
Wavelength (nm)
Key factor in reducing loss is achieving highly regular structures, particularly in the core region
 Improving reach: New record 11km
length Strut t/L aspect ratio: ~10 11
>80% yield
• 1m long preform
• ~20mm diameter
• Max yield: ~14 km
@ ,
20 20 • Loss measured via long
20m 18 cutback
0
Transmission (dBm)

16
Min Loss: 5.2dB/km 14
(~11km to 20m)

Loss (dB/km)
-20
@1560nm
11570m
520m (20m cutback) 12 • Using SC source
Loss
-40 10
8 • Transmission BW: >160nm
-60 6 (subsequent • Core diameter: 29.7 ± 1𝜇𝜇𝜇𝜇
4
-80
11,07km 2
measurements ≈200nm - • Average Pitch: 6.2 ± 0.2𝜇𝜇𝜇𝜇
-100 0 previous SOTA 160nm) • Average Cladding strut thickness: 48.0 ± 10𝑛𝑛𝑛𝑛
1470 1500 1530 1560 1590 1620 • Thinnest strut thickness : 27.5 ± 5𝑛𝑛𝑛𝑛
Wavelength (nm) • Loss in the region of • Average 𝑑𝑑/Λ: 0.992
Chen et al, OFC PDP 2015 5dB/km • Air-filling fraction > 96%
 Transmission: Low latency -11km
• Single channel, 10G
RZ, scanned across C
band
• Main challenge: 65-
67dB loss
• Nonlinearity not a b2b
problem
 Error free transmission, (BER<1e-9) no error floor on all tested
channels
 1.4-3.9dB power penalty likely due to OSNR limitation
 11km transmission: 16μs latency reduction from all-glass
equivalent fibre link
+
 First recirculating loop experiment with 2 sub-sections of this Chen et al, OFC PDP 2015
fibre (2.7km and 3.5km)
Kuschnerov et al. ECOC 2015 Th 1.2.4
 Total length = 74.8km
 SoTA long-length HC-PBGF
OFC 2015 PDP
Mangan 2004
3.85km 3dB/km
BW x Length / Loss

Wheeler 2012
Frosz 2013

100 van Uden 2013

(nm km /dB)

Zhang 2014
2

Liu 2015
Mangan 2015

Chen 2015
This work Liu et al., JLT 33,1373, 2015.
10 Reference λ (nm) Loss (dB/km) BW (nm) Length(km)
Mangan et al. 2004 1565 1.7 20 0.8
Wheeler et al. 2012 1510 3.5 160 0.25
Frosz et al. 2013 1530 1.8 40 0.5
van Uden et al. 2013 1510 8 160 0.95
2003 2005 2007 2009 2011 2013 2015 Zhang et al. 2014 1990 2.8 85 1.15
Liu et al. 2015 2010 3.2 160 3.85
Year
Mangan et al. 2015 1540 6.5 12 2.75
This
Chen work
et al. 2015 1560 5.2 200 11.07
Transition from HC-PBGF to
Antiresonant Fibre
HC-PBGF NANF
Hollow Core Antiresonant Fibres
2016 FORC 2017 Xlim 2018 BUT
• Pioneering work from
- Bath
- Limoges
- FORC Moscow
- Erlangen
- ORC
- …
2018 Max Planck, Erlangen 2019 ORC 2020 University of Bath 2020 ORC

30 µm

1) Kosolapov et al, Qnt Elec 46 (3), 2016, 2) Debord et al, Optica 4 (2), 2017, 3) Gao et al, Nat Comms (9), 2018, 4) Roth et
al, Optica 5 (10), 2018 , 5)Yerolatsitis et al, JLT 38 (18) 2020
First Data-transmitting ARF

22.3 µm

40.2 µm

Width = 359.6 nm

20μm
Hayes et al. OFC 2016 PDPTh5A.3
 1 octave
160

140

120
Loss (dB/km)

100

80

60

40

20
400 nm @ <30 dB/km 100 m
0
800 1000 1200 1400 1600 1800 2000
Wavelength (nm)
Hayes et al. OFC 2016 PDPTh5A.3
OFC 2016, PDP Th5A.3

Nested Antiresonant Nodeless Fibre (NANF)

F. Poletti, Opt Express 22, 23807 (2014)
Guidance Mechanism - NANF
Core struts act as Fabry-Perot resonators– transmission windows open
between resonances given by:
1
2𝑡𝑡
𝜆𝜆𝑗𝑗 = 𝑗𝑗
𝑛𝑛𝑔𝑔2 − 1 2
j = 1,2,3.. , t = thickness
ng = glass refractive index

Pros:
• Lower mode overlap with glass, 10-5
• Larger mode field diameter
• Huge bandwidth
• Large range of core sizes
Cons:
• Higher loss (so far)
• Higher bend sensitivity

F. Poletti, Opt Express 22, 23807 (2014)

26
Loss in NANF Over Past 5 Years 3
10
BUT 2018 ORC 2018 ORC 2019

2
10
Loss in NANF (dB/km)

1
10
Wheeler et al. OFC2012 PDP

0
10
Bath 2015
Glass core SMF Mangan et al. OFC2004 PDP
-1
10

OFC 2020
-2 FORC 2016
10
2014 2015 2016 2017 2018 2019 2020 2021
Year

27
NANF Beating HC-PBGF 2.4

9.6 µm 24.9 µm 2.2

90nm <2 dB/km

2

1.8

Attenuation (dB/km)
1.6 65nm <1.5 dB/km
1.4

1.3 ± 0.1 dB/km 1.2 Cutback

40nm <1.4 dB/km Cutback error

1
1400 1410 1420 1430 1440 1450 1460 1470 1480 1490 1500

Wavelength (nm)
-25

30.4 µm -30
OTDR @1550nm 10
1

-35
0

505 m
10

-40

Power ( dBm)

Attenuation / dB/km
-45
-1
10

-50 Modelling
Confinement Microbend Scattering

-55 Total Simulated Measured

-2
10
0 100 200 300 400 500
800 1000 1200 1400 1600 1800 2000 2200

Length (m) Wavelength / nm

28
NANF – Below 1 dB/km
C L
~13.5 µm ~26.5 µm

120nm

37.2 µm

First sub 1dB/km HCF

0.65 dB/km +/- 0.08 dB/km

between 1520 and 1640nm
State of the art developments

4.0 µm
We have:
1. Improved the fabrication process and
reduced asymmetries in the stacked
preform
34.5 µm
2. Reduced core size to 34.5 μm to reduce
microbending
3. Reduced average azimuthal gap to 4 μm
to maintain low leakage loss

30
 Record Low Loss NANF
White light; launched through SMF and a Mode Field Adaptor 0.5
-55 1 C L
0.4
-60 500 m 0.8

Loss (dB/km)
90 nm
-65 0.6 0.3

Loss (dB/km)
Power (dBm)

-70 0.4
0.8

0.2
1700 m 0.6

1450 1500 1550 1600 1650

-75 0.4

0.2
0.2 Wavelength (nm)
0
1300 1350 1400 1450 1500

Wavelength (nm)

-80 0 0.28 dB/km at 1550 nm

1000 1100 1200 1300 1400 1500 1600 1700

Wavelength (nm) 0.28 dB/km +/- 0.04 dB/km from

1510 nm to 1600 nm
≤0.3 dB/km in C and L bands
 Modelling
0.051
dB/km

0
10

Loss (dB/km)
0.046
dB/km
-1 Measured loss Confinement Loss
10
Simulated total Loss Microbend Loss

0.036
dB/km

0.04
-2
dB/km 10
1200 1300 1400 1500 1600 1700
• Small and uniform gaps prevent leakage Wavelength (nm)
• Leakage alone accounts for 0.26 dB/km
Surface scattering limit may be lower than previously thought!
Fibre Fabrication Overview
• Hollow Core Photonic Bandgap Fibre
developed over a number of years.

• From 2018 to 2020 NANF loss has been

reduced from 1.3 dB/km to 0.28 dB/km

• Loss in NANF is now only 2x higher than

pure silica SMF
Further developments are within sight
Applications of HCFs Fibre Optics Sensors
Telecoms

High Power Lasers Nonlinear Optics

Medical Imaging
 Integration of HCFs
• Fusion Splicing
• Hermetic Seal
• Packaged Micro Optics
• Can create functional devices
• Photonic Lightwave Circuits
• Extremely low loss Yongmin Jung, Hyuntai Kim, Yong Chen, Thomas D. Bradley, Ian A. Davidson, John R. Hayes, Gregory Jasion, Hesham Sakr, Shuichiro Rikimi,
• Not hermetically sealed Francesco Poletti, and David J. Richardson, "Compact micro-optic based components for hollow core fibers," Opt. Express 28, 1518-1525 (2020)

• Complex Alignment

1) D. Suslov, M. Komanec, S. Zvánovec, T. Bradley, F. Poletti, D. J. Richardson, and R. Slavík, "Highly-efficient and low return-loss coupling of standard and antiresonant hollow-core fibers," in Frontiers in Optics
+ Laser Science APS/DLS, OSA Technical Digest (Optical Society of America, 2019), paper FW5B.2
2) M. Komanec et al., "Low-Loss and Low-Back-Reflection Hollow-Core to Standard Fiber Interconnection," in IEEE Photonics Technology Letters, vol. 31, no. 10, pp. 723-726, 15 May15, 2019, doi:
10.1109/LPT.2019.2902635.
 Nonlinear Optics in HCFs Raman Side Band Generation
• Tight Mode Confinement
• Long Interaction Length
• Strong Light Matter Interaction

Benoît, A.; Beaudou, B.; Alharbi, M.; Debord, B.; Gérôme, F.; Salin, F.; Benabid, F.
Over-five octaves wide Raman combs in high-power picosecond-laser pumped
H2-filled inhibited coupling Kagome fiber. Opt. Express 2015, 23, 14002–14009
UV Light Generation

Ka Fai Mak, John C. Travers, Philipp Hölzer, Nicolas Y. Joly, and Philip St. J. Russell,
F. BENABID, J. C. KNIGHT, G. ANTONOPOULOS, P. ST. J. RUSSELL, “ Stimulated Raman Scattering in Hydrogen- "Tunable vacuum-UV to visible ultrafast pulse source based on gas-filled Kagome-
Filled Hollow-Core Photonic Crystal Fiber”, SCIENCE11 OCT 2002 : 399-402 PCF," Opt. Express 21, 10942-10953 (2013)
 Laser Power Delivery
CW Laser Delivery
• Cutting/Engraving/Machining/Additive
Manufacturing
• Automated
• Precise
• Repeatable
• Tailorable

M. Michieletto, J. Lyngsø, C. Jakobsen, J. Lægsgaard, O. Bang, and T. Alkeskjold,

https://www.spilasers.com/application-
"Hollow-core fibers for high power pulse delivery," Opt. Express 24, 7103-7119
cutting/fiber-laser-cutting-of-stents/
(2016).
 Guiding UV & Mid IR Light 20

Norm. Transmission [dB] Absorbance [dB]

CH in HC-PBGF
4
1)15 Fei Yu, Maria
HITRAN DataCann, Adam Brunton, William
• Silica Low Loss 10
Wadsworth, and Jonathan Knight, "Single-mode
solarization-free hollow-core fiber for ultraviolet
Transmission ~ 5
pulse delivery," Opt. Express 26, 10879-10887
200 – 2000nm 0 (2018) 5m
2)-5 Shou-Fei Gao, Ying-Ying Wang, Wei Ding,
and Pu Wang, "Hollow-core negative-
-10

-15
curvature fiber for UV guidance," Opt.
Lett. 43, 1347-1350 (2018) 58 m
-20
3.15 3.2 3.25 3.3 3.35 3.4
Wavelength [µm]

Yu et al, APL Phtn 4 (080803), 2019

 Thermal Sensitivity of HCFs Processed data
Measured data 140

Phase change, rad/m

b)
SMF-28
• Temperature
a) changes refractive index of 120
Phase change, rad/m

140 100
the120
optical fiber and itsSMF-28
length. 80
60
100 40 Slope:
20 45.4 rad/m/oC
80 0
40.0 40.5 41.0 41.5 42.0 42.5 43.0
60 o
7
c) Tempearature, C

Phase change, rad/m

HCF
Refractive index
40 6
5
change20 HC-PBGF 4
3
0
0 200 400 600 800 1000 1200 1400 1600 2 Slope:
1 2.4 rad/m/oC
• This in turn changes the phase
Time, sec of the light at the fibre output as well as 0
40.0 40.5 41.0 41.5 42.0 42.5 43.0
HCF has a sensitivity 18.5 times smaller than SMF-28
its arrival time.
o
Tempearature, C
Slavík et al., Scientific reports 5, 15447, 2015.
 Anti-resonant HCF sensitivity &
Fabry Perot etalon
5-m long, 0.9-dB/km, NANF (ARF) fibre FP

Thermal sensitivity: Comparison with SMF

14
12
/°C

Freq. shift, GHz

• Two FPs (HCF & SMF) with similar Free Spectral Range (FSR): 5 m z
HCF & 3.3 m SMF28 10
. 6GH
8 : 1
.3 m
• HCF FP: Finesse of 130 6 P, 3
4 F-F
• NANF (ARF-type) 21 times less thermally sensitive than SMF (per SM
unit length; corresponding to 14.5 times per unit delay/ FSR). 2 HCF-FP, 5.0 m: 110MHz/°C
M. Ding et al., J of Lightwave Technol. 2020 (early access) 0
0 1 2 3 4 5 6 7 8 9 10
Temperature change, °C
 Polarisation in Anti-resonant HCFs
7-Tube ARF
PM - PBGF
PM Solid-Core Fibre PM Hollow-Core Fibre
10-5

10-6
ORC Nodeless
Polarization coupling, h (1/m) 10-7
Antiresonant Fibres

10-8 Anisotropic Rayleigh

scattering limit for SMF-28
10-9

10-10

https://www.nature.com/articles/s41566-020-0633-x
10-11
0.1 1 10 100 1000 104 105 106 Commercial PM
6-NANFs Group birefringence, β (rad/m) (Bow-Tie)
Material-free propagation in ARFs provides…
• Pure, interaction-free propagation for long interferometers & resonators
• Ultra-low polarization coupling despite minimal birefringence
41
 Data transmission in HCF: why?
• longest transmission in any HCF: 618 km with
• Ultimate low latency: 1.54 μs/km lower latency PM-QPSK
than SMF • previous record was 341km but only center channel
• Low Nonlinearity: can launch higher signal measured
power
• Wide Bandwidth: more than C band
100 turns, 772km
• Potential for ultra low loss….
0.20 GMI = 3.55 b/symb
No fibre Poletti et al., Nat. Photonics 7 279 2013
0.15
260 m
Output, a.u.

0.10 260 m NZ-DSF

HC-PBGF
0.05 1270 ns
Nespola et al, OFC 2020, PDP TH4B.5, 2020
870 ns
0.00
-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Time (µs)
Record distance transmission in HCFs
1000

800
OFC 2020 PDP Th4B.5
HCF transmission distance

600

400
NANF ECOC 2019 PDP Th1F.5
(km)

PBGF
200

0
2004 2006 2008 2010 2012 2014 2016 2018 2020
Year

43
Outlook
• Huge range of exciting applications
• Nonlinear Optics
• Telecoms
• …..
• Splicing/Interconnection
• Fusion splicing
• Packaged micro optics
• Rapid development in NANF
technology
• now only 2x higher than pure silica SMF
Further developments are within sight
Acknowledgements