Centrifugal Separations in

Biotechnology 2nd Edition Wallace

Woon Fong Leung
Visit to download the full and correct content document:
https://textbookfull.com/product/centrifugal-separations-in-biotechnology-2nd-edition-
wallace-woon-fong-leung/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Modern Industrial Microbiology and Biotechnology 2nd

Edition Nduka Okafor

https://textbookfull.com/product/modern-industrial-microbiology-
and-biotechnology-2nd-edition-nduka-okafor/

Beginning Power Apps The Non Developer s Guide to

Building Business Applications 2nd Edition Tim Leung

https://textbookfull.com/product/beginning-power-apps-the-non-
developer-s-guide-to-building-business-applications-2nd-edition-
tim-leung/

Beginning Power Apps: The Non-Developer's Guide to

Building Business Mobile Applications 2nd Edition Tim
Leung

https://textbookfull.com/product/beginning-power-apps-the-non-
developers-guide-to-building-business-mobile-applications-2nd-
edition-tim-leung/

An Introduction to Polysaccharide Biotechnology 2nd

Edition Stephen E. Harding

https://textbookfull.com/product/an-introduction-to-
polysaccharide-biotechnology-2nd-edition-stephen-e-harding/
Environmental biotechnology : principles and
applications 2nd Edition Perry L. Mccarty

https://textbookfull.com/product/environmental-biotechnology-
principles-and-applications-2nd-edition-perry-l-mccarty/

Membrane-Based Separations in Metallurgy: Principles

and Applications 1st Edition Lan Ying Jiang

https://textbookfull.com/product/membrane-based-separations-in-
metallurgy-principles-and-applications-1st-edition-lan-ying-
jiang/

Transferring Information Literacy Practices Billy Tak

Hoi Leung

https://textbookfull.com/product/transferring-information-
literacy-practices-billy-tak-hoi-leung/

Wallace at Bay Wallace of the Secret Service Book 7

3rd Edition Wilson

https://textbookfull.com/product/wallace-at-bay-wallace-of-the-
secret-service-book-7-3rd-edition-wilson/

Separation Processes in Biotechnology First Edition

Asenjo

https://textbookfull.com/product/separation-processes-in-
biotechnology-first-edition-asenjo/
Centrifugal Separations in
Biotechnology
Centrifugal Separations in
Biotechnology

Second Edition

Wallace Woon-Fong Leung

Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
Copyright © 2020 Elsevier Ltd. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage
and retrieval system, without permission in writing from the publisher. Details on how to
seek permission, further information about the Publisher’s permissions policies and our
arrangements with organizations such as the Copyright Clearance Center and the
Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright
by the Publisher (other than as may be noted herein).

Notices
Knowledge and best practice in this field are constantly changing. As new research and
experience broaden our understanding, changes in research methods, professional practices,
or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in
evaluating and using any information, methods, compounds, or experiments described herein.
In using such information or methods they should be mindful of their own safety and the
safety of others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors,
assume any liability for any injury and/or damage to persons or property as a matter of
products liability, negligence or otherwise, or from any use or operation of any methods,
products, instructions, or ideas contained in the material herein.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress
ISBN: 978-0-08-102634-2

For Information on all Elsevier publications

visit our website at https://www.elsevier.com/books-and-journals

Publisher: Susan Dennis

Acquisitions Editor: Kostas Marinakis
Editorial Project Manager: Sara Pianavilla
Production Project Manager: Joy Christel
Neumarin Honest Thangiah
Cover Designer: Greg Harris
Typeset by MPS Limited, Chennai, India
 Contents

In God, I trust xvii

Preface to Second Edition (2019) xix
Preface to First Edition (2007) xxiii

1 Introduction 1
1.1 Introduction 1
1.1.1 Common Host Cells for Secreting Recombinant
Protein 6
1.1.2 Platform for Protein Expression 8
1.1.3 Extracellular Protein 9
1.1.4 Intracellular Protein—Liquid 9
1.1.5 Intracellular Protein—Inclusion Body 10
1.2 Centrifugal Separation and Filtration 12
1.2.1 Sedimenting Centrifuge 14
1.2.2 Filtering Centrifuges 15
1.3 Pros and Cons of Filtration Versus Centrifugation 16
1.4 Generic Flow Sheet for Biopharmaceutical Process 18
1.5 Other Centrifugal Separations 19
1.6 Inputs and Outputs of Centrifuge 19
1.7 Separation Metrics 20
1.7.1 Protein Yield 20
1.7.2 Centrate Suspended Solids 20
1.7.3 Throughput Rate 21
1.7.4 Cell Viability 21
1.8 Text Organization 22
1.9 Summary 23
References 23
Problems 25

v
vi CONTENTS

2 Principles of Centrifugal Sedimentation 27

2.1 Introduction 27
2.2 Nonintuitive Phenomena 27
2.2.1 Pressure Gradient in a Fluid Under Centrifugal
Acceleration 27
2.2.2 Combined Centrifugal and Gravitational
Accelerations 28
2.2.3 Coriolis Effect 32
2.3 Intuitive Phenomena 35
2.3.1 Centrifugal Acceleration 35
2.3.2 Fluid in a Centrifuge Bowl Not at Solid-Body
Rotation 38
2.3.3 Regimes of Sedimentation 40
2.3.4 Stokes’ Law 42
2.3.5 Settling With Concentrated Solids 43
2.4 Process Functions 44
2.5 Summary 46
References 46
Problems 47

3 Batch and Semibatch Centrifuges 49

3.1 Spintube 49
3.2 Centrifugal Filter 53
3.3 Ultracentrifuges 54
3.3.1 Analytical Ultracentrifuge 55
3.3.2 Preparative Ultracentrifuge 56
3.3.3 Centrifugal Elutriation 58
3.4 Tubular Centrifuge 60
3.4.1 General Tubular Bowl Geometry 60
3.4.2 Ribs and Solids Scraper 63
3.4.3 Automatic Plunger Cake Discharge 67
3.5 Summary 69
References 70
Problems 70

4 Disk Centrifuge 73
4.1 Lamella/Inclined Plate Settler 73
4.1.1 Inclined Plate Settler Principle 73
4.1.2 Complications in Inclined Plate Settler 74
CONTENTS vii

4.2 Disk-Stack Centrifuge 75

4.2.1 General Disk Geometry 75
4.2.2 Disk Angle 77
4.2.3 Disk Spacing 77
4.2.4 Process Functions of Disk Centrifuge 79
4.2.5 Feed Solids 80
4.2.6 Manual Disk Centrifuge 80
4.2.7 Intermittent Discharge 82
4.2.8 Chamber Bowl 87
4.2.9 Continuous Concentrate Discharge 87
4.2.10 Liquid Discharge 95
4.2.11 Solution to Adverse Heating Effect 100
4.3 Feed Inlet and Accelerator 100
4.3.1 Introduction to Low Shear 101
4.3.2 Hydro-Hermetic Feed Design 101
4.3.3 Power Loss 102
4.3.4 Feed Acceleration Visual and Quantitative
Testing 104
4.3.5 Improved Feed Accelerator 108
4.4 Other Considerations 111
4.4.1 Materials of Construction 111
4.4.2 Clean-in-Place 112
4.4.3 Sterilization-in-Place 113
4.4.4 Containment 114
4.4.5 Surface Finish 114
4.4.6 Temperature Control 115
4.4.7 Water Requirements 115
4.4.8 Noise Level 115
4.4.9 Explosion Proof Design 115
4.5 Examples of Commercial Disk-Stack Centrifuge 115
4.6 Summary 119
References 119
Problems 120

5 Decanter Centrifuge 121

5.1 Solid Bowl or Decanter Centrifuge 121
5.2 Feed Rate 122
5.3 Pool Depth 122
5.4 Rotation Speed and G-Force 123
viii CONTENTS

5.5 Differential Speed 124

5.6 Sedimentation Enhancement Using Chemical 127
5.7 Three-Phase Separation 127
5.8 Cake Conveyance 129
5.8.1 Dry Beach 129
5.8.2 Hydraulic Assist 130
5.9 Summary 132
References 132
Problems 133

6 Commercial Applications of Centrifugation in

Biotechnology 135
6.1 Generic Flow Sheet of Biopharmaceutical 136
6.2 Mammalian Cell 138
6.3 Yeast Processing 141
6.4 Hormones Processing 144
6.5 Insulin Production 145
6.6 Biotech Separation of Inclusion Bodies 145
6.7 Vaccines Processing 147
6.7.1 Concentrated Cell-Based Product 147
6.7.2 Serum Product 148
6.8 Enzymes Processing 148
6.8.1 Extracellular Enzymes 148
6.8.2 Intracellular Enzymes 149
6.9 Probiotic Processing 150
6.10 Aquaculture 151
6.11 Alternative Meat 152
6.12 Baker Yeast Processing 153
6.13 Omega-3 From Microalgae 154
6.14 Ethanol Production 155
6.15 Other Biotech Processing 157
6.15.1 Recovery of Coagulation Factors From
Blood Plasma 157
6.15.2 Tissue From Animal Cells 157
6.15.3 Laboratory Concentration and Buffer
Exchange Using Centrifugal Filter 158
6.16 Summary 159
References 159
Problems 160
CONTENTS ix

7 Concentrating Solids by Centrifugation 161

7.1 Introduction 161
7.2 Concentrating Underflow 161
7.3 Compaction 163
7.4 Expression or Percolation 164
7.5 Compaction Testing 166
7.6 Compaction Pressure 167
7.6.1 Test-Tube Compaction 168
7.6.2 Decanter Compaction 168
7.6.3 Considerations of Cake Compaction 170
7.7 Recommendations for Increasing Solid Concentration
in Underflow 170
7.8 Summary 171
References 171
Problems 172

8 Laboratory and Pilot Testing 173

8.1 Process Objectives 173
8.2 Solid, Liquid, and Suspension Properties 174
8.2.1 Solids Properties 174
8.2.2 Mother Liquid Properties 175
8.2.3 Feed Slurry Properties 175
8.3 Bench-Scale Testing 175
8.3.1 Separability 175
8.3.2 Flocculant and Coagulant in Bench Tests 176
8.3.3 Test Variables 177
8.3.4 Material Balance 177
8.3.5 Acceleration and Deceleration Time Duration 180
8.3.6 Settling Velocity 180
8.4 Centrifugal Filter Testing 185
8.4.1 Steady-State Membrane Centrifugal Filtration
to Determine Protein Diffusivity and Solubility 185
8.4.2 Transient Membrane Centrifugal Filtration to
Determine Protein Osmotic Pressure and
Membrane Resistance 186
8.5 Pilot Testing 187
8.5.1 Material Balance Consideration for Pilot/
Production Scale 188
8.5.2 Product (Protein) Yield 190
8.5.3 Pilot Test Factors 192
8.6 Summary 199
References 199
Problems 200
x CONTENTS

9 Selection and Sizing of Centrifuges 203

9.1 Selection 203
9.1.1 Introduction 203
9.1.2 Tubular Centrifuge Selection 204
9.1.3 Disk Centrifuge Selection 204
9.1.4 Centrifuge Comparison 205
9.2 Centrifuge Sizing 207
9.2.1 Sizes and Rates 207
9.2.2 Dimensionless Le Number 208
9.2.3 Spintube (Bottle) Centrifuge 210
9.2.4 Sizing for Disk Centrifuge 213
9.2.5 Sizing for Tubular, Chamber, and Decanter
Centrifuge 219
9.3 Feed Particle Size Distribution 222
9.4 Performance of Tubular Centrifuge 223
9.5 Summary 224
References 224
Further Reading 225
Problems 225

10 Troubleshoot and Optimization 229

10.1 Troubleshooting 229
10.1.1 Timescale of Occurrence 229
10.1.2 Mechanical or Process Problem 230
10.1.3 Process Problems 230
10.1.4 Mechanical Problem 233
10.2 Optimization 235
10.2.1 Separation Metrics 235
10.2.2 Monitored Variables 236
10.2.3 Controlled Variables 237
10.2.4 Simple Optimization Scheme 237
10.3 Summary 240
Problems 240

11 Visualization and Modeling of Flow and Separation in

Tubular Centrifuge 243
11.1 Flow Visualization 243
11.2 Improved Moving Layer Flow Model 248
11.3 Effect of Velocity Profile 251
11.4 Effect of Friction Within the Flow Layer 252
CONTENTS xi

11.5 Dimensionless Le Parameter 252

11.6 Quantitative Prediction 253
11.6.1 Total Solids Recovery in Cake 253
11.6.2 Total Solids Recovery in the Centrate 253
11.6.3 Particle Size Distribution of Supernatant/
Overflow 254
11.6.4 Cumulative Size Recovery 254
11.7 Sedimentation Tests 255
11.7.1 Experiments on Sedimentation in Rotating
Bowl Centrifuge 255
11.8 Summary 258
References 259
Problems 259

12 Disk-Stack Modeling 261

12.1 Disk Model 261
12.1.1 Continuum Phase 264
12.1.2 Dispersed Phase 264
12.2 Model Validation 268
12.3 Complications 270
12.4 Summary 271
References 271
Problems 272

13 Performance Projection of Centrifuges in Bioseparation 273

13.1 Disk Centrifuge 273
13.1.1 Baseline Case (400-mm Disk) 275
13.1.2 Effect of Fine Size Distribution
(400-mm Disk) 276
13.1.3 Effect of G-Force (580-mm disk) 277
13.1.4 Efficiency η in Le Number (580-mm Disk) 280
13.1.5 Disk Centrifuge for Yeast Processing
(500-mm Disk) 282
13.1.6 Disk Centrifuge for Inclusion Body
Separation (260-mm Disk) 284
13.1.7 Enzymes (580-mm Disk) 286
13.2 Tubular Centrifuge 288
13.2.1 High-G Tubular (150- and 300-mm Tubular) 289
13.2.2 Low-G Tubular (150- and 300-mm Tubular) 290
13.3 Decanter 292
xii CONTENTS

13.4 Spintube 294

13.5 Further Discussion on Numerical Simulations 296
13.6 Summary 297
References 297
Problems 297

14 Rotating Membrane in Bioseparation 299

14.1 Membrane 299
14.1.1 Osmotic Pressure Resistance 300
14.1.2 Gel Resistance 301
14.1.3 Membrane Fouling and Cake Formation 302
14.1.4 Two Scenarios on Rotation 302
14.2 Rotating Disk Membrane With Surface Parallel to
the G-Force 303
14.2.1 Dimensionless Numbers 304
14.2.2 Governing Equations and Solutions 306
14.2.3 Gel Concentration 310
14.2.4 Determining Diffusivity 311
14.2.5 Parametric Effects 313
14.3 Rotating Membrane With Membrane Perpendicular
to the G-Force 315
14.3.1 Spintube Equipped With Membrane
Module—Centrifugal Filter 316
14.3.2 Model on Swinging Bucket Equipped With
Ultrafiltration Membrane 320
14.3.3 Comparing Test Results With Predictions 323
14.4 Summary 328
References 329
Problems 329

15 Flocculation With Decanter Centrifuges 331

15.1 Introduction 331
15.1.1 Coagulation and Flocculation 331
15.1.2 Decanter Centrifuge 332
15.1.3 Problems 332
15.2 Monotonic Size Distribution Model 333
15.2.1 Moving Layer 334
15.2.2 Floc Model 336
15.2.3 Exponential Floc Size Distribution 336
15.2.4 Leung Number Calculation 338
15.2.5 Model Solution 339
CONTENTS xiii

15.3 Field Test 340

15.3.1 Two Decanter Tests at Wastewater
Treatment 340
15.3.2 Determining In Situ Floc Size 341
15.4 Prediction 346
15.5 Scale-Up 346
15.5.1 Le scale-Up 346
15.5.2 Sigma Scale-Up 348
15.5.3 G/g-Volume Scale-Up 348
15.5.4 Surface Area Scale-Up 348
15.6 Summary 350
References 350
Further Reading 351
Problems 351

16 Case Studies of Monotonic and Unimodal Size

Distribution Models 353
16.1 Introduction 353
16.2 Monotonic Model Equivalent for Disk-Stack and
Tubular Centrifuges 354
16.3 Disk-Stack Centrifuge for Processing Protein From
Mammalian Cell Culture 356
16.3.1 Low Cell Density Cell Culture 358
16.3.2 High Cell Density Cell Culture 361
16.3.3 Centrate Solids and Turbidity 363
16.4 Tubular Centrifuge for Separating E. coli Lysate 364
16.5 Tubular Centrifuge for Separating S. pneumoniae
Flocculate 366
16.6 Unimodal Size Distribution Model 367
16.6.1 Unflocculated Suspension 369
16.6.2 Flocculation in Disk-Stack Centrifuge 370
16.7 Comparing the Solids Recovery Between Monotonic
and Unimodal Size Distributions 375
16.8 Summary 377
References 378
Problems 378

17 Classifying Bimodal Particle Size Distribution and Case

Study of Inclusion Body Classification 381
17.1 Introduction 382
17.2 Mammalian Cells With Cell Debris 383
xiv CONTENTS

17.3 Processing Hybridoma Cell Broth 384

17.4 Bimodal Size Model 387
17.5 Application of Bimodal Model on Hybridoma Cell
Separation 388
17.6 Variation in Fine Fractions From Debris 390
17.7 Size of Whole Cells 391
17.7.1 Whole Cells Average Size 392
17.7.2 Size Range on Whole Cells 392
17.8 Effect of Smaller Size (The Debris) 394
17.9 Size Recoveries 398
17.9.1 Size Recovery of Small (S) Particles in
Centrate 399
17.9.2 Size Recovery of Large (L) Particles in
Centrate 400
17.9.3 Example on Classification 401
17.9.4 Smaller Size Fraction Further Apart From
Larger Size Fraction in Bimodal Feed 405
17.9.5 Smaller Size Fraction Closer to the Larger
Size Fraction in Bimodal Feed 407
17.10 Classification of Inclusion Bodies 408
17.10.1 Conventional Inclusion Bodies Processing 409
17.10.2 New Inclusion Bodies Processing 411
17.11 Centrifuges for Inclusion Bodies processing 413
17.11.1 Disk-Stack Centrifuge 413
17.11.2 Tubular Centrifuge 415
17.11.3 Spintube Centrifuge 417
17.12 Separation by Size and Density Difference 418
17.13 Summary 422
References 422
Problems 423

18 Integration of Unified Modeling With Practice in

Centrifugal Separations 425
18.1 Introduction 425
18.2 Unified Modeling to Centrifugal Separation 425
18.3 Applications of the Unified Separation Models 428
18.3.1 Analysis of Test Data 428
18.3.2 Prediction/Forecast 429
18.3.3 Guiding Testing 429
18.3.4 Optimization 429
CONTENTS xv

18.3.5 Troubleshooting 429

18.3.6 Scale-Up/Scale-Down 430
18.4 Integration of Unified Separation Models With
Practice 431
18.5 Summary 433

Appendix A: Nomenclature 435

Appendix B: Buckingham-π Analysis for Decanter and
Tubular, Disk-Stack, and Spintube Centrifuges 439
Appendix C: Centrate or Concentrate Discharge Through
Rotating Impeller 445
Appendix D: Answers to Problems in Chapters 2 17 449
Index 457
In God, I trust

xvii
 Preface to Second Edition (2019)

Since the Centrifugal Separation in Biotechnology (CSB) has been

published in 2007, it has well served the community of biotechnology
separation. I have been contacted by practitioners, researchers, engineers,
students, users, and equipment manufacturers in the biotechnology and
biopharmaceutical community, who told me that they have recommended
the text whole-heartedly to others as they find the contents in it are informa-
tive and helpful. They said CSB is the most comprehensive treatise in deal-
ing with centrifugal separation in biotechnology. There is a good balance of
practice and fundamentals on the subject matter in the text. The text is well
illustrated with tables, figures, sketches, and photographs, and is written in
layman’s language that is easy to understand. The readers also found the
questions at the end of each chapter both challenging and stimulating,
which help them to think and digest the contents that have been discussed.
Over a decade since the first publication of CSB, there has been an
exponential growth in biotech activities. This is especially for the world
of biopharma. As an example, there are much more therapeutic mono-
clonal antibodies that have been under various stages of testing than
before and they have been approved, in record numbers, by the
European Medicines Agencies and the United States Federal Drug
Administration. There is a popular demand from the biotechnology and
biopharmaceutical community for a revised edition of the text. After a
year and a half of preparation we finally come up with the second edi-
tion of the book. We have kept the same approach as our first edition,
again with good “balance” between practice and fundamentals, together
with the author’s 44 years of experience in separation and filtration from
both industry and academia injecting into this delicate balance. We have
updated the text with new technologies and new designs for both the
disk-stack and tubular centrifuges, which are the workhorse of the indus-
try. A decade ago, flocculation was unheard of for disk-stack centrifuge.
Given the bottom feed design which is available today, flocculation can
be implemented for disk-stack centrifuge, when process permits. In
some applications processing flowable concentrate containing biological
cells or bacteria, after separation the concentrate is discharged through
external nozzles. The impact of the concentrate discharge stream at high
speed from the nozzle disk-stack centrifuge results in destroying the
cells or bacteria and losing their probiotic function. Today this probiotic

xix
xx PREFACE TO SECOND EDITION (2019)

flowable concentrate stream can be routed to a small diameter near the

axis of the centrifuge, where the concentrate stream can be gently dis-
charged under pressure without being exposed to air causing oxidation
and being destroyed from high-speed impact. These are only a few of the
new technologies that are presented in the second edition. We have
expanded the discussion on clean-in-place (CIP) and sterilization-in-place
(SIP), which are very important practice for biotechnology and biophar-
maceutical to avoid cross-contamination of their products. Despite there is
no satisfactory standard to execute CIP and SIP, the text has discussed
good practice that can be used as a general guideline for practitioners.
Chapter 6, Commercial Applications of Centrifugation in Biotechnology,
discusses several typical flow sheets for biotech and biopharma processing,
and this has received very favorable response from the readers. We have
included a few more typical flow sheets in biotech and biopharma proces-
sings, and they are by no means exhaustive but serve as good examples.
Some of the flow sheets are related to separation of recombinant protein,
while others are for different applications. Not only have recombinant pro-
tein being used for diagnostic, therapeutics, and disease prevention, they
have been expanded into human food (beef, pork, and chicken made from
recombinant protein) and aquaculture. As medical scientists discover new
antigens that are specific to different types of cancer cells, more monoclo-
nal antibodies are also discovered for identifying the antigens associated
with these cancer cells so that human immune system can launch self-
defense against them. Despite each type of protein being secreted by the
common hosts (mammalian cells, microbial, yeast, virus infected insect
cells, etc.) are different, there are certain commonalities in the separation
and purification steps. CSB continues to use case studies throughout to
illustrate these commonalities so that biotech practitioner not only can use
them in generic form but tailor-make for their own process.
In the first edition, the author has developed basic models of separa-
tion for tubular/decanter and disk-stack centrifuges (Chapter 11:
Visualization and Modeling of Flow and Separation in Tubular
Centrifuge, and Chapter 12: Disk-Stack Modeling). These models have
been used to complement with case studies to discuss separation
(Chapter 13: Performance Projection of Centrifuges in Bioseparation).
One shortcoming is that the particle size distributions, which were used
in Chapter 13, Performance Projection of Centrifuges in Bioseparation,
were often lacking. To that end, some of the models being developed
may be handicapped in short of that measurement. In the second edition,
we have come up with three analytical forms of particle size distribution
for the feed to the separating centrifuge, which depend on two to five
 PREFACE TO SECOND EDITION (2019)

parameters. These analytical forms include monotonic size distribution,

unimodal size distribution, and bimodal size distribution. By modifying
the parameters, we can easily change the particle size of the feed and
investigate the sensitivity of the separation outcome from the centrifuge.
We have used these size distributions together with a unified approach
on separation using spintube, disk-stack, tubular, and decanter centrifuges
to study a lot more cases. This includes a very important application on
inferring the flocculated size in centrifuge (Chapter 15: Flocculation With
Decanter Centrifuges), which has been a culprit on scale-up/scale-down
problems. We have also used the monotonic size distribution to interpret
tests being conducted on disk-stack and tubular centrifuges (Chapter 16:
Case Studies of Monotonic and Unimodal Size Distribution Models) with
both dispersed and flocculated feeds, respectively, and to demonstrate the
benefits of flocculation with disk-stack centrifuge (Chapter 16: Case
Studies of Monotonic and Unimodal Size Distribution Models).
Next, we have adopted the bimodal model to classify biologics with
both a smaller and a larger size fraction (Chapter 17: Classifying Bimodal
Particle Size Distribution and Case Study of Inclusion Body Classification)
in the feed to the centrifuge. The size recovery of the smaller fraction, as
well as the larger fraction, in centrate and concentrate, respectively, have
been developed. One can fine-tune the classification in accordance to the
S-shaped size recovery curves to maximize recovery of smaller/larger size
fraction with minimum cross contamination. This can speed-up the devel-
opment or optimization process improving the purity of the products.
Finally, in Chapter 18, Integration of Unified Modeling With Practice
in Centrifugal Separation, we propose an integrated approach, irrespec-
tive of the type of centrifuges, to separation by sedimentation in using
known/measured, or estimated analytical representation of, particle size
distribution to tackle separation problem. The author has suggested inte-
grating the models into practices at all stages from laboratory testing, pilot
testing, and clinical manufacturing to full-scale production. This reassures
the reliability of the separation process despite frequently we have only
limited test results as a basis to make important decisionon the process.
Overall, the second edition not only provides an update to various
centrifugal technologies for biotechnology, but fills in a number of miss-
ing gaps. I believe this new edition of CSB will provide a more power-
ful resource for readers to tackle their centrifugal separation problem.
I sincerely hope that the text can help to push the biotech and biopharma
forward with good practice and advanced technologies in centrifugal
separation. Consequently, drugs substances and intermediates, diagnostic
and therapeutic protein-based drugs, and health supplements, based on
xxii PREFACE TO SECOND EDITION (2019)

recombinant protein and other biotech processes, becomes available in

shorter development time and become more affordable to the general
public.
Finally, I give thanks to all those who have provided valuable inputs
into this new edition. I also thank my wife, Stella, in giving me the
extra-time and space to devote in preparation of the new edition of CSB
that has taken much longer time than I originally estimated. The support
from the rest of my family, my mom, Jessica, Daniel, Jeffrey, and
Sandar has been tremendous. I am grateful to God for His keeping and
strengthening of me, without which I cannot complete this revision.
In Him, I trust.

Wallace Woon-Fong Leung

July 2019
 Preface to First Edition (2007)

In processing biological materials to produce high value-added

intermediates or finished products, no matter whether it is liquid or
solid, often involves a separation step, especially after the fermenter or
bioreactor. In one instance, the high-value solid products in a dilute con-
centration have to be separated from the waste liquid with spent cells
and debris; therefore it is essential to prevent loss of valuable solids in
the liquid stream. A further requirement is that contaminants have to be
washed from the solids. Alternatively, the high-value liquid containing
dissolved protein needs to be separated from the biomass and that the
liquid product should be free of solid particles to avoid downstream sepa-
ration and contamination of purification equipment, such as chromatogra-
phy column. The difficulty in carrying out separation step is that biosolids
do not filter well and often foul and blind the filter, such as microfiltration
and ultrafiltration membranes. Instead of filtration, separation by sedimen-
tation utilizing the density difference between the solid and the suspending
liquid can be employed. However, the density difference between bioso-
lids and liquid (typically water based) is very small, rendering the separa-
tion very slow and ineffective, especially under the Earth’s gravity. In
addition, if RNA and other protein materials are dissolved in the liquid,
the liquid phase can be very viscous, which further slows down sedimenta-
tion. Another difficulty is that the solids concentration in suspension for
processing is relatively dilute and requires equipment that has large volu-
metric capacity for handling the flow and process.
Centrifugation has proven to be a rather robust process for enhancing
settling by using thousands to almost millions of times the Earth’s gravita-
tional acceleration. In biopharmaceutical processing for producing a
recombinant therapeutic protein for antibiotics and drug substances from
yeast, microbial, and mammalian cells, such as the Chinese Hamster
Ovary cells, centrifuges have been widely used to perform separation,
classification of cell debris, concentration of suspension, and separation
and washing of solids, such as inclusion body or crystalline protein. No
doubt, given the escalating research activities in biotechnology, many new
sources of therapeutic proteins and other valuable biological materials
will be discovered and developed, and more stringent requirements are
demanded from separation/recovery and purification. There will be more

xxiii
xxiv PREFACE TO FIRST EDITION (2007)

growing needs of centrifugation in combination with other separations and

filtrations to perform the often overlooked, yet important, duty.
While there are many texts and reference books on bioseparation,
there is very little coverage on centrifugation. This book is the first ref-
erence book of its kind devoted to centrifugal separation in biotechnol-
ogy. It is an outgrowth of a series of seminars, short courses and
presentations that the author has delivered to biopharmaceutical compa-
nies all over the world. This new and challenging topic has received
excellent global reception, which is quite comforting and rewarding. The
contents of this book are also based on the author’s research and exten-
sive experiences, respectively, in practice, mentoring and lecturing on
the subject for over 20 years.
The book starts out with an introduction on the topic (Chapter 1) fol-
lowed by Chapter 2 on sedimentation, which is the key step of separation.
Subsequently, various batch (spintube centrifuge, ultracentrifuge) and
semi-batch (tubular centrifuge) centrifuges are discussed in Chapter 3. The
workhorse of the industrial separation process, disk-stack centrifuge, is
presented and discussed in detail in Chapter 4. Also, decanter centrifuge,
which is more applicable to high-solids feed under relatively lower centrif-
ugal acceleration, has also been included in Chapter 5. However, the dis-
cussion will be brief in favor of giving room to various other topics.
Commercial applications of centrifuges in biotechnology are discussed in
Chapter 6. This is perhaps one of the most interesting topics for practi-
tioners who are more concerned about where proven processes are and
how their new process may build on what is already known and practiced.
Despite there being lot of applications discussed in this chapter, unfortu-
nately there might be applications that have been inadvertently omitted,
given that the biotech applications are very diverse. Subsequently, we dis-
cuss in Chapter 7 the importance and practice of increasing, or at least
maintaining, high-solids concentration in the underflow stream of the cen-
trifuge. Laboratory and pilot testing and selection and sizing are essential
functions for establishing and implementing the biotech process and they
are discussed in Chapters 8 and 9, respectively. A new unified approach in
scale-up and prediction with use of a dimensionless Leung (Le) number is
introduced. The Le number works for all types of centrifuges, including
spintube, tubular, chamber, disk-stack, and decanter centrifuges. This pro-
vides a solid foundation for practitioners to scale-up equipment and analyze
test results. Troubleshooting and optimization are two important topics of
general interest, especially for installed machines, and these are discussed in
Chapter 10. Subsequently, modeling of tubular and disk-stack centrifuges
are covered in Chapters 11 and 12, respectively, for researchers who are
 PREFACE TO FIRST EDITION (2007)

interested. Readers who are not interested in modeling can jump directly
to Chapter 13. The Le number provides a basis for the scale-up and per-
formance prediction covered in Chapter 13. Here, numerous examples
are used to demonstrate the versatility of the numerical simulator built on
the Le-approach to forecast performance in parallel with concurrent test-
ing, which is often limited for various reasons. Numerical simulation can
also be used to analyze laboratory, pilot, and production test results to
validate machine and process performance. Therefore numerical simula-
tion can be used for laboratory screening, pilot testing, clinical
manufacturing testing, full-scale production testing, and even for small-
scale testing in the laboratory to investigate alternatives and improve-
ments of the existing process under production. Membrane processes,
such as microfiltration, ultrafiltration, and diafiltration, are frequently
used in bioseparation. Lastly, Chapter 14 is devoted to combining two
separation processes: centrifugation and membrane separation. Two
examples, respectively, on centrifugal filter in spintube and large rotating
membrane systems are discussed. The general approach can be extended
to other rotating membrane geometry.
Centrifugation has been treated as a black box in the past, as the sub-
ject is quite complex and nonintuitive. The subject involves multiple dis-
ciplines, such as fluid dynamics, mechanics and vibration, design,
material science, rheology, chemical and process engineering, chemistry,
biology, and physics. I hope this text will fulfill the quest of knowledge
rendering centrifuge a lot more “transparent” to biologists, biotechnolo-
gists, chemists, physicists, scientists, researchers, and practicing engi-
neers. The more they know the better they can deploy, comfortably and
without reservation, centrifuge as handy process equipment.
Problems are listed at the end of each chapter in the text, and they
complement and supplement the contents in the chapter. They are also
meant to reinforce the concept for the readers through practices, chal-
lenging their thoughts and understanding on the topic. Apart from practi-
tioners and researchers, this book is written primarily for 4th year
university student in the 4-year undergraduate study and research gradu-
ates in their MS or PhD research program in university taking biosepara-
tion, bioprocessing, advanced unit-operation/process engineering, or
similar courses.
I am grateful to Stella, Jessica, Jeffrey, my mother, and my late father
for putting up with me while I was devoted to preparing this book. Both
my late father and my dear friend and mentor, the late Professor Ascher
H. Shapiro, demonstrated dedication and perseverance in their lives,
which inspired me all along, especially during the trying times when I
xxvi PREFACE TO FIRST EDITION (2007)

was working on the manuscript among other responsibilities that also

demanded my undivided attention. I also thank Alice Tang for skillfully
helping out with the manuscript work and meeting the publisher’s
deadline.

Wallace Woon-Fong Leung

2007
1
Introduction

1.1 Introduction
Biotechnology has revolutionized our life in the 20th century [1]. Its
impact is only now being felt from engineering food [2 4], engineering
and delivering drugs [5,6], to engineering consumable products. Without
doubt it will continue to influence our daily lives for years to come. One
of the many successful examples is that drugs, such as monoclonal anti-
bodies (mAbs) and some basic drug substances (i.e., the building block
for various drugs), can now be manufactured and formulated from bior-
eaction. One of the commonly used methods in biotechnology is the
recombinant DNA technique [7 9]. A desired gene is isolated from one
organism, and this is inserted into a small piece of carrier DNA called a
vector. It is highly desirable that the recombined DNA (vector plus
gene) can propagate in a similar or unrelated host/recipient cell.
The mammalian cell, such as the Chinese Hamster Ovary cell, is a
popular host cell. Fig. 1.1 shows a schematic of an animal cell which is
very similar to that of a mammalian cell. A characteristic size of the
mammalian cell is about 10 20 µm. Unlike a plant cell, there is no cell
wall for animal and mammalian cells, so they rely on a plasma mem-
brane to keep the intracellular contents intact. High shear stress acting
on the cell can rupture the fragile membrane releasing the intracellular
material.
Yeast (see schematic in Fig. 1.2), in eukaryotic single-celled microor-
ganisms classified as a member of fungus kingdom, has been commonly
used as a host cell in the recombinant DNA process, the knowledge and
experience of which we have gained from the brewery industry. Unlike
a mammalian cell, the yeast cell has a strong cell wall. Yeast cells are
smaller than mammalian cells and are typically between 7 and 10 µm.
Some common yeast hosts include, Saccharomyces cerevisiae (referred
commonly as baker yeast) and Pichia pastoris.
Bacteria, such as Escherichia coli (hereafter abbreviated as E. coli)
and Bacillus subtilis, have been used as host cells for the recombinant

Centrifugal Separations in Biotechnology. DOI: https://doi.org/10.1016/B978-0-08-102634-2.00001-5

© 2020 Elsevier Ltd. All rights reserved. 1
Figure 1.1 Animal cell schematic showing plasma membrane.

Figure 1.2 Yeast cell schematic showing both cell wall and membrane.
 Introduction 3

Figure 1.3 Escherichia coli cell schematic.

DNA technique. A schematic of E. coli bacteria cell is shown in

Fig. 1.3. Again, E. coli has a sturdy cell wall with both an outer mem-
brane and an inner membrane. E. coli is typically elongated with a
dimension of 3 µm long by 1 µm width. Therapeutic protein can be
“expressed” by these host cells or organisms with the recombinant
DNA. The protein of interest may remain in the cell (intracellular) or be
secreted to the exterior of the cell (extracellular). The aforementioned
biosynthesis provides more engineering flexibility, specificity, versatil-
ity, reliability, and cost-effectiveness.
Therapeutic proteins are quite diverse in the application treatments,
such as human insulin for diabetes, erythropoietin for anemia and chronic
renal failure, interferon-beta and gamma for cancer, DNase for pulmonary
treatment, vaccines for hepatitis B, interleukin-2 for AIDS, prourokinase
for heart attacks, and tissue plasminogen activator (enzyme) for strokes.
Therapeutic proteins are present in many different kinds of mAbs.
mAbs are antibodies that are identical, because they are produced by
one type of immune cells, and they are all copies or clones of a single
parent cell. mAbs are first produced by Kohler and Milstein in 1975 [10],
for which they were awarded the Nobel Prize in Physiology or Medicine
in 1984. By virtue of the mAb being identical copies produced by one
type of immune cells, they have a high specificity for their targets. mAb
has been used in diagnosis. There are over 100 different diagnostic pro-
ducts available in the world that are mAb [11]. mAb is also used for
4 Centrifugal Separations in Biotechnology

therapeutics. For example, mAb has been a popular antibody made in the
laboratory used for cancer treatment where the antibody is designed to
attach as a label to their counterpart protein (antigen) on a specific cancer
cell so that immune cells can spot and attack the cancer cells. As an
example, the mAb known commercially as alemtuzumab drug (note all
mAb drugs have the last three alphabets labeled as “mab,” which distin-
guishes them being mAb-based drugs) can target at the antigen CD52
found on the cancer cells that causes chronic lymphocytic leukemia [12].
As antigens are discovered to be linked to more specific cancers, more
mAbs have also been developed for cancer treatment. Some mAbs work
better on certain cancers than others. mAb can also attach to the antigen
on breast cancer cells blocking the growth of breast cancer cells.
Cancer cells can “turn off the switch” of immune cells to avoid being
attacked by the immune system in our bodies. Inhibitors, or commonly
known as checkpoints, are mAb produced by the recombinant protein
process, that inhibit the protein secreted from the cancer cells in “fool-
ing” the immune cells, thereby allowing the immune cells to carry out
their normal functions. As an example, PD-1 is a protein on the immune
T-cells. The immune T-cells are normally in a switch-off condition
because PD-1 has been attached by their counterpart PD-L1, another
protein that both normal cells and cancer cells have. In other words,
they have been switched off. Some cancer cells have an abundant PD-
L1 that is used to attach to the PD-1 of the immune T-cells thereby
evading being attacked. On the other hand, mAb can target at either PD-
1 or PD-L1 and inhibit their binding, thereby allowing the immune cells
to attack the cancer cells. For example, pembrolizumab is a PD-1 inhibi-
tor that can treat skin melanoma, non-small-cell lung cancer, kidney can-
cer, bladder cancer, head and neck cancer, and Hodhkin lymphoma [12].
As another example, atezolizumab is a PD-L1 inhibitor that can treat
bladder cancer, non-small-cell lung cancer, and Merkel cell carcinoma
[12]. New inhibitors are being developed rapidly over time as more
knowledge is being gained on the specifics of different cancers and their
behavior. The examples mentioned in the forgoing are just a few under
the broad umbrella of immunotherapy for which mAb plays an impor-
tant role. The main objective of immunotherapy is to enable the immune
system of patients to recognize or target specific cancer cells and destroy
them [13]. In 2005 the total mAb therapeutics entering first-in-human
studies per year is 35, 16 out of which are for cancer treatment. In 2017
the total mAb therapeutics rose to 105, and nearly 80 were for cancer
treatment. Indeed, the antibody therapeutics entering clinical study and
being approved are in record numbers [14].
 Introduction 5

Extracellular proteins secreted from yeasts are produced for making

insulin, human serum albumin, and hepatitis vaccines. Insulin drug has
reached over USD 24 billion market in 2018 according to a market study
in 2019. The fast-growing biopharmaceutical business in producing ther-
apeutic proteins is getting so popular that all major drug manufacturers
also carry a parallel line of this business.
Unfortunately, the protein expressed from the bioprocess is in very
small amounts in a large volume of suspension, that is, low concentration.
The two key hurdles in recombinant DNA techniques to produce therapeu-
tic protein [15] are (a) to recover this small concentration of protein after
fermentation by separation and (2) to provide high purity of the protein
product through separation and purification. It is prudent that both separa-
tion and purification processes should be robust and cost-effective for the
biopharmaceutical technology to be viable and competitive. Although this
text is focused on separation, one should bear in mind that given these two
steps are sequential, poor separation can adversely affect purification
downstream. Therefore it is prudent to have an integrated approach for
downstream processing. To say the least, if there is an upset from the fer-
menter upstream producing, say, off-spec finer feed, the centrifuge should
take on the upset feed and try to produce a consistent output downstream
to the filter, membrane and chromatography column downstream in the
interim, while the upset condition is being fixed. Otherwise, the entire
chain of downstream processes can be seriously affected.
Other biotechnology involves synthesis and/or modification of inter-
mediates or final products. Frequently, this is in a suspension form so
that mechanical mixing, separation, spray or thermal drying, and other
allied processes are required.
Given separation is an important task [16 20] in biotechnology in
lieu of the above, it can be a very difficult task due to the low concen-
tration of the protein present and the large volume of liquid to handle,
the fragility of the cells, the presence of cell debris, fine particulates and
colloids, and the high viscosity due to dissolution of intracellular sub-
stances, such as RNA. Typically, separation can be achieved by filtration
and sedimentation. There are some specific problems relating to each as
discussed in the following.
Filtering a suspension containing biomass is quite tricky as the mate-
rial can foul the filter surface, reducing permeate or filtrate flow regard-
less whether the media is a microfiltration or an ultrafiltration
membrane. It is equally challenging to settle biomass as the density of
the biomass material is just slightly greater than that of the liquid phase,
which often is aqueous based. Given that settling rate is proportional to
Another random document with
no related content on Scribd:
The affinities with the order Stolonifera are clearly seen in the genera
Xenia and Telesto. Some species of Xenia form flattened or domed
colonies attached to stones or corals, with non-retractile anthocodiae
and body-walls united for only a short distance at the base. Young
Xenia colonies are in fact Stolonifera in all essential characters. In
Telesto prolifera we find a network of stolons encrusting coral
branches and other objects after the manner of the stolons of many
species of Clavularia, although the zooids do not arise from these
stolons singly, but in groups, with their body-walls fused together for
a certain distance. In Telesto rubra the spicules of the body-walls are
fused together to form a series of perforated tubes very similar in
some respects to the tubes of Tubipora.

A remarkable genus is Coelogorgia. Here we find a branching colony

arising from a basal stolon, and the axis of the main stem and of
each branch consists of a single very much elongated zooid bearing
on its thickened walls the branches of the next series and other
zooids. It is true that in this genus there is very little fusion of
neighbouring zooids, and the amount of true coenenchym is so small
that it can hardly be said to exist at all. Bourne[373] has united this
genus with Telesto into a family Asiphonacea, which he joins with the
Pennatulida in the order Stelechotokea; but their affinities seem to
be closer with the Alcyonacea than with the Pennatulacea, from
which they differ in many important characters.
 Fig. 154.—Alcyonium digitatum, a single-lobed specimen, with some of the
zooids expanded.

The genus Alcyonium not only contains the commonest British

Alcyonarian (A. digitatum), but it is one of the most widely distributed
genera of all Alcyonaria that occur in shallow water.

The genera Sarcophytum and Lobophytum occur in shallow water in

the tropics of the old world. The former frequently consists of huge
toad-stool shaped masses, soft and spongy in consistency, of a
green, brown, or yellow colour. On some reefs the colonies of
Sarcophytum form a very conspicuous feature, and from their very
slimy, slippery surface, add to the minor dangers of wading in these
regions. Both genera are dimorphic. Some species of the genus
Sclerophytum,[374] which occur in the Indian Ocean, are so hard and
brittle that they might readily be mistaken for a Zoantharian coral.
This character is due to the enormous number of tightly packed
spicules borne by the coenenchym. Some of these spicules in S.
querciforme are 7 mm. × 1.7 mm.; the largest, though not the longest
(vide p. 335) of any spicules occurring in the order.

Another very important genus occurring on coral reefs, and of very

wide distribution, is Spongodes. This genus forms bushy and rather
brittle colonies of an endless variety of beautiful shapes and colours.
Arising from the neck of each anthocodia there are one or two long,
sharp, projecting spicules, which give the surface a very spiny or
prickly character.

The genera Siphonogorgia and Chironephthya form large brittle,

branching colonies which might readily be mistaken for Gorgonians.
The strength of the branches, however, is mainly due to the large,
densely packed, spindle-shaped spicules at the surface of the
coenenchym, the long coelenteric cavities of the zooids penetrating
the axis of both stem and branches. Siphonogorgia is usually
uniformly red or yellow in colour. Chironephthya, on the other hand,
exhibits a great variety of colour in specimens from the same reef,
and indeed in different branches of the same colony.

Fam. 1. Xeniidae.—Alcyonacea with non-retractile zooids. Spicules

very small discs, usually containing a relatively small proportion of
lime.

Xenia, Savigny; Indian Ocean and Torres Straits. Heteroxenia,

Kölliker; Red Sea, Cape of Good Hope, and Torres Straits.

Fam. 2. Telestidae.—Colonies arising from an encrusting

membranous or branching stolon. The erect stem and branches are
formed by the body-walls of two or three zooids only, from which
secondary zooids and branches of the next order arise.

Telesto, Lamouroux, widely distributed in warm waters of the

Atlantic, Pacific, and Indian Oceans. The genus Fascicularia, Viguier,
from the coast of Algiers, seems to be related to Telesto, but the
groups of zooids are short, and do not give rise to branches.

Fam. 3. Coelogorgiidae.—The colony arborescent, attached by

stolon-like processes. The stem formed by an axial zooid with
thickened body-walls. Branches formed by axial zooids of the
second order, and branchlets by axial zooids of the third order, borne
either on two sides or in spirals by the main stem. Genus
Coelogorgia, Zanzibar.

Fam. 4. Alcyoniidae.—The colonies of this family are usually soft

and fleshy, and the spicules, evenly distributed throughout the
coenenchym, do not usually fuse or interlock to form a continuous
solid skeleton. They may be unbranched or lobed, never dendritic in
form. The principal genera are:—Alcyonium, Linnaeus,
cosmopolitan, but principally distributed in temperate and cold
waters. Alcyonium digitatum is the commonest British Alcyonarian. It
is found in shallow water, from the pools left at low spring tides to
depths of 40 or 50 fathoms, at most places on the British shores. It is
stated by Koehler to descend into depths of over 300 fathoms in the
Bay of Biscay. There are two principal varieties; one is white or pale
pink in the living condition, and the other yellow. In some localities
the two varieties may be found in the same pools. Another species,
Alcyonium glomeratum, placed in a distinct genus (Rhodophyton) by
Gray, and distinguished from the common species by its red colour
and long digitate lobes, is found only off the coast of Cornwall.
Paralcyonium, Milne Edwards; Mediterranean. Sclerophytum, Pratt;
sometimes dimorphic, Indian Ocean. Sarcophytum, Lesson;
dimorphic, principally tropical. Lolophytum, Marenzeller; dimorphic,
tropical. Anthomastus, Verrill; dimorphic, Atlantic Ocean, deep water.
Acrophytum, Hickson; dimorphic, Cape of Good Hope.

Fam. 5. Nephthyidae.—Colonies dendritic. Usually soft and flexible

in consistency. Nephthya, Savigny; Indian and Pacific Oceans.
Spongodes, Lesson; widely distributed in the Indian and Pacific
Oceans.

Fam. 6. Siphonogorgiidae.—Colonies often of considerable size.

Dendritic. Spicules usually large and abundant, giving a stiff, brittle
consistency to the stem and branches. Siphonogorgia, Kölliker; Red
Sea, Indian, and Pacific tropics. Chironephthya, Wright and Studer;
Indian and Pacific Oceans. Lemnalia, Gray; Zanzibar. Agaricoides,
Simpson;[375] Indian Ocean, 400 fathoms.

Order IV. Gorgonacea.

This order contains a very large number of dendritic and usually
flexible corals occurring in nearly all seas and extending from
shallow waters to the very great depths of the ocean. A large
proportion of them are brightly coloured, and as the principal
pigments are fixed in the spicules, and are therefore preserved when
the corals are dead and dried, they afford some of the most
attractive and graceful objects of a natural history museum.
The only character that separates them from the Alcyonacea is that
they possess a skeletal axis that is not perforated by the coelenteric
cavities of the zooids. The coelenteric cavities are usually short. The
order may conveniently be divided into two sub-orders.

Sub-Order 1. Pseudaxonia.
The axis in this sub-order consists of numerous spicules tightly
packed together, or cemented together by a substance which is
probably allied to horn in its chemical composition. This substance
may be considerable in amount, in which case it remains after
decalcification as a spongy, porous residue; or it may be so small in
amount, as in Corallium, that the axis appears to be composed of
solid carbonate of lime. The statement is usually made that the axis
is penetrated by nutritive canals in certain genera, but the evidence
upon which this is based is unsatisfactory and in some cases
unfounded. There can be no doubt, however, that in some genera
the axis is porous and in others it is not, and this forms a useful
character for the separation of genera.

Fam. 1. Briareidae.—The medullary substance consists of closely

packed but separate spicules embedded in a soft horny matrix,
which is uniform in character throughout its course. Nearly all the
genera form dendritic colonies of considerable size.

The principal genera are:—Solenocaulon, Gray; Indian Ocean and

North Australia. Many of the specimens of this genus have fistulose
stems and branches. The tubular character of the stem and
branches is probably caused by the activity of a Crustacean,
Alpheus, and may be regarded as of the nature of a gall-formation.
[376] Paragorgia, M. Edwards; Norwegian fjords, in deep water. This
genus forms very large tree-like colonies of a ruby-red or white
colour. It is perhaps the largest of the dendritic Alcyonarians. It is
dimorphic. Spongioderma, Kölliker; Cape of Good Hope. The surface
of this form is always covered by an encrusting sponge. Iciligorgia,
Ridley; Torres Straits. The stem and branches are compressed and
irregular in section.

Fam. 2. Sclerogorgiidae.—The medullary mass forms a distinct

axis consisting of closely packed elongate spicules with dense horny
sheaths.

Suberogorgia, Gray, has a wide distribution in the Pacific Ocean,

Indian Ocean, and the West Indies. Keroeides, W. and S., comes
from Japan.

Fam. 3. Melitodidae.—The axis in this family exhibits a series of

nodes and internodes (Fig. 155), the former consisting of pads
formed of a horny substance with embedded spicules, the latter of a
calcareous substance with only traces of a horny matrix. The
internodes are quite rigid, the nodes however give a certain degree
of flexibility to the colony as a whole. Neither the nodes nor the
internodes are penetrated by nutritive canals, but when dried the
nodes are porous.
 Fig. 155.—Melitodes dichotoma, showing the swollen nodes and the internodes.

The principal genera are:—Melitodes, Verrill; widely distributed in the

Indian and Pacific Oceans, Cape of Good Hope, etc. This genus is in
some localities extremely abundant and exhibits great brilliancy and
variety of colour. The branching is usually dichotomous at the nodes.
Wrightella, Gray. This is a delicate dwarf form from Mauritius and the
coast of South Africa. Parisis, Verrill; Pacific Ocean from Formosa to
Australia but not very common. One species from Mauritius. The
branches arise from the internodes.

Fam. 4. Coralliidae.—The axis is formed by the fusion of spicules

into a dense, solid, inflexible, calcareous core.

Corallium, Lamarck. Corallium nobile, Pallas, the "precious coral,"

occurs in the Mediterranean, chiefly off the coast of North Africa, but
also on the coasts of Italy, Corsica, Sardinia, and it extends to the
Cape Verde Islands in the Atlantic Ocean. C. japonicum, Kishinouye,
called Akasango by the fishermen, occurs off the coast of Japan,
and C. reginae, Hickson, has recently been described from deep
water off the coast of Timor.[377] The genus Pleurocorallium, Gray, is
regarded by some authors as distinct, but the characters that are
supposed to distinguish it, namely, the presence of peculiar "opera-
glass-shaped spicules," and the occurrence of the verrucae on one
side of the branches only, are not very satisfactory. The following
species are therefore placed by Kishinouye[378] in the genus
Corallium:—C. elatius, Ridley (Momoirosango); C. konojoi,
Kishinouye (Shirosango); C. boshuensis, K.; C. sulcatum, K.; C.
inutile, K.; and C. pusillum, K.,—all from the coast of Japan. Of the
coral obtained from these species, the best kinds of Momoirosango
vary in price from £30 per pound downwards according to the quality.
The Shirosango is the least valuable of the kinds that are brought
into the market, and is rarely exported.[379] Three species of
Corallium (Pleurocorallium) have been described from Madeira,[380]
and one of these, C. johnsoni, has recently been found in 388
fathoms off the coast of Ireland.[381] Other species are C.
stylasteroides, from Mauritius; C. confusum, Moroff,[382] from
Sagami Bay in Japan; and an undescribed species obtained by the
"Siboga," off Djilolo. These corals range from shallow water to
depths of 300-500 fathoms. Pleurocoralloides, Moroff, differs from
the others in having very prominent verrucae and in the character of
the large spindle-shaped and scale-like spicules. It was found in
Sagami Bay, Japan. Specimens attributed to the genus
Pleurocorallium have been found fossil in the white chalk of France,
but Corallium has been found only in the tertiaries.[383]

Sub-Order 2. Axifera.
The axis in this sub-order may be horny, or horny with a core of
calcium carbonate, or composed of horn impregnated with calcium
carbonate, or of nodes of horn alternating with internodes of calcium
carbonate. It may be distinguished from the axis of the Pseudaxonia
by the fact that in no case have definite spicules been observed to
take part in its formation. It has been suggested that as the Axifera
represent a line of descent distinct from that of the Pseudaxonia they
should be placed in a separate order. Apart from the character of the
axis, however, the two sub-orders show so many affinities in their
general anatomy that it is better to regard the two lines of descent as
united within the Gorgonacean limit. It is very improbable that the
two groups sprang independently from a stoloniferous ancestor.
Fam. 1. Isidae.—This family includes all those Axifera in which the
axis is composed of alternate nodes of horn and internodes of
calcareous substance.

There can be little doubt of the close affinities of many of the genera
of this family with the Melitodidae among the Pseudaxonia. In both
the coenenchym is thin and the coelenteric cavities short. No
important differences have been observed between the structure of
the zooids of the two families, and now that we know that the
"nutritive canals" of Melitodes do not perforate the nodes there is no
important difference left between the coenosarcal canal systems.
The structure and method of calcification of the internodes of the two
families are very similar. The main difference between them is that
the nodes of the Isidae are purely horny, whereas in the Melitodidae
the horny substance of the nodes contains calcareous spicules.

The principal genera are:—Isis, Linnaeus; Pacific Ocean. This genus

forms substantial fan-shaped colonies with, relatively, a thick
coenenchym, short stout internodes and black horny nodes.
Mopsea, Lamouroux; Coast of Australia. The verrucae are club-
shaped and are arranged in spiral rows round the stem. Acanella,
Gray; principally found in deep water in the Atlantic Ocean but also
in the Pacific. The internodes are long and the branches arise from
the nodes. Most of the species occur in deep water, some in very
deep water (A. simplex, 1600 to 1700 fathoms). In this and the
following genera the coenenchym is thin and the zooids imperfectly
or not retractile. Ceratoisis, Wright; Atlantic Ocean, extending from
shallow to deep water. The branches arise from the nodes.
Chelidonisis, Studer; deep water off the Azores. Isidella, Gray;
Mediterranean Sea. Bathygorgia, Wright; off Yokohama, 2300
fathoms. This genus is unbranched, with very long internodes and
short nodes. The zooids are arranged on one side only of the stem.

Fam. 2. Primnoidae.—This is a well-marked family. The axis of the

colonies is horny and calcareous. The coenenchym and the non-
retractile zooids are protected by scale-like spicules, which usually
overlap and form a complete armour for the protection of the soft
parts. On the aboral side of the base of each tentacle there is a
specialised scale, and these fit together, when the tentacles are
folded over the peristome, to form an operculum.

The principal genera are:—Primnoa, Lamouroux; Atlantic Ocean,

occurring also in the Norwegian fjords. This genus is usually found in
moderately deep water, 100 to 500 fathoms. Primnoella, Gray. This
genus seems to be confined to the temperate seas of the southern
hemisphere. It is unbranched. The zooids are arranged in whorls
round the long whip-like stem. Plumarella, Gray; southern
hemisphere, in moderately deep water. This is branched pinnately in
one plane. The zooids are small and arise at considerable intervals
alternately on the sides of the branches. Stenella, Gray; widely
distributed in deep water. The zooids are large and are arranged in
whorls of three situated at considerable distances apart. Stachyodes,
W. and S.; Fiji, Kermadecs, Azores, in deep water. Colony feebly
branched. Zooids in regular whorls of five. Other genera belonging to
this group of Primnoidae are Thouarella, Gray, and Amphilaphis,
Antarctic seas.

The following genera are placed in separate sub-families:—

Callozostron, Wright; Antarctic Sea, 1670 fathoms. The axis is
procumbent and the zooids are thickly set in rows on its upper
surface. The zooids are protected by large imbricate scales, of which
those of the last row are continued into long spine-like processes.
Calyptrophora, Gray; Pacific Ocean, in deep water. The base of the
zooids is protected by two remarkably large scales. Primnoides, W.
and S.; Southern Ocean. The opercular scales are not distinctly
differentiated and the calyx is therefore imperfectly protected.

Fam. 3. Chrysogorgiidae.[384]—The axis in this family is composed

of a horny fibrous substance with interstratified calcareous particles,
and it springs from a calcareous plate, which sometimes gives off
root-like processes. It may be unbranched or branched in such a
way that the branches of the second, third, and subsequent orders
assume in turn the direction of the base of the main axis. The axis is
frequently of a metallic iridescent appearance. The zooids usually
arise in a single straight or spiral row on the branches, and are not
retractile. The coenenchym is thin. The spicules vary considerably,
but in a very large proportion of the species they are thin, oval, or
hour-glass plates (Fig. 149, 10, p. 336).

By some authors this family is considered to be the simplest and

most primitive of the Axifera; but the delicate character of the axis of
the main stem and branches, the thinness of the coenenchym, the
position of the zooids on one side of the branches only, and the
tenuity of the calcareous spicules may be all accounted not as
primitive characters, but as special adaptations to the life in the slow
uniform currents of deep water.

The principal genera are:—Lepidogorgia, Verrill; Atlantic and Pacific

Oceans, 300 to 1600 fathoms. Axis unbranched. Zooids large and
arranged in a single row. Trichogorgia, Hickson; Cape of Good Hope,
56 fathoms. Colony branching in one plane. Zooids numerous and
on all sides of the branches. Chrysogorgia, D. and M.; deep water.
Axis branched. Spicules on the zooids always large. Metallogorgia,
Versluys; Atlantic Ocean, 400 to 900 fathoms. Basal part of the stem
unbranched (monopodial). Iridogorgia, Verrill. Spiral stem and
branches. Pleurogorgia, Versluys. Axis branched in one plane.
Coenenchym thick. Riisea, D. and M. Monopodial stem and thick
coenenchym.

Fam. 4. Muriceidae.—This is a large family, exhibiting very great

variety of habit. The spicules are often very spiny, and project
beyond the surface of the ectoderm, giving the colony a rough
appearance. A great number of genera have been described, but
none of them are very well known. The family requires careful
revision.

The more important genera are:—Acanthogorgia, Gray; principally in

deep water in the Atlantic Ocean. The calices are large, cylindrical,
and spiny. Villogorgia, D. and M.; widely distributed. Delicate,
graceful forms, with thin coenenchym. Echinomuricea, Verrill;
Muricea, Lamouroux; Paramuricea, Köll; Acamptogorgia, W. and S.;
Bebryce, Philippi.

Fam. 5. Plexauridae.—In this family we find some of the largest and

most substantial Gorgonids. The axis is usually black, but its horny
substance may be impregnated with lime, particularly at the base.
The coenenchym is thick, and the zooids are usually completely
retractile, and the surface smooth. The species of the family are
principally found in shallow water in warm or tropical regions.

The principal genera are:—Eunicea, Lamouroux. The calices are

prominent, and not retractile. Plexaura, Lamouroux; Euplexaura,
Verrill. Eunicella, Verrill. With an outer layer of peculiar torch-shaped
spicules. The only British species of this order is Eunicella cavolini
(formerly called Gorgonia verrucosa). It is found in depths of 10 to 20
fathoms off the coast of the English Channel and west of Scotland.
Occasionally specimens are found in which a gall-like malformation
with a circular aperture is seen, containing a Barnacle. Such gall
formations, common enough in some species of Madreporaria, are
rarely found in Alcyonaria.

Fig. 156.—Eunicella cavolini. Some branches of a large dried specimen, showing

a gall formed by a Cirripede.

Fam. 6. Gorgoniidae.—This family contains some of the

commonest and best-known genera of the order. They usually form
large flexible branched colonies with delicate horny axes and thin
coenenchym. The zooids are usually completely retractile.

The principal genera are:—Gorgonia, Linn. This genus includes

Gorgonia (Rhipidogorgia) flabellum, the well-known fan Gorgonia
with intimately anastomosing branches, from the warm waters of the
Atlantic Ocean. The genera Eugorgia, Verrill, and Leptogorgia, Milne
Edwards, differ from Gorgonia in the character of the spicules. In
Xiphigorgia, Milne Edwards, from the West Indies, the branches are
much compressed, forming at the edges wing-like ridges, which bear
the zoopores in rows. Malacogorgia, Hickson, has no spicules. Cape
of Good Hope.

Fam. 7. Gorgonellidae.—In this family the horny axis is

impregnated with lime. The surface of the coenenchym is usually
smooth, and the spicules small. The colonies are sometimes
unbranched (Juncella). In the branching forms the axis of the
terminal branches is often very fine and thread-like in dimensions.

Fig. 157.—Verrucella guadaloupensis, with an epizoic Brittle star (Oph.) of similar

colour.

The principal genera are:—Gorgonella, with a ramified flabelliform

axis; Ctenocella, with a peculiar double-comb manner of branching;
and Juncella, which forms very long unbranched or slightly branched
colonies, with club-shaped spicules. All these genera are found in
shallow water in the tropical or semi-tropical regions of the world.
Verrucella is a genus with delicate anastomosing branches found
principally in the shallow tropical waters of the Atlantic shores. Like
many of the Gorgonacea, with branches disposed in one plane
(flabelliform) Verrucella frequently carries a considerable number of
epizoic Brittle stars, which wind their flexible arms round the
branches, and thus obtain a firm attachment to their host. There is
no reason to suppose that these Brittle stars are in any sense
parasitic, as a specimen that bears many such forms shows no sign
of injury or degeneration, and it is possible they may even be of
service to the Verrucella by preying upon other organisms that might
be injurious. An interesting feature of the association is that the
Brittle stars are of the same colour as the host, and the knob-like
plates on their aboral surface have a close resemblance to the
verrucae (Fig. 157).

Order V. Pennatulacea.
The Sea-pens form a very distinct order of the Alcyonaria. They are
the only Alcyonarians that are not strictly sedentary in habit, that are
capable of independent movement as a whole, and exhibit a bilateral
symmetry of the colony. No genera have yet been discovered that
can be regarded as connecting links between the Pennatulacea and
the other orders of the Alcyonaria. Their position, therefore, is an
isolated one, and their relationships obscure.

The peculiarities of the order are due to the great growth and
modification in structure of the first formed zooid of the colony. This
zooid (Oozooid, Hauptpolyp, or Axial zooid) increases greatly in
length, develops very thick fleshy walls, usually loses its tentacles,
digestive organs, and frequently its mouth, exhibits profound
modification of its system of mesenteries, and in other ways
becomes adapted to its function of supporting the whole colony.
 Fig. 158.—Diagram of a Sea-pen. L, leaves composed of a row of autozooids; R,
rachis; St, stalk; T, anthocodia of the axial zooid, usually suppressed. (After
Jungersen.)

The axial zooid shows from an early stage of development a division

into two regions: a distal region which produces by gemmation on
the body-wall numerous secondary zooids, and becomes the rachis
of the colony; and a proximal region which becomes the stalk or
peduncle, and does not produce buds (Fig. 158). The secondary
zooids are of two kinds: the autozooids and the siphonozooids. The
former have the ordinary characters of an Alcyonarian zooid, and
produce sexual cells; the latter have no tentacles, a reduced
mesenteric system, and a stomodaeum provided with a very wide
siphonoglyph.

The arrangement of the autozooids and siphonozooids upon the

axial zooid is subject to great modifications, and affords the principal
character for the classification of the order. In the Pennatuleae the
autozooids are arranged in two bilaterally disposed rows on the
rachis, forming the leaves or pinnae of the colony (Fig. 158). The
number in each leaf increases during the growth of the colony by the
addition of new zooids in regular succession from the dorsal to the
ventral side of the rachis[385] (Fig. 159). In other Pennatulacea the
autozooids are arranged in rows which do not unite to form leaves
(Funiculina), in a tuft at the extremity of a long peduncle (Umbellula),
scattered on the dorsal side of the rachis (Renilla, Fig. 160), or
scattered on all sides of the rachis (Cavernularia, Fig. 161). In those
forms in which the autozooids are scattered the bilateral symmetry of
the colony as a whole becomes obscured. The siphonozooids may
be found on the leaves (Pteroeides), but more frequently between
the leaves or rows of autozooids, or scattered irregularly among the
autozooids. Usually the siphonozooids are of one kind only, but in
Pennatula murrayi there is one specially modified siphonozooid at
the base of each leaf,[386] which appears to have some special but
unknown function.

Fig. 159.—Diagram of a portion of a rachis of a Sea-pen, aut, The rows of

autozooids; 1-6, the order of age of the autozooids composing a leaf; D, the
dorsal side of the rachis; Si, the siphonozooids; V, the ventral side of the
rachis. (After Jungersen.)

In Umbellula gracilis each siphonozooid bears a single pinnate

tentacle, and in some other species of the same genus there is a
tentacle which is not pinnate.[387]

The zooids and coenenchym are usually protected by a crust of

coloured or colourless, long, smooth, needle-like, calcareous
spicules, situated principally in the superficial layer, so as to leave
the subjacent tissues soft and spongy in texture. In some cases the
spicules are smooth double clubs, rods, discs, or irregular granules,
and in Sarcophyllum, Chunella, some species of Umbellula and
others, there is no calcareous skeleton. The tuberculated spindles,
so common in other Alcyonaria, are not found in any species. In
most genera a horny, or calcified horny rod is embedded in the
central part of the axial polyp, serving as a backbone or support for
its muscles. It is absent, however, in Renilla, and reduced or absent
in Cavernularia.
The sexual organs are borne by the mesenteries of the autozooids
only, and each colony is either male or female. There is no record of
hermaphroditism in the order. The eggs contain a considerable
amount of yolk, and fertilisation is effected in the sea-water after their
discharge. The segmentation is irregular, and the free-swimming
ciliated larva (of Renilla) shows the rudiments of the first buds from
the axial polyp before it settles down in the mud.

The Sea-pens are usually found on muddy or sandy sea-bottoms,

from a depth of a few fathoms to the greatest depths of the ocean. It
is generally assumed that their normal position is one with the
peduncle embedded in the mud and the rachis erect. Positive
evidence of this was given by Rumphius, writing in 1741, in the case
of Virgularia rumphii and V. juncea at Amboina,[388] and by Darwin in
the case of Stylatula darwinii at Bahia Blanca.[389]

"At low water," writes Darwin, "hundreds of these zoophytes might be

seen projecting like stubble, with the truncate end upwards, a few
inches above the surface of the muddy sand. When touched or
pulled they suddenly drew themselves in with force so as nearly or
quite to disappear."

It is not known whether the Pennatulids have the power of moving

from place to place when the local conditions become unfavourable.
It is quite probable that they have this power, but the accounts given
of the Sea-pens lying flat on the sand do not appear to be founded
on direct observation. The fable of Pennatula swimming freely "with
all its delicate transparent polypi expanded, and emitting their usual
brilliant phosphorescent light, sailing through the still and dark abyss
by the regular and synchronous pulsations of the minute fringed
arms of the whole polypi," appears to be based on a statement made
by Bohadsch in 1761, and picturesque though it be, is undoubtedly
erroneous.
The brilliant phosphorescence of many species of Pennatulacea has
been observed by many naturalists, and it is very probable that they
all exhibit this property to some degree. The phosphorescence
appears to be emitted by the mesenteric filaments of the autozooids,
but it is not yet determined whether the phenomenon is confined to
these organs or is more generally distributed.

The Pennatulacea are usually devoid of epizoites, but occasionally

the parasitic or semi-parasitic Entomostracan Lamippe is found in
the zooids. A small crab is also frequently found between the large
leaves of species of Pteroeides. The most remarkable case of
symbiosis, however, has recently been observed in the form of an
encrusting Gymnoblastic Hydroid[390] living on the free edge of the
leaves of a species of Ptilosarcus.

The order Pennatulacea is divided into four sections.

Sect. 1. Pennatuleae.—In this section the colony is distinctly

bilaterally symmetrical, and the autozooids are arranged in rows with
their body-walls fused to form leaves.

The genus Pteroeides, the representative genus of the family

Pteroeididae, is a fleshy Sea-pen found in shallow sea water in the
warm waters of the Pacific Ocean and in the Mediterranean. It has
large leaves with long spiny, projecting spicules, and the
siphonozooids are borne by the leaves. Pennatula, the
representative genus of the family Pennatulidae, has a wider
distribution in area and in depth. Pennatula phosphorea is a common
British species, found in depths of 10 to 20 fathoms in many
localities off our coasts. It is about 5 inches in length. There are
several varieties of this species distributed in Atlantic waters.
Pennatula grandis is a magnificent species found in Norwegian
fjords, in the Faeroe Channel, and off the northern coasts of N.
America, in depths of from 50 to 1255 fathoms. Specimens have
been obtained no less than 2½ feet in length. P. murrayi and P.
naresi are species of the genus found at depths of a few hundred
fathoms in tropical seas.

The genus Virgularia, belonging to the family Virgulariidae, is

represented in the British seas by V. mirabilis, a long slender Sea-
pen found in many localities off the Scottish coasts.

Sect. 2. Spicatae.—This section includes those Sea-pens in which

the autozooids are arranged bilaterally on the axial zooid in rows or
more irregularly, but do not unite to form leaves. It is a large section
and contains many widely divergent genera.

The family Funiculinidae is represented on our coasts by Funiculina

quadrangularis, a long and slender Sea-pen 2 to 3 feet in length. The
autozooids are arranged in oblique rows, and the siphonozooids are
on the ventral side of the rachis. There is one point of special interest
in this genus. The siphonozooids appear to change as the colony
grows and to become autozooids. If this is the case it may be more
correct to describe the genus as devoid of true siphonozooids.

The family Anthoptilidae contains the species Anthoptilum

grandiflorum, which has a wide distribution in depths of 130 to 500
fathoms in the N. and S. Atlantic Ocean. It is perhaps the largest of
all the Pennatulacea, specimens having been obtained from the
Cape of Good Hope over 4 feet long with expanded autozooids,
each more than half an inch in length.

The family Kophobelemnonidae contains a number of forms with

remarkably large autozooids arranged in irregular rows on the two
sides of the rachis. The siphonozooids are numerous and scattered,
and their position is indicated by small papilliform calices on the
coenenchym. The surface of these pens is usually rough, owing to
the presence of numerous coarse projecting spicules.
Kophobelemnon occurs in the Mediterranean in deep water, off the
coasts of Ireland and Scotland, and in other regions.

The family Umbellulidae contains some of the most remarkable and

interesting examples of the deep-sea fauna. The peduncle is very
long and the rachis stunted and expanded. The autozooids are of
great size, non-retractile, and arranged in a cluster or rosette on the
terminal rachis. There is a wide structural range between the
species. Some species have numerous large spicules, others have
none. In some species the siphonozooids have a single pinnate or
digitate tentacle, in others the siphonozooids are of the usual type.
Umbellula appears to be a somewhat rare but cosmopolitan genus in
deep water, extending from the Arctic to the Antarctic region in water
ranging from 200 to 2500 fathoms.

The interesting genus Chunella was discovered by the German

"Valdivia" Expedition at a depth of about 420 fathoms off the coast of
E. Africa, and subsequently by the Dutch "Siboga" Expedition at a
depth of about 500 fathoms in the Malay Archipelago. According to
Kükenthal,[391] this genus with another closely allied genus
Amphianthus should form a new section of Pennatulacea, the
Verticilladeae. Chunella has a long and very delicate rachis and
peduncle, and the former terminates in a single autozooid and has
five or six whorls of three autozooids, situated at considerable
distances from one another. Spicules are absent. The full description
of this genus has not yet been published, but it is clear that it
occupies a very isolated position in the order.

Fig. 160.—Renilla reniformis, a small specimen (34 mm.), showing the dorsal
side of the expanded rachis. A, autozooid; H, the mouth of the axial zooid;