How to Write an Article

The scientific publication
Lecture no. 4, March 27 2013
Enrico Rubiola
Lecture series for PhD students
of the Universities of Bourgogne and Franche Comté
March – April 2013
Guest attendees are welcome

home page http://rubiola.org

2

How to write an article

3

Anatomy of a short article

RAPID COMMUNICATIONS

journal, and reference
title
authors

Thermal-noise-limited crystalline whispering-gallery-mode resonator for laser stabilization
J. Alnis,1 A. Schliesser,1,2,* C. Y. Wang,1 J. Hofer,1 T. J. Kippenberg,1,2 and T. W. H¨ansch1

affiliations

Max-Planck-Institut f¨ur Quantenoptik, D-85748 Garching, Germany
´ erale
´
Ecole Polytechnique Fed
de Lausanne (EPFL), CH-1015 Lausanne, Switzerland
(Received 20 February 2011; published 15 July 2011)

We have stabilized an external cavity diode laser to a whispering gallery mode resonator formed by a
< 1 cm3 ) reduce
protrusion of a single-crystal magnesium fluoride cylinder. The cylinder’s compact dimensions (∼
the sensitivity to vibrations and simplify the stabilization of its temperature in a compact setup. In a comparison
to an ultrastable laser used for precision metrology we determine a minimum Allan deviation of 20 Hz at an
optical wavelength of 972 nm, corresponding to a relative Allan deviation of 6 × 10−14 , at an integration time
of 100 ms. This level of instability is compatible with the limits imposed by fundamental fluctuations of the
material’s refractive index at room temperature.

DOI

reference author

4

1

2

Introduction

short
article

PHYSICAL REVIEW A 84, 011804(R) (2011)

DOI: 10.1103/PhysRevA.84.011804

[email protected]

1050-2947/2011/84(1)/011804(4)

abstract

PACS number(s): 42.60.Da, 42.62.Fi, 06.30.Ft

In many applications of lasers in metrology, sensing, and
spectroscopy, the quality of a measurement critically depends
on the spectral purity of the employed laser source. For
this reason, researchers’ efforts have been directed toward
the reduction of the laser’s phase or frequency fluctuations
since the early days of laser science [1]. Tremendous progress
has been made since then, culminating in light sources with
linewidths below 1 Hz, and a relative Allan deviation of their
frequency below the 10−15 level [2,3]. Such ultrastable lasers
are operated in a number of specialized laboratories and are
realized by electronically locking the laser frequency to a
resonance of an external reference cavity defined by two
high-reflectivity mirrors attached to a low-thermal-expansion
spacer [4–8]. Their performance has been enabled by sophis> 1 dm3 -sized spacers’ material and
ticated engineering of the ∼
geometry, as well as their vibration and thermal isolation. It is
now understood to be limited by thermodynamic fluctuations
in the mirror substrates and coatings [9,10]. Several proposals
to overcome this limitation are discussed in the literature,
including cryogenic operation of reference cavities [11–14]. In
an alternative method, the optical reference cavity is replaced
by a long (∼km) all-fiber delay line yielding a simplified setup
at the expense of slightly reduced stability [15].
In this work we explore a different approach [16,17] to
provide a reference for the laser’s frequency, by fabricating
whispering-gallery-mode (WGM) resonators from a single
crystal of magnesium fluoride (Fig. 1). The WGM resides in
a protrusion of a MgF2 cylinder, whose compact dimensions
< 1 cm3 ) and monolithic nature inherently reduce the res(∼
onator’s sensitivity to vibrations. It also enables the operation
in more noisy and/or space-constrained environments, such
as a cryostat or a satellite. Furthermore, in contrast to the
highly wavelength-selective and complex multilayer coatings
required for mirror-based resonators, WGM resonators are
intrinsically broadband, limited only by optical absorption in
the host material.
Experiments with silica microspheres [18–20] have
already demonstrated the potential to realize compact, narrowlinewidth laser sources with WGM resonators. But only

*

date

in recent experiments [21,22] with crystalline CaF2 WGM
resonators has the laser frequency stability been quantitatively
assessed with sufficient resolution. The lowest relative Allan
deviations have been determined at the 10−12 level for
submillisecond integration times. We are able to significantly
improve this performance and explore the thermodynamic
limits of this approach at room temperature.
We fabricate WGM resonators following the pioneering
work of Maleki and coworkers [23,24], combining a shaping
and several polishing steps on a home-built precision lathe
[25]. Several MgF2 resonators were produced with a typical
radius of 2 mm, one of which is shown in Fig. 1(a). The WGMs
are located in the rim of the structure, which has been fine
polished with a diamond slurry of 25 nm average grit size in
the last step. The resulting surface smoothness, together with
very low absorption losses in the ultrapure crystalline material
(Corning), enables quality factors in excess of 2 × 109 .
A high-index prism is used to couple the beam of an external
cavity diode laser into the WGM. For laser stabilization, we
choose to work in the undercoupled regime, in which the
presence of the coupling prism has only a weak effect on

FIG. 1. (Color online) MgF2 resonator used for laser frequency
stabilization. (a) Photograph of the resonator on brass mount and
coupling prism (right). (b) Typically observed WGM resonance with
laser-limited 0.6 MHz linewidth (sidebands are due to laser frequency
modulation at 5 MHz). (c) Ringing in the transmission signal observed
in a fast laser sweep. The envelope of the oscillations (dashed green
line) indicates an exponential field decay from the WGM within
2.1 µs, corresponding to Q ≈ 2.0 × 109 . (d) Transmission for a laser
scan over ca. 3 free spectral ranges showing different transverse
WGMs.

011804-1

©2011 American Physical Society

keywords

Methods

additional info

J. ALNIS et al.

RAPID COMMUNICATIONS

running head

1st
author

FIG. 2. (Color online) Thermal and acoustic isolation of MgF2
cavity and coupling setup (not to scale).

the WGM quality factor Q and resonance frequency ωc (see
below). To accurately determine the linewidth κ = ωl /Q of
such a resonance, we sweep the laser frequency ωl = 2π c/λ
(c is the vacuum light speed, λ = 972 nm the laser wavelength)
through the WGM resonance at a rate exceeding κ 2 . When the
circulating power built up in the WGM decays during the
ringdown time κ −1 , it interferes with the laser field swept
already beyond the cavity resonance. A transient, chirped
heterodyne beat is therefore observed in the transmitted power
(Fig. 1) , the envelope of which decays with the cavity field
amplitude ringdown time 2κ −1 . The linewidth determined in
this way is unobscured by the linewidth of the probing laser
and is given by κ = 2π × 150 kHz for the WGM used in this
work (finesse F = 87,000).
For the laser stabilization experiments, a MgF2 resonator
is mounted into a prism-coupling setup shielded against
vibrations and thermal fluctuations. In this proof-of-concept
experiment (Fig. 2), we have made no attempts to miniaturize
the assembly but have drawn on standard techniques developed
for isolating larger reference resonators from their environment
[7]: To suppress acoustic perturbations, the setup is kept in high
vacuum (p < 10−6 mbar) maintained in an aluminum chamber
by an ion pump. An additional box of heavy loudspeaker
plywood shields the experimental setup from acoustic noise,
while a passive vibration isolation platform (Minus-K) and
an optical table suppress the transmission of vibrations to the
coupling setup.
The resonance frequency ωc of the WGM sensitively
depends on the temperature of the resonator due to induced
changes in the radius R and refractive index n of the device,
as described by the respective coefficients αl = R1 dR
and
dT
dn
αn = n1 dT
. Since αl is as large as 10−5 /K in MgF2 at room
temperature, it is important to decouple the resonator from the
unstable (±0.2 K) laboratory temperature. The low pressure
in the chamber already suppresses heat exchange through
the residual gas, while heat conduction through the support
of the coupling setup is reduced by minimizing the contact
area to three thermally insulating spacers. Heat exchange via
thermal radiation is reduced by introducing two aluminum
shields between the chamber’s walls and the actual coupling
setup (Fig. 2). These measures increase the inertial time for
the transmission of temperature changes from the laboratory

PHYSICAL REVIEW A 84, 011804(R) (2011)

environment to the resonator to several hours. In addition,
electronic feedback to a two-stage thermoelectric element is
used to actively stabilize the temperature of the outer shield to
a value close to room temperature. With the loop activated, the
inner shield’s temperature is stable to 1 mK/day according to
an out-of-loop temperature sensor.
The beam of a commercial Littman-type extended cavity
diode laser (New Focus TLB) is focused on the face of the
coupling prism through antireflection coated windows in the
vacuum chamber and small (diameter 1 cm) bore holes in
the aluminum shields. The laser is locked to a high-Q WGM
using the Pound-Drever-Hall (PDH) method [26] implemented
here with an external electro-optic phase modulator driven at
11.4 MHz and demodulation of the transmission signal at the
same frequency (Fig. 3). The power, polarization, position,
and pointing of the laser beam incident on the coupling prism,
as well as the prism-resonator gap are then optimized for the
best error signal. This signal is fed back via a two-branch
control system actuating both the grating tilt in the laser (via
a piezoelectric transducer) and the diode pump current. The
error signal is amplified by home-built proportional-integral
(PI) controllers in both branches; the current controller has an
additional phase advance (PID) for faster regulation up to a
feedback frequency of ca. 1 MHz.
To assess the frequency fluctuations of the laser locked
to the WGM resonator, its frequency is compared to that of
an independent diode laser locked to an ultrastable mirrorbased resonator in the same laboratory [7]. The extraordinary
stability of the latter affords a direct measurement of the
WGM-stabilized laser by analyzing the spectral properties
of the radio-frequency (rf) beat generated between the two
lasers in a heterodyne detector. Using a rf spectrum analyzer,
the width of the beat signal at 586 MHz could be fitted to a
Lorentzian of 290 Hz linewidth, close to the 300 Hz resolution

PID

11.4 MHz

APD

journal and
reference

f=18mm
MgF2

λ/2

λ/2

50:50

ECDL1

5

vibration isolation platform

PDH signal

PI

short
article

APD

optical table 1

beat note
spectral
analysis

~20 m
optical table 2

PDH lock
ECDL2
ULE cavity

laser linewidth 0.5 Hz, drift 0.05 Hz/s

FIG. 3. (Color online) Laser stabilization to a MgF2 resonator using the PDH method, and comparison to an ultrastable
laser locked to a mirror-based cavity on another optical table.
APD: avalanche photodiode; ECDL: external cavity diode laser;
EOM: electro-optic modulator; λ/2: half-wavelength retardation
plate; PI(D): proportional-integral-(differential) feedback controller;
ULE: ultralow-expansion glass.

011804-2

Article no 011804.
Pages are numbered
011804-1 to 011804-4
Necessary for online
early print

running head

PHYSICAL REVIEW A 84, 011804(R) (2011)

bandwidth of the recording (see inset of Fig. 3). For a more
systematic characterization, we determine the Allan deviation
σy (τ ) of the stabilized laser’s frequency as a function of
gate time τ . To that end the electronic beat signal is mixed
to a frequency close to 30 MHz using a stable synthesizer
and then filtered with a 6 MHz-wide bandpass filter. For
τ < 1 ms, a counter computing the Allan deviation internally
(Stanford Research SR620) was used. Allan deviations for
longer gate times were obtained by offline processing a large
set of dead-time-free frequency readings from a different
counter (K + K Messtechnik FX80) measured at gate times
of 2 and 100 ms. All determined Allan deviations correspond
to classical #-type [27] counting results with the counters’
backgrounds well below 10 mHz/(τ/ s).
The results of these measurements are shown in Fig. 4 along
with data from earlier experiments [7]. We attribute the shortterm (! 1 ms) fluctuations to uncompensated fluctuations of
the diode lasers’ phase due to the limited feedback bandwidth
and gain of the stabilization electronics. For the same reason,
we also expect Brownian noise in the cavity due to thermal
excitation of its mechanical modes—as observed recently
in CaF2 WGM resonators [25]—to be insignificant in this
regime. For gate times between ca. 10 ms and 1 s, the Allan
deviation is approximately constant at a level of several
10 Hz, corresponding to a relative Allan deviation below
10−13 (Fig. 4). For gate times beyond 1 s, the Allan deviation
increases again, with an asymptotic σy (τ ) ∝ τ dependence
for long gate times, indicative of linear frequency drift. The
extracted drift rate of 38 Hz/s is compatible with drift of the

resonator temperature due to the imperfect stabilization of
the latter and can be subtracted from the data for further
analysis. The residual Allan deviation at longer time scales
(τ > 10 s) is equally expected to be caused mainly by temperature instability. Other sources specific to this setup—such
as the resonator-prism gap d controlled by a piezoelectric
transducer—were checked to be less critical (|∂ωc /∂d| ≈
2π 1.6 kHz/nm). Importantly, temperature fluctuations can be
induced not only by a changing environment temperature, but
also by fluctuations of the absorbed optical power, as caused by
amplitude fluctuations of the intracavity laser field. Indeed, the
large measured resonance frequency dependence on the power
sent into the coupling setup (∂ωc /∂Pin ≈ −2π 28 kHz/µW)
indicates that photothermal noise could be another important
source of instability. In an independent measurement, we have
determined the power fluctuations of the diode laser in the
(Fourier) frequency range between 30 mHz and 300 Hz and
< 6 × 10−5 ,
have found relative rms-power fluctuations of ∼
which could induce cavity frequency fluctuations of up to
25 Hz for Pin = 15 µW. Figure 4 shows the estimated upper
limit for the photothermally induced instability, where an
instantaneous response of the cavity resonance frequency
to power fluctuations has been assumed, and a spectral
dependence SP P (') ∝ '−2 measured in the range relevant
for the shown Allan deviations was used.
While the influence of environment temperature fluctuations and photothermal noise can be further suppressed by
technical improvements, more fundamental sources of noise
were observed to limit the lowest Allan deviations measured.
These noise sources originate from the fact that the temperature
of a body of heat capacity CV at an average temperature T
intrinsically fluctuates as 'δT 2 ( = kB T 2 /CV , where kB is
Boltzmann’s constant [28]. This implies, for example, that
the temperature of the resonator fluctuates by 'δTr2 ( =
kB T 2 /cV ρVr ≈ (2 nK)2 with the specific heat capacity cV =
1020 J/kg K and density ρ = 3180 kg/m3 of MgF2 [29]. The
resulting fluctuations of the resonator radius [16] lead to
comparably small resonance frequency fluctuations on the
order of 'δωc2 ( = ωc2 αl2 'δTr2 ( ≈ (2π 7 Hz)2 . In contrast, the
fluctuations of temperature Tm sampled by the optical mode
are larger
to the! significantly smaller mode volume [30]
! due
) 2 d 3 r)2 / |E|
) 4 d 3 r of approximately 2 × 10−12 m3
Vm = ( |E|
in the used resonator, leading to 'δTm2 ( = (0.4 µK)2 . This
fluctuating temperature directly modulates the refractive index
in the WGM and, as a consequence, its resonance frequency
according to 'δωc2 ( = ωc2 αn2 'δTm2 ( ≈ (2π 60 Hz)2 , where we
have used the mean value of αn = 5 × 10−7 /K for ordinarily
and extraordinarily polarized light at 972 nm. The large fluctuations due to this thermorefractive noise [30,31] therefore
limit the frequency stability of the WGM.
For a quantitative comparison with our measurements,
we calculate the expected Allan deviation as a function
of measurement time. Approximating the resonator as an
infinitely long cylinder of MgF2 , the (double-sided) spectral
density of fluctuations of the average temperature u¯ of the
material accommodating the optical mode can be estimated
to [16,30,31]
#
kB T 2 " 2π DR +2 (kp ,km ) dkm
Su¯ u¯ (') ≈
,
(1)
cV ρVm p
'2 + D 2 k 4
2π

5

10

-10

10
4

10
10

-12

10
2

10

1

-13

10

10

0

10

290 Hz

-14

10

Relative Allan deviation

-11

10
3

RF power (a.u.)

Results

THERMAL-NOISE-LIMITED CRYSTALLINE WHISPERING- . . .

Allan deviation at 308 THz (Hz)

title

RAPID COMMUNICATIONS

-15

10

Beat frequency
-1

10

-6

10

-5

10

-4

10

-3

10

-2

10

-1

10

Gate time (s)

0

10

1

10

2

10

3

10

FIG. 4. (Color online) Allan deviation of the optical beat note in
Hz, and normalized to the optical carrier at 308 THz. Inset shows a
beat note measured at 586 MHz with a spectrum analyzer (blue points)
and Lorentzian fit (red line). Diamonds are reference measurement
of beat-note stability between two lasers locked to two mirror-based
resonators [7]. Red symbols are Allan deviation of beat frequency
between two lasers locked to a mirror-based and the WGM resonator
obtained with two different counters (triangles and squares; see text).
Full (open) symbols show data after (before) removal of a linear drift
of 38 Hz/s. For comparison, gray circles show the Allan deviation
measured between a laser stabilized to a two-mirror cavity and
one tooth of an optical frequency comb stabilized by a hydrogen
maser. Blue dashed and orange dotted lines indicate the estimated
Allan deviation due to thermorefractive and photothermal noise,
respectively.

011804-3

short
article

6

journal and
reference

Article no 011804.
Pages are numbered
011804-1 to 011804-4
Necessary for online
early print

PHYSICAL REVIEW A 84, 011804(R) (2011)

J. ALNIS et al.
2

Discussion /
Conclusion

2
km

kp2 ,

∗

−6

2

where k =
+
and D = λ /cV ρ ≈ 7.9 × 10 m /s is
MgF2 diffusivity (assuming, for simplicity, a mean thermal
conductivity of λ∗ ≈ 26 W/K m). The wave numbers kp (km )
characterize the thermal waves in the radial (axial) direction
permitted by the boundary conditions, implying in particular
that kp R are the roots of the first-order cylindrical Bessel
function in this model. The overlap function #(kp ,km ) of
the thermal and optical modes is approximately constant for
all thermal waves with wave numbers k/2π that are much
smaller than the inverse transverse dimension b−1 of the
optical mode. Thermal modes with larger spatial frequencies
make only small contributions to the sum or integrals at
Fourier frequencies % % Db−2 , so that the spectrum of WGM
frequency fluctuations can be calculated from
√ Eq. (1) using
Sωω (%) = ωc2 αn2 Su¯ u¯ (%) with #(kp ,km ) ≈ 1/ π R. Figure 4
shows the resulting frequency fluctuations converted to an
Allan deviation. In spite of this only rough estimation, it is clear
that thermorefractive noise constitutes a critical limitation
for WGM frequency references. The lowest Allan variances
determined in our measurements are more than one order
of magnitude above the level expected from laser phase
noise and thermal drifts, and compatible with estimations for
thermorefractive noise.
In conclusion, we have demonstrated the stabilization of
a diode laser by locking it to a crystalline WGM resonator
used as an optical reference and characterized the stability
of the laser’s frequency by comparison with an ultrastable

optical reference. At an integration time of 100 ms, the lowest
measured Allan deviation was 20 Hz, a value that is compatible
with thermorefractive fluctuations in the resonator material.
While the stability does not yet reach the level achieved
by optical references based on two-mirror resonators, WGM
resonators could be used to significantly reduce the frequency
fluctuations of diode lasers or free-running frequency combs
in simple, compact setups, or be employed to transfer a first
laser’s stability to other lasers across a wide wavelength region.
The references’ performance is expected to improve if the
strong temperature sensitivity can be reduced or eliminated.
For example, the operation close to temperatures (∼ 200◦C) at
which the thermorefractive coefficients αn of MgF2 vanish [17]
is practically feasible. On the other hand, self-referenced
temperature stabilization [32] could dramatically improve
temperature stability. Finally, the inherent compatibility of
these resonators with cryogenic operation opens a promising
approach to not only an improved temperature stability and
reduced sensitivity to temperature fluctuations, but also a
strong suppression of thermodynamic fluctuations limiting
also today’s best optical flywheels.

indivuduals

The authors thank M. Gorodetsky, E. Rubiola, and T. Udem
for valuable discussions and acknowledge funding by a Marie
Curie IAPP, the EU project SOC 2, and the Swiss National
Science Foundation.

funding agency

References

[1] A. D. White, IEEE J. Quantum Electron. 1, 349 (1965).
[2] B. C. Young, F. C. Cruz, W. M. Itano, and J. C. Bergquist, Phys.
Rev. Lett. 82, 3799 (1999).
[3] Y. Y. Jiang et al., Nature Photon. (London) 5, 158 (2011).
[4] H. Stoehr et al., Opt. Lett. 31, 736 (2006).
[5] A. D. Ludlow et al., Opt. Lett. 32, 641 (2007).
[6] S. A. Webster, M. Oxborrow, S. Pugla, J. Millo, and P. Gill,
Phys. Rev. A 77, 033847 (2008).
[7] J. Alnis, A. Matveev, N. Matveev, T. Udem, and T. W. H¨ansch,
Phys. Rev. A 77, 053809 (2008).
[8] J. Millo et al., Phys. Rev. A 79, 053829 (2009).
[9] K. Numata, A. Kemery, and J. Camp, Phys. Rev. Lett. 93, 250602
(2004).
[10] M. Notcutt et al., Phys. Rev. A 73, 031804 (2006).
[11] H. J. Kimble, B. L. Lev, and J. Ye, Phys. Rev. Lett. 101, 260602
(2008).
[12] M. L. Gorodetsky, Phys. Lett. A 372, 6813 (2008).
[13] D. Meiser, J. Ye, D. R. Carlson, and M. J. Holland, Phys. Rev.
Lett. 102, 163601 (2009).
[14] S. Seel, R. Storz, G. Ruoso, J. Mlynek, and S. Schiller, Phys.
Rev. Lett. 78, 4741 (1997).
[15] F. K´ef´elian et al., Opt. Lett. 34, 914 (2009).
[16] A. B. Matsko et al., J. Opt. Soc. Am. B 24, 1324 (2007).
[17] A. A. Savchenkov et al., J. Opt. Soc. Am. B 24, 2988 (2007).

7

Ac

[18] V. V. Vassiliev et al., Opt. Commun. 158, 305 (1998).
[19] K. Kieu and M. Mansuripur, Opt. Lett. 32, 244 (2007).
[20] B. Sprenger, H. G. L. Schwefel, and L. J. Wang, Opt. Lett. 34,
3370 (2009).
[21] B. Sprenger et al., Opt. Lett. 35, 2870 (2010).
[22] W. Liang et al., Opt. Lett. 35, 2822 (2010).
[23] I. Grudinin et al., Opt. Commun. 265, 33 (2006).
[24] I. S. Grudinin, V. S. Ilchenko, and L. Maleki, Phys. Rev. A 74,
063806 (2006).
[25] J. Hofer, A. Schliesser, and T. J. Kippenberg, Phys. Rev. A 82,
031804 (2010).
[26] R. W. P. Drever et al., Appl. Phys. B 31, 97 (1983).
[27] E. Rubiola, Rev. Sci. Instrum. 76, 054703 (2005).
[28] L. D. Landau and E. M. Lifshitz, Statistical Physics (Pergamon
Press, Oxford, England, 1980).
[29] M. J. Weber, Handbook of Optical Materials (CRC Press, Boca
Raton, FL, 2003).
[30] M. L. Gorodetsky and I. S. Grudinin, J. Opt. Soc. Am. B 21, 697
(2004).
[31] V. B. Braginsky, M. Gorodetsky, and S. Vyatchanin, Phys. Lett.
A 271, 303 (2000).
[32] A. B. Matsko, A. A. Savchenkov, V. S. Ilchenko,
D. Seidel, and L. Maleki, Phys. Rev. A 83, 021801
(2011).

011804-4

kn
ow

led

ge

m

en

ts

Useful readings
EASE Guidelines for Authors and
the Editors’ Desks
Translators ofFrom
Scientific
Articles to
Be Published in English
EASE Guidelines for Authors and Translators of Scientific Ar ticles to be Published in English, June 2011

June 2011

2

Guidelines
Appendices

7

Abstracts

8

Ambiguity

9

Cohesion

10

Ethics

11

Plurals

12

Simplicity

13

Spelling

14

Text-tables

15

About EASE

www.ease.org.uk
©2011 European Association of Science Editors (www.ease.org.uk). Non-commercial printing allowed.

8
1

The IMRaD Format for Scientific Papers
Suitable to short articles and letters
Titles may not be appear explicitly

• Introduction What was the question?
• Methods How did you try to answer it?
• Results What did you find?
• Discussion What does it mean?

© R.A.Day, B.Gastel

9

Title

• The fewest possible words that adequately
indicate the contents of the paper
• Important in literature searching
• Should not include extra words, such as “a
study of”
• Should be specific enough but not overly
narrow

© R.A.Day, B.Gastel

10

Authorship
• Only those researchers who give a significant contribution to a research work
should be the authors
• Some journals now request and publish information about the contributions of
each person
• Alphabetical order is sometimes used, necessary in a large list of authors
• Three logical author functions
• First author
• Corresponding author
• Principal investigator
• Funding, technical work, and administration
• Is appropriate in the Acknowledgements
• Should not go with authorship
• Context-dependent interpretation
• Co-authorship at local (national) conferences can be a means to reward
people who should not authors in high-impact articles
See also Vancouver group, since 1978 (bio-med) – ICMJE

11

Abstract
• Summarizes the paper
• Widely read and therefore important
• Commonly organized in IMRAD format (may be
structured abstract, with headings
corresponding to the various sections)
• Content must be consistent with that in the
paper
• Normally should not include figures, tables,
references

© R.A.Day, B.Gastel

12

The Core of the Paper

•Introduction
•Methods
•Results
•Discussion

© R.A.Day, B.Gastel

13

Introduction

• Provides background needed to understand the
paper and appreciate its importance
• Identifies the question the research addressed
• In general, should be fairly short
• Typically should be funnel-shaped, moving from
general to specific.

© R.A.Day, B.Gastel

14

Methods
• Purposes: to allow others to replicate and to evaluate
what you did
• Should describe the study design
• Should identify (if applicable)
• Equipment, organisms, reagents, etc used (and sources thereof)
• Approval of human or animal research by an appropriate committee
• Statistical methods

• May include tables and figures
• An issue: level of detail in which to describe
• Well-known methods
• Methods previously described but not well known
• Methods that you yourself devised

• Helpful to use papers published in the same journal
as models
© R.A.Day, B.Gastel

15

Results

• The core of the paper
• Often includes tables, figures, or both
• An issue: how much the information in the text
should overlap with that in the tables and
figures
• Should present results but not comment on
them

© R.A.Day, B.Gastel

16

Discussion
• Often should begin with a brief summary of the
main findings
• Should answer the question stated in the
introduction
• Some other items commonly addressed:
• Limitations of the study
• Relationship to findings of other research
• Other research needed

• Typically should move from specific to general
(opposite of introduction)

© R.A.Day, B.Gastel

17

Acknowledgments

• A place to thank people who helped with the
work but did not make contributions deserving
authorship
• Permission should be obtained from people you
wish to list
• Sometimes the place where sources of
financial support are stated

© R.A.Day, B.Gastel

18

References
• Functions:
• To give credit
• To add credibility
• To help readers find further information

• Importance of accuracy
• Existence of various reference formats
• Availability of citation management software
(examples: EndNote, Reference Manager)

© R.A.Day, B.Gastel

19

20

Software tools

21

WYSIWYG word processing

22

Microsoft Word
Screenshot

Printed page
Submitting a Book Proposal to
Cambridge University Press
These notes are intended to help you put together a detailed outline of your book
and to give you an idea of the process that Cambridge follows once the proposal
has been submitted.
Don’t feel that you have to follow this framework exactly. However, the
following headings do cover most of the important information that we need in
order to evaluate your proposal.
Title (and subtitle, if any)
A clear and accurate title is important in marketing your book. As a general rule
the main title should have no more than seven words. (If the title looks like it will
be longer than that then consider using a subtitle too. Many people use search
engines when hunting for books so if key words or acronyms are in the subtitle
they will still be seen.)

≈

Author Name and Affiliation

You should include your mailing address, e-mail, and phone/fax numbers.

Background
This should outline the general field, how it has evolved, where it is going, its
commercial importance (if any) and so on.

Brief Description of the Book

Here you should set out, in a few paragraphs, what specifically the book will be
about. You should discuss the approach you intend to take (e.g. the balance
between theory and practice) and any particular presentational or pedagogical
features that will characterize the book. Will it, for example, include real-world
case studies? Practical hints and tips for practitioners?

Your Reasons for Writing the Book
Why do you think this book should be published and how will it benefit the
readers?

Market and Readership

Here you should describe exactly who the book is aimed at (e.g. graduate
students, researchers, practitioners in industry, etc.) and in what subjects they
work/study (e.g. electrical engineering, computer science, applied physics, etc.).
If the readership of your book includes practising professional engineers, please
be as specific as possible in describing their job functions (e.g. microwave circuit
design engineer). If the book can be used as a textbook/short course text or then
you should describe the type of course for which it could be adopted. In this

What You See Is What You Get

23

Multilingual support
computer detects

spell

trovano

orthografia
les fautes
orthographe

Track Changes: a tool for team working

24

Changes made
by Enrico Rubiola

Accept or
reject changes

Changes made by
Vincent Giordano

25

Equations

%
%
!"#$ &' ()*+

+Microsoft Word 2008
+
+ equation editor
1 2
2
x + y " 0+
2
fast typesetting

%
%
!"#$ &' ()*+

+

Professional typography

displayed formula

p
1
2
2
x +y
2

0

inline formula

p
1
2
2
x
+
y
2

0

Quality of graphics
There may be a loss of quality in high-resolution prints

Original graphics
computer detects

Pasted from Word 2008

spell

trovano

orthografia
les fautes
orthographe

26

Quality of graphics
There may be a loss of quality in high-resolution prints

Original vector graphics
control

synt

DUT
detector

synt
VCO

control

dual-channel FFT analyzer

VCO

Pasted from Word 2008

27

Word – The bottom line
• Pros

28

• Cons

• Basic use extremely simple and
quick to learn
• Spell check and grammar check
• Sophisticated multilingual support
• Track changes
• Full integration with other programs
• Generally efficient for small
documents
• Industry standard
• Accepted by all publishers

• Advanced use is terribly complex and
difficult to learn
• Painful search through menus
• Limited set of symbols, difficult to find
• Loss of quality with vector graphics
• Equations are difficult to write
• Sometimes small documents give a
large file
• Print generally inferior to pro quality
• Large documents are difficult or
impossible to manage (split)
• Document damages when changing
version

.doc is proprietary format
.docx is XML, which is open format
Some people (me!) think the document structure, and get distracted by the layout

Other WYSIWYG programs
• OpenOffice
• Similar to Office, and free, but less efficient
• Export in PDF and Latex

• LibreOffice
• Alternate version of OpenOffice, born when Sun/Oracle tried to limit
the freedom with OpenOffice

• Lyx
• Free, yet small community of users and lacks most pro features

• Pages (Apple)
• Simple and beautiful results, but lacks most pro features

• Scientific Word
• Uses Tex/Latex as the typesetting engine
• Outstanding for technical writing

• TexMacs (free)
• Free, but small community of users
• Beautiful prints, but lacks most pro features
• Designed for integration with some mathematical packages

29

30

Structured text processing

Descendants of TROFF

The Tex / Latex family
.tex file (ascii)

• The .tex is an ASCII file
• typesetting commands
• text

manuscript

style
file

latex
library

provided by
the publisher

font
metrics

tex
engine

Latex

device
independent

fonts

dvi to ps
preview
.ps file
pdf to ps
.pdf file

• Extremely compact files
• Latex processes boxes (font metrics)
instead of graphics
• True onts are added at preview/print
time

.dvi file
fonts

31

display

• Scalable graphics
• Full professional quality with early
computers (1980s: 5 MHz clock, 640 kB
RAM, 20 MB hard disk)
• Proved portability over ≥ 30 years (!!!)
• Free, open source
• Supported by the American
Mathematical Society

PdfLatex
.tex file (ascii)
manuscript

provided by
the publisher

latex
library

style
file

• A different flavor of latex which skips
dvi and postscript
• Faster and simpler engine
• Default in most installations
• Same results as regular Latex (almost)

font
metrics
fonts

tex
engine

PdfLatex
.pdf file

printer

display

• Preferred in most cases
aaa • Personal use, up to small/mid-size
publishing companies
• Nowadays, very few use true
postscript (2400 dpi photo-plotters)

32

The Tex engine

33

Pioneering design (D. Knuth, 1978)
• Automated placement of floating bodies (figures,
tables)
• Automated numbering of chapters, sections,
figures, formulas
• Refer to numbered objects by name (label)
• Table of contents automatically generated and
updated
• Also list of tables, figures, etc.
• Index automatically generated and updated (via
Makeindex)
• Bibliography management (via Bibtex)
• Virtually unlimited font set (via Metafont)

34

2.4.2
8.7

Dashes and Hyphens

2.4.2 Dashes and Hyphens
35
LATEX knows
four
kinds of dashes.
Access three
ofthis
them with differe
The
source
file
looks
like
A
L TEX knows four kinds of dashes. Access three of them with diff
of consecutive
dashes.dashes.
The fourth
sign issign
actually
not anot
dash
at all
of consecutive
The fourth
is actually
a dash
at
mathematical
minus sign:
mathematical
minus sign:
Code
Printout
Examples of Multiple-Line Equation Structures
⇥⇤
⇥
⇥
⇥
|I1 | = ⇥⇥ gRu d⇥⇥⇥
C3

⇤ ⌥⇤

2

x

g( , t) d

d⇥

a

⌅

⇤ ⌃

u2x

1
+
k

⌥⇤

⇧1/2
2

x

cut d

a

⌅

c⇥

257

(8.49)

⇧1/2

⇥⇥ ⇥
⇥⇥ ⇥
⇥⇥
⇥
⇥⇥ ⇥ 1,0
⇥⇥ ⇥
⇥⇥
A
C4 ⇥⇥f ⇥Sa, W2 (⇥, l )⇥⇥ ⇥|u| ⇤ W2 (⇥; r , T )⇥⇥ .
⇥⇤
⌃
⌅ ⇥
⇤ ⇥
⇤ a
⇥ T
⇥
d⇤
⇥
⇥
⇥(t) u(a, t) ⇥
c( )ut ( , t) d dt⇥
|I2 | = ⇥
⇥ 0
⇥
(t) k(⇤, t) a
⇥
⇥
⇥⇥ ⇤ ⇥
⇥⇥ ⇥
⇥⇥
⇥⇥
⇥
⇥ 1,0
⇥⇥ ⇥
⇥
A
⇥
⇥
C6 ⇥⇥f
⇥Sa, W2 (⇥, l )⇥⇥ ⇥|u| ⇤ W2 (⇥; r , T )⇥⇥⇥ .

ghter-in-law,
X-rated\\
daughter-in-law,
X-rated\\
es 13--67\\
pages 13--67\\
(8.50)
---oryes---or
no? \\ no? \\
, $1$$0$,
and $1$
$-1$and $-1$

8.7

The input for the above formulae is:

\begin{align}
\begin{split}
|I_1| &= \left| \int_\Omega gRu \,d\Omega \right|
&\le C_3 \left[ \int_\Omega \left( \int_{a}^x
g(\xi,t) \,d \xi \right)^2d \Omega \right]^{1/2}
&\quad\times \left[ \int_\Omega \left\{ u^2_x + \frac{1}{k}
\left( \int_{a}^x cu_t \, d\xi \right)^2 \right\}
c \Omega \right]^{1/2}
&\le C_4 \left| \left| f \left| \widetilde{S}^{-1,0}_{a,-}
W_2(\Omega,\Gamma_l) \right| \right|
\left| |u| \overset{\circ} \to W_2^{\widetilde{A}}
(\Omega;\Gamma_r,T) \right| \right|.
\end{split}\label{eq:A}
\\
\begin{split}
|I_2| &= \left| \int_{0}^T \psi(t) \left\{ u(a,t)
-\int_{\gamma(t)}^a \frac{d\theta}{k(\theta,t)}
\int_{a}^\theta c(\xi) u_t(\xi,t) \,d \xi \right\} dt
\right|
&\le C_6 \left| \left| f \int_\Omega
\left| \widetilde{S}^{-1,0}_{a,-}
W_2(\Omega,\Gamma_l) \right| \right|
\left| |u| \overset{\circ} \to W_2^{\widetilde{A}}
(\Omega;\Gamma_r,T) \right| \right|.
\end{split}
\end{align}

daughter-in-law,
X-rated
daughter-in-law,
X-rated
pages pages
13–6713–67
yes—or
yes—or
no? no?
0, 1 and
1
0, 1 and
1

Examples of Multiple-Line Equation Structures

⇥⇤
⇥
⇥
⇥
⇥
⇥
|I1 | = ⇥ are:
gRu d⇥
dashes
‘-’
⇥

The for
names
fordashes
these are: ‘-’ hyphen,
hyphen,
‘–’ en-dash,
The names
these
‘–’ en-dash,
‘—’ ‘—’
em
⇧1/2
⇤ ⌥⇤ x
‘
’
minus
sign.
2
‘ ’ minus sign.
C3
g( , t) d
d⇥

2.4.3

2.4.3

\\

\\

\\

Tilde (≥)

Tilde (≥)

a

⌅

⇤ ⌃

u2x +

1
k

⌥⇤

a

2

x

cut d

⌅

c⇥

⇧1/2

⇥⇥ ⇥
⇥⇥ ⇥
⇥⇥
⇥
⇥
⇥
⇥
⇥
⇥
⇥
⇥⇥
1,0
A
seen in web addresses
tilde. To generate
C4 ⇥⇥f ⇥Sa, W2is
(⇥, the
l )⇥⇥ ⇥|u| ⇤ W2 (⇥; r , T )⇥⇥ .
⇥⇤
⌃
⌅ ⇥
⇤ a you ⇤want.
⇥
T
⇥
result
(˜)
is
not
really
what
Try
th⇥⇥
\\
d⇤
⇥
⇥(t) u(a, t) ⇥
c( )ut ( , t) d dt⇥
|I2 | = ⇥
⇥ 0
⇥
k(⇤,
t)
(t)
a
⇥⇥ ⇤ ⇥
⇥⇥⇥ ⇥
⇥⇥⇥
⇥⇥
⇥
⇥ 1,0
⇥ ⇥
⇥
C6 ⇥⇥⇥⇥f
⇥Sa, W2 (⇥, l )⇥⇥⇥ ⇥|u| ⇤ W2A (⇥; r , T )⇥⇥⇥ .

A character often
A character
oftenbut
seen
use \~{}
thein web addresses is the tilde. To generate thi

use \~{} but the result (˜) is not really what you want. Try this in
http://www.rich.edu/\~{}bush \\

http://www.rich.edu/˜bush

The Metafont concept
.mf file (ascii)
experts or
databases

Metafont

.tfm files

.pk files
scalable graphics

∂α ➜
∜∞ ∄ ⡝
✞‰ ✯
aBc

baseline

font metrics

a f gM

36

As you see, draw is here followed by the coordinates of three rather than two points, also separated
by “..”. This is the very secret of the powerful draw command of METAFONT. You can feed it
with any number of coordinate pairs (always separated two by two with “..”), and it will draw a
smooth curve passing by all those points (in the order you typed them in) as well as it can! Check
Figure 2.5 to see the draw command in action.

Metafont and Bezier lines

draw z 0 ..z 1 ..z 2 ..z 3 ..z 4 ..z 5 ;
(a)

draw z 0 ..z 5 ..z 4 ..z 2 ..z 0 ;
2.4 Pens
(d)

draw z 0 ..z 1 ..z 2 ..z 5 ..z 4 ..z 3 ;
(b)

draw z 0 ..z 5 ..z 2 ..z 1 ;
(c)

draw z 0 ..z 5 ..z 3 ..z 4 ..z 2 ..z 1 ;
(e)

draw z 0 ..z 2 ..z 4 ..z 3 ..z 5 ..z 1 ;
(f)

37

Round tip

45

Figure 2.5: Examples of curves and the commands that drew them.
As you see in those examples, METAFONT doesn’t mind bending the curve quite a lot to get a
very smooth result. That’s because as it is used right now, it will always construct curves where
bending is minimal at the positions of the given points (you can check on Figure 2.5 that when a

Flat tip

2

Note that my use of the words “curve” and “line” is strictly personal and in no way representative of some
metaphysical distinction METAFONT would do between straight lines and bent curves. Indeed, METAFONT doesn’t
make a distinction at all and treats all curves and lines identically.

C. Grandsire

8
The Name of the Ga
Now open the text editor you have chosen to use to write METAFONT
sources with,
38
and possibly (on a Unix or Unix-like platform) some graphics of a letter looking like a
and write the following 26 lines (without the line numbers):

1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6

Metafont example

Greek beta. If you check your directory, you will see that indeed, a beta.2602gf file
has appeared, as well as a beta.log file, which just contains the same text as above.
Carry on and enter now the following line:

u#:=4/9pt#;
gftodvi beta.2602gf
define_pixels(u);
The resulting output is uninteresting, but you should find that you now have also a
beta.dvi file in your directory. View it with your DVI viewer, and you should get
beginchar(66,13u#,16u#,5u#);"Letter beta";
something similar to Figure 1.
x1=2u; x2=x3=3u;
bot y1=-5u; y2=8u; y3=14u;
Greek letter “beta”
x4=6.5u; top y4=h;
z5=(10u,12u);
z6=(7.5u,7.5u); z8=z6;
z7=(4u,7.5u);
z9=(11.5u,2u);
z0=(5u,u);
penpos1(2u,20);
penpos2(.5u,0);
penpos3(u,-45);
penpos4(.8u,-90);
penpos5(1.5u,-180);
penpos6(.4u,150);
baseline
penpos7(.4u,0);
penpos8(.4u,210);
penpos9(1.5u,-180);
penpos0(.3u,20);
Figure 1: The Greek letter .
pickup pencircle;
penstroke z1e..z2e..z3e..z4e..z5e..z6e..{up}z7e..z8e..z9e..{up}z0e;
To understand how METAFONT drew this figure out of the program you fed it with,
labels(range 1 thru 9);
endchar;
The compressed file takes a few bytes
C. Grandsire
end

\fmflabel{$d$}{i2}
\fmfright{o1,o2}
⌦ \Rightarrow
=⌦\fmfleft{i1}
\Longrightarrow
\fmfright{o1,o2} 39
\fmflabel{$\bar{d}$}{o1}
\fmf{photon}{i1,v4}
⌃ \leftrightarrow
⇥⇤
\longleftrightarrow
\fmflabel{$b$}{o2}
\fmf{quark}{o1,v1,v2,v3,v4,v5,v6,v7,
\Leftrightarrow
⌦\fmf{gluon}{v1,v7}
\Longleftrightarrow
\fmf{fermion}{i1,v1}
\fmf{gluon}{v2,v6}
\fmf{fermion,tension=.5,label=$\bar{t}
\fmf{fermion}{vp,oq1}
l.side=right}{v1,v3}
◆⇤
\mapsto
◆ ⇤\fmf{gluon}{v3,v5}
\longmapsto
\fmf{fermion}{vp,oq2}
\fmf{fermion}{v3,o1}
\fmf{photon}{vl,vq}
⇥⌃ \hookleftarrow
⇧⇤
\hookrightarrow
d
b
\fmf{fermion}{o2,v4}
\fmf{fermion}{il,vl,ol}
t, c, u
What
wrong? Obviously the gluons are
\fmf{fermion,tension=.5,label=$t,,c,,u
\leftharpoonup
⇤ went\rightharpoonup
\fmfblob{.15w}{vp}
strongly. The fix is simple: just create a ske
\fmf{fermion}{v2,i2}
\fmfdot{vq,vl}
⇥
\leftharpoondown
⌅
\rightharpoondown
\fmffreeze
\fmf{photon,tension=.2,label=$W^+$,l.s
⌥
\rightleftharpoons
⇡ \fmfleft{i1}
\leadsto
\fmfright{o1,o2}
W − shifted (thick*(0,2))}
W+
\fmfi{plain}{vpath
(__ip,__vp)
\fmf{photon,tension=.2,label=$W^-$,l.s
\fmf{photon}{i1,v4}
\fmfi{plain}{vpath (__ip,__vp) shifted (thick*(1,-2))}
\fmfdotn{v}{4}

Unlimited fancy graphics

\fmf{quark}{o1,v1,v2,v3,v4,v5,v6,v7,

"

\dashrightarrow
#
\dashleftarrow
t¯, c¯, u
¯
\leftrightarrows
\Lleftarrow
¯
As it stands, ball vertices come out too far
right,'
because
the greater
d¯ to the
Thisand
result
is
much
nicer
than the original.
add
the
gluons
later:
number
of
outgoing
lines
pulls
them
over.
Adding
\fmf{phantom}
makes
 \leftarrowtail
\looparrowleft
the bond
the incoming vertices and the interactions
tighter and
✏ between
\curvearrowleft
⌃ \fmfleft{i1}
\circlearrowleft
\fmfright{o1,o2}
produces a better balanced picture: 2.6 Miscellaneous commands
\upuparrows
⌘ \fmf{photon}{i1,v4}
\upharpoonleft
2.6.1 Graphs
in \leftrightsquigarrow
graphs
\fmfleft{ip,il}
\multimap
⇢\fmf{quark}{o1,v1,v2,v3,v4,v5,v6,v7,
\fmffreeze
\fmfright{oq1,oq2,d1,oq3,d2,d3,ol}
\fmf{gluon}{v1,v7}
\rightleftarrows
\rightrightarrows
fmfsubgraph
The fmfsubgraph
environment
contains a subgraph
\fmf{fermion}{ip,vp,vq,oq3}
the \rightarrowtail
rectangle of width !width" and
\twoheadrightarrowbe placed inside
◆\fmf{gluon}{v2,v6}
\fmf{phantom}{ip,vp}
\fmf{gluon}{v3,v5}
\fmf{fermion}{vp,oq1}
left
corner
at
(!x
⌥ \rightleftharpoons
⇣ ", !y"):
\curvearrowright
\fmf{fermion}{vp,oq2}
⇠\fmf{photon}{vl,vq}
\Rsh
\downdownarrows
\begin{fmfsubgraph}(!x
",!y")(!width",!he
Alternatively,
we
can
use
a vanishing tensi
!body"
⇣\fmf{fermion}{il,vl,ol}
\downharpoonright
⇡
\rightsquigarrow
clude the gluons from the layout decisions:

\fmf{phantom}{il,vl}
\end{fmfsubgraph}
\fmfblob{.15w}{vp}
\fmfleft{i1}
\fmfright{o1,o2}
\fmfdot{vq,vl}
⇤
\nleftarrow
⌅four
\nrightarrow
The center and
corners are
available as c, nw, n
\fmf{photon}{i1,v4}
\fmffreeze
⌃
\nRightarrow
etc). Because of the\nleftrightarrow
restrictions on the overall siz
\fmfi{plain}{vpath (__ip,__vp) shifted (thick*(0,2))}
environment will, mainly be useful for pr
\fmfi{plain}{vpath (__ip,__vp) FONT,
shiftedthis
(thick*(1,-2))}

META O T.

8

Miscellaneous symbols

25

Hereinisquestion
a not very
serious
Equivalently, we could add tension to the lines
and we
will application of this
get
result:
⌘the same
\infty

\forall
k \Bbbk

40

Bibliography management

ng
of

33)

34)

35)

ich
t
nce

by
nal
otor
sis
it
als,
in

ree
y

All documentation
for 28,
AMS
X-related products is available in PDF form from
116,117
mer, Opt. Lett.
272TE#2003$.
Chin, Appl. Phys. Lett. 92, 101122 #2008$
the AMS web server as indicated below. If you are reading this handbook on-line,
41 9
The
proof
of
Theorem
1
can
now
be
completed.
Assume
!6"
P.
Del’Haye,
A.
Schliesser,
O.
Arcizet,
T.
Wilken,
R.
!30"
M.
F.
Yanik
and
S.
Fan,
Phys.
Rev.
Lett.
the links for each item should be “live”.

have been developed
; however, high-Q optical microcavities, as
a sensor
transducer,
offer the
potential
greatly
detection
Holzwarth,
T. J. Kippenberg,
Nature
#London$
450,
1214
!31" M. F. Yanik
and S. Fan, Phys. Rev. Lett. 9
isand
constant.
Then
Let itsto
gacv
be enhance
that
39
118
[AFG]#2008$.
User’s
to
AMSFonts,
version 2.2d,
Amer.
Math.
Providence,
sensitivity
.one
Recently,
based
on Soc.,
both
monolithic
!32" Q. Xu, P. Dong,and
and M. Lipson, Nat. Phys.
From
(36)Guide
sees
that sensors
is a polynomial.
RI, 2002. 119
Link at www.ams.org/tex/amsfonts.html
!7"
L.
Matone,
M.
Barsuglia,
F. Bondu,gallery
F. Cavalier,
H. Heitmann, have
!33" Q. Xu, J. Shakya, and M. Lipson, Opt.
equals
By
Proposition
1,
the
microsphere
whispering
transducers
[ALG]
User’s Guide gacv
for theofamsmath Package,
version
2.0, Amer. Math. Soc., been demonand
N.aMan,
Phys.
Lett.and
A
271,
314 #2000$.
#2006$.
Old, APS (American
Physical Society)
plus
polynomial,
is attherefore
also a polynomial.
Providence,
RI, 2002.
Link
www.ams.org/tex/amslatex.html
strated.
It
seems
likely
that
this
will
become
an
important
application
!8"
T. Aoki,
B. 3,
Dayan,ofE.names
Wilcut,
P. Bowen,
T. J. Reviews],
!34"
S.same
Manipatruni,
[ASMR]
Abbreviations
of W.
serials
[reviewed
Mathematical
Footnote
on
the
page C. B. Poitras, Q. Xu, and M
a.e.A. S.inParkins,
By
Proposition
Amer.
Math.
Providence,
www.ams.org/msnhtml/serials.pdf
area
for
these
Likewise,
the
broad
technological
impact
that
Kippenberg,
K.devices.
J.Soc.,
Vahala,
and
H. RI.
J. Kimble,
Nature
#London$
33, 1644
#2008$.
Van Vugt, User’s
M., Hogan,
R.,
R. B. (2008).
[ATG]
Guide
to&AKaiser,
2.2, Amer. Math.
Soc.,(American
Providence,
MS-T
EX, version
APA
Psychological
Association)
443,
671
#2006$.
!35" L. Maleki,
A. B. Matsko,
A. A. Savch
Leadership,
followership,
and
evolution:
Some
lessons
resonant
devices
have
had
at
acoustic,
radio
and
microwave
frequenRI,
2001.
LinkAND
at www.ams.org/tex/amstex.html
from
the
past.
American
Psychologist,
63(3), V.
182-196.
DUMEIGE,
TREBAOL,
FÉRON
PHYSICAL
REVIEW
A 79,
013832
!9"
A.
A.
Savchenkov,
A.
B.
Matsko,
S. Ilchenko,Intext
and reference
L.
Ilchenko,
Opt.and
Lett.
29, 626 #2004$.
(Smith,
1992),
bibliography
at end
CKNOWLEDGMENT
A
doi:10.1037/0003-066X.63.3.182
[ATH]
Using
the
amsthm
Package,
version
2.20,
Amer.
Math.
Soc.,
Providence,
cies Maleki,
suggests
that many
devices
will K. A. Fuller, A. T. R
Opt. Express
15, 6768 other
#2007$. applications for these
!36" D. D.
Smith, H. Chang,
RI,
2004.
Link
at
www.ams.org/tex/amslatex.html
The
author Optical
wouldMicrocavities
like to thank
hisScientific,
colleagues
at the JPL
!1"
K.
Vahala,
#World
Singapore,
Lett. 25,W.
1732
#2000$.
Physical
Review,
!10"
J.
M.
Choi,
R.
K.
Lee,
and
A.
Yariv,
Opt.
Lett.
26, 1236
Boyd,
Phys. Rev.
emerge
in
the
optical
domain.
!A 69, 063804 #2004$
[FAQ]
Frequently Asked Questions for AMS Authors, www.ams.org/authors/

Many different bibliography styles!

Frequency
Standards Laboratory for their suggestions and
2004$.
!25" C.(American
Sauvan,
P. Lalanne,
J.-P.Society)
Hugonin,
Phys. and
Rev.
author-faq.html.
#2001$.
!37"
A. Naweed,
G.and
Farca,
S.
I. Shopova,
APS
Physical
!2" T. J. Kippenberg,
M. Preparation
Spillane, andofK.Papers
J. Vahala,
Rev.
#2005$.
encouragement.
[INL]
Instructions
for
andPhys.
Monographs:
AM165118
S-LA
TEX,
doi:10.1038/nature01939
!11"
K. Totsuka
andS.M.
Tomita,
Opt. Lett.
32, 3197
#2007$.
Phys.inRev.
A 71,
043804 #2005$.
Square
brakets
the text
[11]
Lett. 93,
083904
D. D. Smith, H. Chang, and K. A. Fuller, J. Opt. Soc.
version
2.20,#2004$.
Amer. Math. Soc., Providence, RI, 2004. Link!26"
at www.ams.

!38" K. Totsuka, N. Kobayashi, and M. Tomita, P
!12"V.A.S.Yariv,
Electron.
Lett. 36, 321 #2001$.
A. A. Savchenkov,
A. B.
Matsko,
and L. microstructures:
20, 1967 #2003$.
1.!3"
Gerard, org/tex/amslatex.html
J.Ilchenko,
M. et al. Quantum
boxes as active
probes
for photonic
The pillar
Nature
213904
#2007$.
!13"
M.
C.
Cai,
O.
Painter,
and
K.
J.
Vahala,
Phys.
Rev.
Lett.
85,
74
Maleki,
Phys. Rev. for
Lett.
92,
043903of#2004$.
!27"
J.EK.
Poon,
A. Yariv, IEEE
Photonics Te
[INT]
Instructions
Preparation
Papers and Monographs: AM
S-T
X,Superscript,
ver- J. Scheuer,
EFERENCES
R
and and
bibliography
at end
microcavity
case.
A
pp
l
.
P
h
y
s
.
L
e
tt
.
69,
449–451
(1996).
!39"
Dumeige,
sion 2.2b,
Amer.and
Math.
Soc., Providence,
RI, Chem.
2002. Link at www.ams.
!4" G.#2000$.
Berden,
R. Peeters,
G. Meijer,
Int. Rev. Phys.
Lett.
16,Y.1331
#2004$.Thi Kim Ngan Nguyên, L. Gh
2.
Solomon,
G.
S.,S.
Pelton,
M.of
& Z.
Yamamoto,
Y. Modification
ofProc.
spontaneous
emission
of
aLett.
single
quantum
[1]
D.
Allan,
“Statistics
atomic
frequency
IEEE,
19,
565
#2000$.
!28" I.IEEE
Golub,P.Opt.
31,
507
Féron,
Phys.
Rev.#2006$.
Athe
78,text
013818
!14"
L.W.
V.org/tex/amstex.html
Hau,
E. Harris,
Dutton,
and
C.standards,”
H. Behroozi,
Nature
Square
brakets
in
[11] #2008$.
vol.
Feb.
1966.
P54,
hys.pp.
StM.
a221–230,
t397,
uKhoshsima,
s So594
lidi 178,
341–344
(2000).
!5"dot.
S.#London$
Arnold,
I.
Teraoka,
S. Holler, and F. Voll!29" L. Y.
M.J.Tobing,
D. C. S.V.Lim,
P. Dumon,
R. Baets, andR
!40"
E. Heebner,
Wong,
A. Schweinsberg,
#1999$.
[2]Jewell,
J. A. J.Barnes,
“Atomic
timekeeping and
statistics
of precision A
signal
3.
L. eof
tup-to-date
aCormack
lY.-Q.
. Lasing
characteristics
ofthe
GaAs
microresonators.
pp
l. PhChin,
ys. LeleceAppl.
tt.J.54,
1400–1402
An
list
of TJin,
X-related
publications
(both in
print
and
mer,
Opt.
Lett.
28, Li,
272
#2003$.
Phys.
Lett.
92,J.101122
#2008$.
E
…the
work
(1994).
Jackson,
IEEE
Quantum
Electron. 40,
!15"extensive,
M.
Xiao,
S.-Z.
and
J.
Gea-Banacloche,
Phys.
generators,” Proc. IEEE, vol. 54, pp. 207–220, Feb. 1966.
Harvard
tronic)
onA.pp.32-33)
the
web
page
!6"(1989).
P.Rev.
Del’Haye,
A.
Schliesser,
O.that…
Arcizet,
T. Wilken,
R. T.!30"
M. !41"
F.links
Yanik
S. Fan,and
Phys.
Lett. IEEE
92, 083901
Cormack
(1994,
[3]
J.
A.appears
Barnes,
R.666
Chi,
L.states
S.www.ams.org/tex/publications.html
Cutler,
D. J. Healy,
D. B. Leeson,
E. with
H. J.and
Schmitt
H. Rev.
Zimmer,
Trans.#2
Lett.
74,
#1995$.
to sources
where
these
publications
can be
obtained.
Holzwarth,
and
J.
#London$
450,
!31"
M. F. Yanik
and14,
S. 206
Fan,and
Phys.
Rev. Lett. 93,
McGunigal,
J.audience
A.T.
Mullen,
Smith,
R.
R. F. 1214
C.
Vessot,
G.Singapore,
1994).
Intext
reference
(Smith,
1992),
bibliography
at 173903
end #2
4.…professional
Chang,
K. (ed.)
O
pS.
tiKippenberg,
cAdams,
al (Cormack,
PW.
rocL.
essand
es Nature
inI.M
icrL.
ocaSydnor,
vities (World
Scientific,
1996).
Tech.
#1966$.
!16"
G. T.R.Purves,
C.
G.
Hughes,
Phys.
Rev.
A 74,
M.
R. Winkler,
“Characterization
of frequency
stability,” IEEE Trans.
#2008$.
!32"New
Q. Xu,
P. Dong, and M. Lipson, Nat. Phys. 3, 406 #200
Smith
(1946)
and
(1948)
have
shown…
[Gr]
George
antzer,
into
LATEX, fourth
ed.,Cambridge,
Springer,
York,
5.
Rayleigh,
L. in
SGr¨
cieJones
tific PMore
apers Math
617–620
(Cambridge
Univ.,
1912).
!42"
Z. K. Ioannidis, P. M. Radmore, and I. P. G
023805
#2006$.
Meas.,
vol. IM-20,
pp.
105–120,
Apr. 1971.
!7" Instrum.
L. Matone,
M.
Barsuglia,
F.
Bondu,
F.
Cavalier,
H.
Heitmann,
!33"
Q.
Xu,
J. Shakya, and M. Lipson, Opt. Express 14
2007.
6.
Armani,
D.
K.,
Kippenberg,
T.
J.,
Spillane,
S.
M.
&
Vahala,
K.
J.
Ultra-high-Q
toroid
422
#1988$. on a
!17"
Z.
Shi,
R.
W.
Boyd,
R.
M.
Camacho,
Praveen
Kumar
Vudya
Cox,W.
C., 2002.
What health
care
know about
clean hands.
Nursingbased
today, Spring Issue, microcavity
pp.647-85.
[4]
C.
Lindsey
and Lett.
C.
M.assistants
Chie,
“Theory
of oscillator
instability
and
N.
Man,
Phys.
A
271,
314
#2000$.
#2006$.
[Hi] upon
Nicholas
J. Higham,
Handbook
ofvol.
Writing
for 1652–1666,
the Mathematical
Baron,
D. P.,structure
2008. Business
and the
organisation.
Chester:
Pearson.
functions,”
Proc.
IEEE,
64,99,pp.
Dec. Sciences,
!43" J. Morville,
D. Romanini,
M.M.Chenevier,
Setu,
and
J.
C.
Howell,
Phys.
Rev.
Lett.
240801
#2007$.
chip.
N
a
t
u
r
e
421,
925–928
(2003).
!8" 1976.
T. Aoki,
B. Dayan,
E. Wilcut,
W. P. Bowen,
S. Parkins, T. J.
!34" S. Manipatruni,
C. B. Poitras,
Q. Xu, and
Lipson, Op
second
ed., SIAM,
Philadelphia,
PA,A.1998.
Opt.(American
41, 6980 #2002$.
!18"
J. Scheuer
A.
Yariv,
Lett.
96,to053901
#2006$.
AMS
7.
Painter,
O.
et aVliet
lK.
.and
Two-dimensional
band-gap
laser.
Scie1644
nce Appl.
284,
1819–1821
AT X, mode
Kippenberg,
J. and
Vahala,
andPhys.
H.photonic
J. Rev.
Kimble,
Nature
#London$
33,
#2008$.
[KD]
Kopka
and
W.
Daly,
Guide
Ldefect
fourth
ed., AddisonE theorem
[5]
C.
M.Helmut
Van
P. Patrick
H.
Handel,
“A new
transform
for
M. S.
Shahriar,
G.with
S.
Pati,
R. Tripathi,
M. statistics,”
Messall,
J. Poirson,
F. Bretenaker,
M.
Vallet, andandA
!19"
Wesley,
Boston,
2004.
443,
671
#2006$.
!35" L. !44"
Maleki,
A. B. Matsko,
A. A.
Savchenkov,
(1999).
stochastic
processes
special
applicationV.toGopal,
counting
Mathematical
Society)
[Kn]
E.
Knuth,
The
TE
Xbook,
MA,Ilchenko,
1984. Soc.
Phys.
A,
vol.
pp.A.261–276,
1982.
K.
Salit,
Rev.
ASpatial
75,
053807
#2007$.
Am.
!9"
A.and
A. Donald
Savchenkov,
B.
V. Addison-Wesley,
S. Ilchenko,
andReading,
L. microresonators.
Opt.
Lett.
29,
8.
Taranenko,
V. B.113,
&Phys.
Weiss,
C.
O.Matsko,
solitons
in semiconductor
ISquare
EEE
J. SBebrakets
le14,
ct626
. T2811
op#2004$.
.in #1997$.
the text [Kn]
A
[6]
D.
B.
Percival,
“Characterization
of
frequency
stability:
Frequency[La]
Leslie
Lamport,
L
T
X
:
A
document
preparation
system,
second
ed.,
E Wicht,
Maleki,
Express
15, 6768
#2007$.R.-H. Rinkleff, and K. Dan!36" D. !45"
D. Smith,
H. Chang,S.
K.Trebaol,
A. Fuller,L.
A. Ghişa,
T. Rosenberger,
!20"
M.
Müller,
A.
Y. Dumeige,
T. K. N
Qdomain
uG.
antMüller,
uAddison-Wesley,
mOpt.
E
l
e
c
t
.
8,
488–496
(2002).
estimation of stability
measures,”
Proc.
IEEE,
vol.
79,
pp.
Reading,
1994.
!10" J.zmann,
M. Choi,
R. K.
Lee,
and
A. MA,
Yariv,
Opt. Lett. 26, 1236
W. Boyd,ernier,
Phys. and
Rev.P.A Féron,
69, 063804
#2004$.
Phys.
Rev.
A
56,
2385
#1997$.
J.
Opt.
Soc. Am. B 25
A
961–972,
July
1991.
9.
Barland,
S.
e
t
a
l
.
Cavity
solitons
as
pixels
in
semiconductor
microcavities.
N
a
t
u
r
e
419,
699–702
(2002).
of
abbreviations.
Example:
PRA
/
Phys.
Rev.
A
/
Physical
Review
A
[MG] List
Frank
Mittelbach,
Michel Goossens,
et al., The
L TEX
companion,
second
#2001$.
!37"
Naweed,
G. Haus,
Farca, Waves
S. I. Shopova,
and A.
Rosen
!21"J.
M.
D. Lukin,
M. Fleischhauer,
M. and
O. Scully,
andinstabilities
V. L. VeliH. A.
and Fields
in T.Optoele
[7]
Rutman,
“Characterization
of phase
frequency
in A. !46"

Bibliography software
• Ancient time
• Typeset your bibliography and format it by hand
• Old time
• Typeset your bibliography database
• Call “by name” the documents you cite
• The computer formats the bibliography
• Modern time
• Build your database when searching on the web
• PDF files / Saved searches / Database records
• Compile your bibliography by clicking on the
database

42

43

Building the bibliography
style

extract citations

web

database

local

get citations
text with raw citations

publisher

template

raw text

biblio
style

add formatted citations
text with full bibliography

• Start: a bibliographic item is cited by “name” (label)
• Intermediate: the label is replaced by a full database record
• Final: the record is formatted according the publisher rules

44

Example – Latex & Bibtex
Database file (.bib)

\section{Introduction} Lorem ipsum dolor sit amet,
consectetur adipiscing elit \cite[Arthur64im].
Suspendisse eu nunc velit. Morbi fringilla nibh vitae nulla
fringilla imperdiet. Vestibulum rutrum sem ac lorem
vulputate ac aliquet lacus dignissim \cite{Allan-1966}. Ut
nisl tellus, accumsan ut luctus nec, consequat ac enim.
Duis rhoncus sodales magna, vel luctus lorem tristique
ac. Morbi vulputate, sem nec faucibus imperdiet, mi risus
cursus dui, pellentesque dapibus metus nisi nec turpis
\cite[bibey99mtt]. Donec accumsan iaculis sapien, a
accumsan nisl commodo sed. Fusce sit amet mollis
nunc. In fringilla posuere lectus ac sodales. Vivamus
suscipit condimentum ligula, vel mollis eros condimentum
vel. Pellentesque eget diam quis dolor ultricies mattis.

@Article{Allan-1966,
author = {David W. Allan},
title = {Statistics of Atomic Frequency Standards},
journal = PIEEE,
year = 1966, volume = 54, number = 2,
pages = {221-230}, month = feb,}

Latex does this

Text file (.tex)

@Article{Allred62jrnbs, author = … …}
@Article{Arthur64im,
@Article{barber67mtt, author = … …}
@Article{bibey99mtt, author = … …}

Latex does this
Bibtex does this
Printout (.pdf)
Introduction
Lorem ipsum dolor sit amet, consectetur adipiscing elit [2].
Suspendisse eu nunc velit. Morbi fringilla nibh vitae nulla
fringilla imperdiet. Vestibulum rutrum sem ac lorem vulputate
ac aliquet lacus dignissim [1]. Ut nisl tellus, accumsan ut
luctus nec, consequat ac enim. Duis rhoncus sodales magna,
vel luctus lorem tristique ac. Morbi vulputate, sem nec
faucibus imperdiet, mi risus cursus dui, pellentesque dapibus
metus nisi nec turpis [3]. Donec accumsan iaculis sapien, a
accumsan nisl commodo sed.
References
[1] D. W. Allan, Statistics of Atomic Frequency Standards,
Proc. IEEE 54(2):221, 1966
[2] G. B, Arthur … … …
[3] J. A. Bibey, Jr, … … …

Local bibliography (.bbl)
\bibitem{Allan-1966} … … … …
\bibitem{Arthur64im} … … … …
\bibitem{bibey99mtt} … … … …

Bibliography
style file

Endnote – the citation “machine gun”

45

Instant bibliography in Microsoft Word, Apple Pages, OpenOffice Writer, Mathematica
manuscript

organize

citations
biblio
style
provided by
the publisher

organize
saved
searches

format
citations

organize
biblio
database
.pdf for
reading

manuscript
with biblio

Proprietary product, expensive (≈ $300 / €300)

web
databases
Ovid
PubMed
Web of Science
etc…

• Collect and organize
• Files for reading
• Saved searches
• Citation records
• Images and figures
• etc…

Zotero – Web application

46

• Collect all research in a single, searchable interface
• PDFs, images, audio/video files, screenshots…
• Automatically indexes the full-text library content
• Organizes research into iTunes-playlist-like collections
• Named collections and sub-collections…
• Saved searches
• Create footnotes, endnotes, in-text citations, bibliographies
• Create citations in Word and OpenOffice
• Synchronizes data across many devices

Roy Rosenzweig Center for History and New Media
Funded by the Andrew W. Mellon Foundation, the Institute of
Museum and Library Services, and the Alfred P. Sloan Foundation.

Storage pricing, as in 2012:
100 MB Free
1 GB $20 / year

Citation

Citation
MIT

things to think about when you choose:

Overview of Citation Software at MIT
What programs are your colleagues using? For example, if fellow researchers are all using RefWorks to share
citations, you might consider RefWorks so that you can share your research, too.
Use what you like! There is no one perfect program, so use what you're comfortable with.
You can use more than one software. Records can be transferred from one program to another, so don't feel
limited to choosing only one.

of Citation
MIT

EndNote
Strengths?

nagement

Citation
MIT

MIT
MIT

IT
MIT

on
t Tools

ng Sources

nagement

t us at:
@mit.edu

Cost?

Learning curve?

Excellent for
organizing citations
for papers and
theses
Best option for
major research
projects, because it
offers the most
options for
customization and
formatting
Most widely used
product at MIT
Most output styles
for formatting
Most customizable
Can handle a large
amount of
references

RefWorks

Allows users to
share citations
Good for
organizing
citations for
papers
Web-based
Since MIT has a
site license,
there's no cost

Zotero

Simple download of
records
Good for managing
a variety of
formats, including
web pages
Offers most
functionality in a
free, open-source
product
Downloads records
from with several
databases that
don't work with
EndNote and
RefWorks,
including Factiva
and USPTO

Must purchase client
software, available at MIT
Coop, or at a discount
through Gov Connection;
web version free for MIT
users through Web of
Science

Free via MIT license. See
the RefWorks guide for
information. Sign up for
an individual account on
the RefWorks website.
(MIT only)

Free, open source.

Takes longer to learn, but
not difficult with training

Fairly quick to learn;
many online user guides
and demos

Quick to learn; simple
design, many online user
guides and demos

Download Zotero

47

48

Conference presentation

General advices
• Get information about the audience, and design the presentation for them
• Number your slides, helps to aks questions
• Table of contents goes in the first slide (title page)
• Choice of colors –> in the end, you have very little choice
• Printable at a reasonable cost
• Visible on screen and print
• Visible when you don’t print the background image/fill
• Some colors may disturb (bright red, bright light green…)
• Letters and numbers must be visible
• Avoid too small font in figures and plots
(appropriate size is bigger than in printed material)
• Choice of the program, and portability of your presentation
• Beware of fancy fonts, they may not be printed/shown correctly
• The last slide should summarize the results you are most proud of
(don’t show a “thank you” slide)
Slides reflect your personality, and shows it in public
Doing some experiments pays back. When you are under pressure is too late

49

Other advices
• Take a look at the conference room
• Can you use your computer?
• Is a microphone necessary or mandatory?
• Is the position comfortable?
• Is a power outlet available and compatible with your computer?
• Check on laser pointer and battery
• Save a PDF copy of your presentation in a USB key
• Learn by hart the slide order / have a plan for long presentations
• Beginners: practice / record your presentation
• Tradeoff between preparation and improvisation
• A learned-by-hart presentation is deadly boring
(often seen at PhD defenses)
• Good improvisation catches the attention, at some risk
• Experience: you know what can be improvised and what cannot

50

Planning a long presentation
Example
Enrico Rubiola

Fortef 28 March 2012

Frequency stability, phase noise etc
tot
40

dur
40

50
70

10
20

100

30

115
120

15
5

no.
2-26

subject
Spectral analysis
3-6
General
8-15 Fourier
17-26 Nice technical issues
28-33 Phase noise & friends
34-50 Noise in amplifiers and components
36–50 Flicker and white
Leeson effect
67-74 Heuristic approach
76-95 Analysis of some oscillators
96-105 Cross spectrum
107-108 Conclusions

Experimental methods
tot
15
25
40
55
65
75
85
90

dur
15
10
15
15
10
10
10
5

no.
4-10
12-15
17-27
29-42
44-48
50-64
66-71
72

subject
Saturated mixer
Correlation measurements
Oscillator phase noise
Photonic techniques
Calibration
Bridge techniques
AM noise
Conclusion/IMP

51

Advices
• The plan of a long talk cannot
be learned by hart
• Depending on the public, you
may need to slow down
• Divide the available time in
slots
• Keep the slots on schedule
• Add/skip slides within the slot
• Private next-slides preview
helps a lot

52

To be continued

Speaker’s private screen (Keynote)
current slide

next slide

clock
speaker's annotation

presentation time

53

presentation time

next s
lide

previo
us slid
e

Speaker’s private screen (Powerpoint)

r
r
cu
speaker's annotation

t
n
e

e
d
i
l
s

54

Speaker’s private screen (OpenOffice)
The private screen is totally independent of the presentation

55

Latex presentations
Tex/Latex generate outstanding PDFs. Why not for a presentation?

• The strength and the weakness of Latex is that it
hides the layout
• Difficult to figure out what the final result look like
• Available packages and styles
• SliTex is gone
• Foiltex (IBM) is (one of) the simplest to use
• Elderly, no longer maintained
• Beamer is by far the most used
Tex/Latex generate outstanding PDFs. Why not for a presentation?

56

57

Beamer – example
Introduction
Practical use of the Allan variance
Statistics of the Allan variance and the Allan deviation
Prediction of very long term time stability

A statistical estimator as well as a spectral analysis tool
Practical calculation of the Allan variance
Allan variance versus Allan deviation

A spectral analysis tool
as well as a statistical estimator

Convolution product in the time domain. . .
⇤
⇥2 ⌅
+⇥

⇥y2 (⇤ ) =
⇧
⌥

y (t)hy (tk

⇥

hy (t) = ⌅

1

2⇤
+1
with
hy (t) = ⌅
2⇤
⌃
hy (t) = 0

t)dt

if

⇤ ⇥t <0

if

0⇥t <⇤

else

. . . filtering in the frequency domain
⇥y2 (⇤ ) =

⇥
0

2

Sy (f ) |Hy (f )| df

4
sin
( ⇤f)
2
2
with |Hy (f )| = |F.T. [hy (t)]| = 2
( ⇤ f )2
F. Vernotte

Variance measurements

13