A fast scan submillimeter spectroscopic technique

TLDR: In this paper, a fast scan submillimeter spectroscopic technique (FASSST) was developed which uses a voltage tunable backward wave oscillator (BWO) as a primary source of radiation, but which uses optical calibration methods rather than the more traditional phase or frequency lock techniques.

Abstract: A new fast scan submillimeter spectroscopic technique (FASSST) has been developed which uses a voltage tunable backward wave oscillator (BWO) as a primary source of radiation, but which uses fast scan (∼105 Doppler limited resolution elements/s) and optical calibration methods rather than the more traditional phase or frequency lock techniques. Among its attributes are (1) absolute frequency calibration to ∼1/10 of a Doppler limited gaseous absorption linewidth (<0.1 MHz, 0.000 003 cm−1), (2) high sensitivity, and (3) the ability to measure many thousands of lines/s. Key elements which make this system possible include the excellent short term spectral purity of the broadly (∼100 GHz) tunable BWO; a very low noise, rapidly scannable high voltage power supply; fast data acquisition; and software capable of automated calibration and spectral line measurement. In addition to the unique spectroscopic power of the FASSST system, its implementation is simple enough that it has the prospect of impacting a wide range of scientific problems.

Figures

FIG. 5. 3 s scan of 40 GHz of the mm/sub mm spectrum of HNO3 showing four bandheads. The signal amplitude is adjusted to fill the 4048~12 bit! steps of the A/D converter.

FIG. 7. A graphical blowup of the region shown in the box in Fig. 6. The spectrum has been convolved with a first derivative line shape, resulting in line shapes which are approximately second derivative. The ‘‘noise’’ in the figure results from the digital noise of the A/D converter and is not a fundamental limit on the detection of weak lines.

FIG. 8. Portion of a true line shape spectrum recorded for use in linewidth measurements.

TABLE III. A comparison between pressure broadening parameters of HNO3 measured using conventional phase lock techniques and those obtained with the FASSST system.

TABLE I. Example of the line list for a portion of the spectrum shown in Fig. 5.

TABLE II. Rotational constants for then7 vibrational state of HNO3(MHz).

Full text

Wright State University Wright State University
CORE Scholar CORE Scholar
Physics Faculty Publications Physics
4-1997
A Fast Scan Submillimeter Spectroscopic Technique A Fast Scan Submillimeter Spectroscopic Technique
Douglas T. Petkie
Wright State University - Main Campus
, dpetkie@yahoo.com
Thomas M. Goyette
Ryan P. A. Bettens
Sergei P. Belov
Sieghard Albert
See next page for additional authors
Follow this and additional works at: https://corescholar.libraries.wright.edu/physics
Part of the Physics Commons
Repository Citation Repository Citation
Petkie, D. T., Goyette, T. M., Bettens, R. P., Belov, S. P., Albert, S., Helminger, P., & De Lucia, F. C. (1997). A
Fast Scan Submillimeter Spectroscopic Technique.
Review of Scienti>c Instruments, 68
(4), 1675-1683.
https://corescholar.libraries.wright.edu/physics/786
This Article is brought to you for free and open access by the Physics at CORE Scholar. It has been accepted for
inclusion in Physics Faculty Publications by an authorized administrator of CORE Scholar. For more information,
please contact library-corescholar@wright.edu.

Authors Authors
Douglas T. Petkie, Thomas M. Goyette, Ryan P. A. Bettens, Sergei P. Belov, Sieghard Albert, Paul Helminger,
and Frank C. De Lucia
This article is available at CORE Scholar: https://corescholar.libraries.wright.edu/physics/786

A fast scan submillimeter spectroscopic technique
Douglas T. Petkie, Thomas M. Goyette, Ryan P. A. Bettens, S. P. Belov,
a)
Sieghard Albert, Paul Helminger,
b)
and Frank C. De Lucia
Department of Physics, The Ohio State University, Columbus, Ohio 43210
~Received 19 September 1996; accepted for publication 25 November 1996!
A new fast scan submillimeter spectroscopic technique ~FASSST! has been developed which uses
a voltage tunable backward wave oscillator ~BWO! as a primary source of radiation, but which uses
fast scan (;10
5
Doppler limited resolution elements/s! and optical calibration methods rather than
the more traditional phase or frequency lock techniques. Among its attributes are ~1! absolute
frequency calibration to ;1/10 of a Doppler limited gaseous absorption linewidth
(, 0.1 MHz, 0.000 003 cm
2 1
), ~2! high sensitivity, and ~3! the ability to measure many thousands
of lines/s. Key elements which make this system possible include the excellent short term spectral
purity of the broadly (;100 GHz) tunable BWO; a very low noise, rapidly scannable high voltage
power supply; fast data acquisition; and software capable of automated calibration and spectral line
measurement. In addition to the unique spectroscopic power of the FASSST system, its
implementation is simple enough that it has the prospect of impacting a wide range of scientific
problems. © 1997 American Institute of Physics. @S0034-6748~97!02503-3#
I. INTRODUCTION
A. Applications and impact of millimeter and
submillimeter spectroscopy
High resolution millimeter and submillimeter ~mm/
submm! spectroscopy has had a major impact on many im-
portant fields of science and technology. The earliest studies
in this region were of small, fundamental species such as
H
2
O, O
2
, NO, CH
3
F, and OCS and served to both estab-
lish spectroscopic methodologies and to provide basic infor-
mation about molecular structure and interactions.
1–3
Be-
cause these small, fundamental species have intrinsically
interesting collisional properties, their dynamical properties
have been studied as well. These studies have ranged from
investigations of pressure broadening near room temperature
~which are fundamental to the deconvolution of atmospheric
remote sensing data!
4–6
to basic studies of the quantum na-
ture of molecular collisions at low temperature.
7,8
Because the strength of the interaction between electro-
magnetic radiation and molecular rotation peaks sharply in
the mm/sub mm region, a variety of spectroscopically based
remote sensing applications have grown out of this more
basic work. The least remote of these have involved labora-
tory studies of molecular lasers and the collision induced
rotational and vibrational processes which are central to their
operation.
9–11
This spectral region has also played an impor-
tant role in the study of the chemical processes in the upper
atmosphere which are important in ozone formation and
destruction.
12–14
Finally, the vast majority of the over 100
molecular species which have been identified and studied in
the interstellar medium have been observed by means of
mm/sub mm ‘‘radio’’ astronomy.
15–18
B. Current spectroscopic practice: The source and
detector problem
In spite of these and other applications, the mm/sub mm
spectral region is by far the least explored portion of the
electromagnetic spectrum, largely because of the difficulty of
generating and detecting radiation at these wavelengths.
However, over the years, a number of approaches have been
developed. High resolution spectroscopy was first extended
into the submillimeter spectral region by means of nonlinear
harmonic generation
3
and the technique extended by the in-
troduction of sensitive helium temperature detectors
19
and
improved harmonic generator design.
20
Nonlinear techniques
have also been used to generate difference frequencies be-
tween optical lasers
21–23
and to produce microwave side-
bands on far infrared ~FIR! laser sources.
24,25
Additionally, a
number of solid state diode sources have been extended to
higher frequency.
26,27
Of all of the techniques, those most closely related to the
fast scan submillimeter spectroscopic technique ~FASSST!
system described here are also based on backward wave os-
cillator ~BWO! tubes. In 1953, Kompfner and Williams dem-
onstrated that self-sustained oscillation in traveling wave
structures could be achieved by the interaction between the
oppositely directed group and phase velocities.
28
Chief
among the attributes of these devices is the broad band elec-
trical tunability which results from the dispersion relation of
this backward wave interaction. Among the most successful
implementations of this concept were the carcinotrons of
Tompson CSF
29
and the BWOs of the ISTOK Research and
Development Company of Fryazino, Moscow Region,
Russia.
30,31
The latter are particularly advantageous for spectroscopy
because of their broad bandwidth, good spectral purity, high
frequency capability, and relatively small power consump-
tion. The ISTOK BWOs have been used successfully in free-
running spectrometers,
32,33
and in synthesized phase-locked
systems.
34–36
A large amount of important submillimeter
a!
Present address: Physikalisches Institut, Universitat zu Koln, D-50937 Co-
logne, Germany.
b!
Present address: Department of Physics, University of South Alabama,
Mobile, AL 36688.
1675Rev. Sci. Instrum. 68 (4), April 1997 0034-6748/97/68(4)/1675/9/$10.00 © 1997 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:
130.108.169.222 On: Tue, 09 Dec 2014 15:26:59

spectroscopy has been accomplished with these systems, and
interested readers are referred to the cited references for de-
tails.
C. A new spectroscopic system for the mm/sub mm
spectral region
This article describes a new high resolution spectro-
scopic system for the mm/sub mm spectral region, which is
fast, broadband, sensitive, and simple. It is based on broad-
band, voltage tunable BWOs produced by the ISTOK Re-
search and Development Company of Fryazino, Moscow Re-
gion, Russia and uses very fast sweep (;10
5
Doppler
limited spectral resolution elements per second! and optical
calibration methods to replace the phase and frequency lock
techniques more commonly used.
Because the fast scan effectively ‘‘freezes’’ frequency
instability on the time scale of the optical calibration period,
resolution, and frequency measurement accuracy are compa-
rable to that of the much slower and more complex phase/
frequency lock systems. This FASSST approach makes it
possible to fully utilize the inherent instantaneous band-
width of the BWOs and to achieve a spectroscopic system of
unprecedented capabilities. Furthermore, the resultant system
is much simpler and has the potential for wide application.
II. SYSTEM OVERVIEW
A. The basic spectroscopic system
Figure 1 shows a block diagram of the FASSST system.
In this system, an ISTOK OB-30 is used to cover the 240–
375 GHz region. Similar tubes are available from ISTOK for
the ;100–1000 GHz region. The first wire grid polarizer
~WG1! provides a well defined polarization from the output
of the overmoded BWO waveguide. The second polarizer
~WG2! is used to split the output power of the BWO, with
;90% being directed quasioptically through the molecular
absorption cell and detected by an InSb hot electron bolom-
eter operating at 1.5 K. The remaining ;10% of the power
is coupled into a Fabry–Perot ~FP! cavity via a mylar beam-
splitter ~BS1!, which provides fringes for frequency interpo-
lation between reference spectral lines of known frequency.
In order to provide a highly accurate basis for the analysis of
the frequency-voltage characteristic of the BWO, a folded FP
cavity of length ;38.89 m is used to provide modes every
;3.854 MHz. The use of more compact FP cavities in sub-
sequent versions of this system will be discussed below. Pro-
vision for a second molecular absorption cell which can be
used for calibration purposes is also provided.
B. The key system elements
The key system elements include:
~1! The most fundamental element is the excellent short
term spectral purity of the BWO. From studies over
many years, it has been observed that the short-term
spectral purity of free-running ISTOK BWOs is
;10 kHz.
37
Without this spectral purity, the FASSST
system would not be possible.
~2! Second, the BWOs can be voltage tuned continuously
over an ;50% frequency range, which contains ;10
5
spectral resolution elements ~Doppler limited!.
~3! The synthesized frequency reference system typical of
high resolution submillimeter spectrometers is replaced
by a system more typical of optical spectroscopy. How-
ever, the longer wavelength significantly relaxes the re-
quirements for optical precision and much greater fre-
quency accuracy can be achieved.
~4! A fast (;10
5
spectral resolution elements/s, currently
limited by detector bandwidth! sweep and data acquisi-
tion system freeze any drift in the source frequency over
the time required to sweep from one reference fringe to
the next. This eliminates the need for active frequency
stabilization.
~5! Fast data acquisition and calibration hardware and soft-
ware. In a very general sense, the bandwidth of this sys-
tem plays the same role as the bandwidth of the lock
loops of more traditional systems.
C. System attributes
The combination of these five elements make it possible
to measure thousands of spectral lines per second, with a
frequency accuracy of a small fraction of a Doppler width
(;0.1 MHz/3310
2 6
cm
2 1
). Signal averaging is straight-
forward, and for equivalent integration times the sensitivity
is the same as for slow-sweep, synthesized phase locked sys-
tems. Finally, the system is very simple in both concept and
execution and holds the promise of being used in a wide
variety of applications.
III. EXPERIMENTAL DETAILS
A. Tube characteristics
The BWO is an electron beam device whose frequency
depends on the interactions among the electrons, the period-
icity of the BWOs slow wave structure, and the electromag-
netic radiation propagating along with the electron beam in
the slow wave structure. These are often described in the
context of backward wave space harmonics and dispersion
relations.
38
Because the electron velocity has a first order
effect in these relations, the frequency of oscillation f is
strongly dependent on the electric potential V between the
FIG. 1. Block diagram of the FASSST system.
1676 Rev. Sci. Instrum., Vol. 68, No. 4, April 1997 Submillimeter fast scan
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:
130.108.169.222 On: Tue, 09 Dec 2014 15:26:59

cathode and slow wave structure. For an OB-30,
df/dV;75 MHz/V. This increases somewhat with fre-
quency and for tubes centered around 1 THz is of the order
;100 MHz/V. For a given tube, df/dV typically decreases
with increasing voltage, largely because of the quadratic re-
lation between electron energy and velocity. Figure 2 shows
the functional dependence of frequency as well as df/dV on
voltage for the OB-30 used in this work.
Because the slow wave structure does not have a reso-
nant frequency, BWOs can be very broadbanded. However,
because of this low Q, feedback in the form of power re-
flected back into the slow wave structure can dramatically
alter the frequency and power output of the BWO. In a
broadband system in which fast, broad sweeps are executed,
it is not possible to control these reflections by narrow band
matching techniques, and it is important that the reflections
be carefully controlled and minimized. In fact, spontaneous
locking to the FP etalon used as a frequency reference is not
uncommon unless steps are taken to significantly reduce
feedback.
The thermal drift of the frequency of the BWO can be
;500 MHz/K due to the thermal expansion of the slow
wave structure. However, because the time constant for this
drift is of the order of a few seconds ~a time long in com-
parison to the interval between FP modes in the fast scan!,it
is not necessary to thermally stabilize the tube. This is for-
tunate because the power input to the BWO varies by more
than a factor of 2 during the voltage sweep and external
thermal control would be of little utility.
Finally, vibrations which change the orientation of the
magnetic field which guides the electron beam relative to the
slow wave structure of the BWO can modulate both the fre-
quency and power output. In our experience, such vibrations
can be reduced below their observational threshold by
straightforward techniques in an ordinary laboratory environ-
ment.
B. The frequency calibration scheme
If the frequency of the BWO as a function of voltage
were linear, the frequency calibration scheme could be as
simple as having two spectral lines of known frequency
somewhere near the opposite ends of the frequency sweep
range and using linear interpolation. However, from Fig. 3, it
is immediately obvious that the small scale structure is much
more complex than given by a simple dispersion relation.
More specifically, if the effects of the small scale structure
are not properly treated, the frequency accuracy of the
FASSST system will be ;100 MHz, approximately 1000
times worse than that required for high resolution spectros-
copy. However, we will demonstrate below that with a small
FP mode spacing we have been able to use simple linear
interpolation methods to measure line frequencies to ;1/10
of the Doppler limited linewidth.
Based on these considerations, the basic FASSST
scheme is to:
~1! take a fast (10
4
–10
5
MHz/s) scan over the spectral re-
gion of interest,
~2! include two or more ~typically ; 50 are available! ref-
erence lines,
~3! use the known frequencies of the reference lines to de-
termine the FP cavity mode spacing and absolute fre-
quency,
~4! count FP modes to establish the frequency of each
fringe, and
~5! use linear interpolation between the two nearest FP
modes to calculate the frequencies of the unknown lines.
Because the thermal history of the BWO effects the
frequency-voltage function at this level of precision, each
sweep is calibrated separately.
C. The power supply scheme
Ultimately the combination of high spectral purity and
voltage tunability of these BWOs makes the system de-
scribed in this article possible. However, these qualities can
only be exploited in the context of a power supply with
complimentary properties, a stringent requirement. More
specifically, the 240–375 GHz OB-30 BWO tunes ;100
GHz with a voltage variation of ;1000–3500 V, a tuning
rate of ;75 MHz/V. If it is desired that the voltage fluctua-
tions of the power supply result in a frequency variation
which is no more than 10% of a linewidth (;0.1 MHz),
they must be less than ;2 mV in the context of a power
supply capable of sweeping several thousand volts in
;1 ms.
FIG. 2. Dependence of frequency on voltage ~left scale, points! and voltage
sensitivity ~right scale, crosses! for an OB-30 BWO.
FIG. 3. Typical small scale structure of the frequency-voltage characteristic
of a BWO.
1677Rev. Sci. Instrum., Vol. 68, No. 4, April 1997 Submillimeter fast scan
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:
130.108.169.222 On: Tue, 09 Dec 2014 15:26:59

Frequently asked questions

Q1. What are the contributions in "A fast scan submillimeter spectroscopic technique" ?

In this paper, the authors proposed a new kind of mm/sub mm spectrometer of unique and powerful capabilities. 

Q2. What can be done to reduce the frequency of the BWO?

vibrations which change the orientation of the magnetic field which guides the electron beam relative to the slow wave structure of the BWO can modulate both the frequency and power output. 

Q3. Why is the spectral line shape distorted in experiments designed to optimize S/N?

Because it is easiest to maximize dynamic range by insertion of a low fre-quency roll off between the mm/submm detector and the A/D input to the computer, the spectral line shape is distorted in experiments designed to optimize S/N. 

Q4. What is the sweep rate for a FP cavity?

in many applications only a small portion of the 100 GHz avail-able will be scanned, but the same sweep rate in terms of FP cavity modes/s is required in order to eliminate drift and ripple effects. 

Q5. Why is nitric acid studied extensively in the infrared?

Nitric acid has been studied extensively in the mm/ submm and the infrared because of its fundamental role in chemistry and because it is an important component in the ozone destruction cycle in the upper atmosphere. 

Q6. How many FP modes can be used to determine the frequency of each fringe?

reference lines, ~3! use the known frequencies of the reference lines to determine the FP cavity mode spacing and absolute frequency, ~4! count FP modes to establish the frequency of each fringe, and ~5! 

Q7. How did they achieve self-sustained oscillation in traveling wave structures?

In 1953, Kompfner and Williams demonstrated that self-sustained oscillation in traveling wave structures could be achieved by the interaction between the oppositely directed group and phase velocities. 

Q8. Why is it not necessary to thermally stabilize the tube?

because the time constant for this drift is of the order of a few seconds ~a time long in comparison to the interval between FP modes in the fast scan!, it is not necessary to thermally stabilize the tube. 

Q9. How many FP cavity modes are needed to be scanned?

As an example, if 20 FP cavity modes (;100 MHz) are scanned rather than the 20 000 in the full 100 GHz, the 1 s sweep period can be reduced to 1/1000 of a second. 

Q10. How can the authors achieve absolute calibration of the BWO?

Because these amplitude variations are fundamental properties of the BWO and are stable in time, absolute calibration can be achieved by recording a chopped sweep alternatively with a spectral scan. 

Q11. What are the advantages of averaging multiple sweeps?

Although averaging multiple sweeps to increase S/N is possible, the inherent source brightness, detector sensitivity, and strength of the molecular interaction render this unnecessary in many applications. 

Q12. How long does it take to scan from one FP cavity to the next?

For the optical calibration scheme to be accurate, the BWO frequency drift must be linear over the time required to scan from one FP cavity mode to the next. 

Q13. How much ripple will be reduced between FP markers?

contribution to the voltage change between FP markers will be reduced from Vac by the ratio of marker rate of the ripple rate ~a factor of 200 in these examples!. 

Q14. How is the frequency accuracy of the FASSST system affected?

More specifically, if the effects of the small scale structure are not properly treated, the frequency accuracy of the FASSST system will be ;100 MHz, approximately 1000 times worse than that required for high resolution spectroscopy. 

Q15. How would the current linear method reduce the cavity length?

Since the current linear method introduces little error, it would appear that either of these approaches would lead to a reduction of at least a factor of 5 in the required cavity length. 

Q16. How many species have been identified and studied in the interstellar medium?

12–14 Finally, the vast majority of the over 100 molecular species which have been identified and studied in the interstellar medium have been observed by means of mm/sub mm ‘‘radio’’ astronomy. 

Q17. Why is it not possible to display a full spectrum?

Because a single scan contains ;106 resolution elements, it is not possible to graphically display a complete, full band spectrum.