REVIEW OF SCIENTIFIC INSTRUMENTS
VOLUME 73, NUMBER 3
MARCH 2002
A small volume, rapid translation cryostat insert constructed from
commercial components for the detection of ultrafast optical signals
Delmar S. Larsena)
Faculty of Sciences, Division of Physics and Astronomy, Department of Biophysics and Physics of Complex
Systems, Vrije Universiteit Amsterdam, De Boelelaan 1081, 1081 HV Amsterdam, The Netherlands
共Received 10 September 2001; accepted for publication 17 December 2001兲
A linear motion device was designed and built to move small sample cells within the confined space
of a liquid nitrogen Dewar cryostat. Instead of the often-used methods of flowing or rotating
samples to generate motion, this cryostat insert simply translates the sample cell while maintaining
atmospheric isolation. Accurate, repeatable and rapid translation over a 5 cm range with peak linear
velocities of 1 m/s is attainable. The insert is constructed mainly from commercially available
components and can be built with minimal effort. Another benefit is that the cryostat insert uses
commercially available sample cells, allowing for maximal flexibility in satisfying pathlength and
volume requirements and the cells can be easily replaced during the duration of the experiment. The
motion device system can be easily installed into existing experimental apparatuses with confined
volumes with minor modifications and can be used with liquid helium cryostats or for use in high
vacuum applications. This motion is demonstrated by collecting temperature dependent ultrafast
signals from the laser dye Rhodamine 640 in viscous glycerol at 170 and 293 K. © 2002 American
Institute of Physics. 关DOI: 10.1063/1.1448904兴
temperature liquid samples, to limit these detrimental contributions to the collected signals 共e.g., flowing sample or rotating cells兲. Although, these methods have their own problems; dead volume within mechanical pumps may waste or
damage expensive or difficult to obtain sample and the potential exists 共especially for biological systems兲 for macromolecular aggregates 共e.g., proteins and aggregated chromophores兲 to ‘‘stick’’ to the surface of the cell. In contrast,
rotating cells allow the use of smaller sample volumes, but
require the use of bulky machinery to generate the rotational
motion. Additional effort is also required to ensure high quality, flat cell surfaces, with constant pathlengths to reduce
scattering off uneven surfaces or variable signal intensities
from uneven pathlengths, and any movement along the direction of the propagating laser beams will similarly affect
the measured signals. Neither flowing the sample nor rotating the sample works effectively in the confined spaces
within cryostats, where the collection of many time-resolved
signals are often desired. In these cases, experimentalists
have previously been forced to measure signals without the
benefits of sample motion. This note describes the modifications made to a commercial OXFORD® liquid nitrogen
Dewar cryostat 共Optistat-DN®兲 to achieve rapid motion of
the sample cell at variable temperatures.
The limited space within the OXFORD® cryostat precludes the use of flow cells 共due to high viscosity of the
sample at decreased temperatures兲 or rotating cells, and their
required machinery, for use in data collection. A simple solution is to use an external translational motion system consisting of 共1兲 an external air actuated linear feedthrough assembly that generates and controls sample cell motion, and
共2兲 an internal cryostat insert that mounts directly in the inner vacuum chamber of the cryostat to transfer high preci-
Variable temperature studies allow researchers to study
additional dynamic properties not evident with single temperature measurements. Such studies also offer the possibility of observing interesting dynamics that room temperature
effects may obscure. For example, features such as the zerophonon lines in absorption spectra are often observable in
low temperature systems, but are commonly blurred by homogeneous broadening at room temperature.1 This broadening also obscures important information about the degree of
disorder in biological systems,2 and energy transfer dynamics where broad, overlapping spectra limit the selectivity of
exciting of certain transitions. The use of thin sample cells
with well-characterized, small 共⬃100 ␮m兲 pathlengths is integral to nonlinear, time-resolved spectroscopic studies,
where slow dynamic processes 共longer than the inverse repetition rate of the laser兲 may accumulate from one shot sequence to the next. Such accumulated effects can lead to
undesired contributions to the measured signals. These accumulated effects include: thermal gratings,3 long-lived triplet
states,4 and ground or excited state conformational states.4
Many photoactive systems have additional long-lived dynamical processes such as isomerization 共e.g., stilbene5 and
photoactive yellow protein兲6 or time-consuming steps in the
photocycles 共diffusion-limited steps,7 partial denaturation,6
and proton transfer reactions6兲 that require extended time for
returning to the initial nonreacting state before probing with
another pulse sequence. Repeated excitation of sample volumes may result in permanent photodamage and increased
scatter due to the presence of the damaged sample within the
optical pathlength.
Well-established methods do exist, in the case of room
a兲
Electronic mail: dslarsen@nat.vu.nl
0034-6748/2002/73(3)/1325/4/$19.00
1325
© 2002 American Institute of Physics
Downloaded 03 Apr 2002 to 130.37.36.180. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/rsio/rsicr.jsp
1326
Rev. Sci. Instrum., Vol. 73, No. 3, March 2002
Delmar S. Larsen
TABLE I. Summary of commercial parts.a
Part
Linear ball bearing
Ground rod stock
Sleeve coupling
Pneumatic linear motion feedthrough
w/24 V dc air control solenoid valve
ISO KF kwik-flange reducer
NW16 flange assembly
w/flanges, clamp, ring, O-ring
NW25 flange assembly
Reference number
Part number
ALMB-1
ASI-12
ACT-3
K075-ABLM-275-2
687 000
684 001
684 001
662 008
K100-x075
K050-CA
732 000
700 000
K100-CA
700 002
a
With the exception of the homemade retaining plugs and the stationary rod
mount, all parts were purchased from MDC Vacuum Products Corporation.
sion motion to the sample cell mount. With the exception of
the internal cryostat insert housing, all the components used
in the construction of the cryostat insert are commercially
available 共Table I兲.
The feedthrough assembly, displayed in Fig. 1, controls
the translation of the stock rod, and consequently, the sample
cell. The air actuated linear feedthrough generates translational motion by applying a pressure difference between two
air ports. When a pressure differential is applied, the stock
rod translates to equalize the applied pressure. Motion in the
opposite direction is then generated by applying an opposite
pressure differential. Applying an oscillating pressure difference can thus produce an oscillating translational motion. A
pressure control solenoid can generate this oscillating pressure differential when driven by an electric wave form function generator. The frequency and magnitude of the motion is
controlled with the wave form generator and input pressure
to the solenoid respectively, whilst increasing the air pressure
into the solenoid increases the range of motion for the
sample cell; increasing the wave form frequency increased
the frequency of oscillation. The tuning of these two features
allows for the generation of smooth and continuous motion
of the sample cell. A frequency of 10 Hz and a translation of
5 cm half per cycle results in an average velocity of 1 m/s for
the sample cell. For comparison, a 1 kHz train of laser pulses
focused to a 50 ␮m diameter in the sample requires a cross
beam velocity of 5 cm/s to insure the illumination of a new
sample volume between successive laser pulses.
The linear feedthrough assembly alone generates rapid,
oscillating linear motion, however consistently reproducible
motion is imperative for data collection with the thin sample
cells 共⬃100 ␮m兲 often used in nonlinear measurements. The
internal cryostat assembly, displayed in Fig. 2, consists of
two components: a stationary housing and a mobile rod. The
mobile rod is connected directly to the sample cell mount
and the linear feedthrough, and translates during operation.
The stationary housing braces the mobile rod and insures the
precision required to maintain the sample within the spatial
overlap of the laser pulses during translation. To maintain
this accuracy, two linear bearings are used to guide the rod
共Table I兲; one located near the sample cell mount and the
other near the neck of the cryostat and are separated by approximately 25 cm.
Flow holes 共for the cryostat contact gas兲 were drilled
into the sides of the stationary rod mount to allow for effi-
FIG. 1. Linear air-actuator pneumatic feedthrough that generated the linear
motion. The sample cell is shown mounted directly to the moving rod for
illustration, and is connected to mobile rod in the internal assembly when
used in a cryostat.
cient cooling of the cryostat chamber in the presence of the
insert. All components within the cryostat, including the cell
mount, are constructed of stainless steel to reduce complications from thermal expansion. Because the high thermal conductivity of stainless steel leads to a faster loss of liquid
nitrogen in the cryostat Dewar, a newer version of the insert
is currently under construction using Teflon© instead of
stainless steel for the stationary housing. Due to the thermal
properties of Teflon©, the loss of liquid nitrogen will be
greatly reduced. The coupling between the linear
feedthrough and the internal housing is covered with a highvacuum stainless steel bellows. Use of a stainless steel bellows maintains an airtight barrier between the contact gas at
cryostat temperatures and the room temperature atmosphere
which also eliminates moisture contamination and reduces
thermal leakage. The motion device can be adapted for used
in low-vacuum chambers since the steel bellows can maintain a vacuum of 10⫺9 mbar. No evidence of leaks has been
Downloaded 03 Apr 2002 to 130.37.36.180. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/rsio/rsicr.jsp
Rev. Sci. Instrum., Vol. 73, No. 3, March 2002
Notes
1327
FIG. 3. Preliminary temperature dependent echo signals for Rhodamine 640
in glycerol, measured at 170 and 293 K. 共a兲 Zero-population time echo
profiles for moving sample 共solid line兲 and for stationary sample 共dotted
line兲. Temperature was 293 K. 共b兲 3PEPS profiles at 170 K 共solid circles兲
and 293 K 共solid triangles兲. Insert displays the collected 3PEPS signals on a
longer 共log兲 timescale.
FIG. 2. Internal assembly design includes two linear bearings locked with
retraining plugs into a home-built housing and a mobile rod to transfer
motion to the sample cell mount. The insert is attached to the cryostat and
the pneumatic feedthrough via a Kwik Flange© 24 O-ring connection. Insert
is shown inside the inner vacuum chamber of the cryostat.
observed in the bellows after weeks of use with the cryostat.
Time-resolved, resonant, third-order, nonlinear spectroscopies have been employed as valuable techniques in investigating many dynamical processes including solvation
dynamics,10 resonant energy transfer dynamics,8 and protein
fluctuations.9 Previously, it was shown that three pulse
stimulated photon echo peak shift 共3PEPS兲 technique is a
powerful tool to characterize solvation dynamics with sub100 fs resolution across a large dynamic range.10 The details
behind the 3PEPS technique are described elsewhere,10 but
the experimental essentials are presented here for illustration
of the utility of the cryostat insert. In short, three laser
beams, with wave vectors: k1 , k2 , and k3 , are arranged in
an equilateral triangle geometry and are focused into the
sample. The generated echo signals are simultaneously measured in two different phase-matched directions, k⫽k3
⫾(k1 ⫺k2 ), and are collected by varying the time between
the first two pulses 共coherence time兲 and the last two pulses
共population time兲. The resulting two-dimensional data are
then mapped onto a one-dimensional curve by plotting the
peak of the echo signals at fixed population times versus
coherence time. At short population times, the echo profiles
共as a function of coherent time兲 peak at nonzero values, but
as the population time is increased, the echo peak steadily
moves toward smaller values, until eventually the signals
peak at zero. The 3PEPS profiles are constructed by plotting
the peak of the echo profiles versus the population time;
hence reliable determination of the echo profile is necessary
for measurement of the peak shift value.
The OXFORD® cryostat with the constructed motion device was used to collect preliminary variable temperature
echo signals of the Rhodamine 640 laser dye dissolved in
Downloaded 03 Apr 2002 to 130.37.36.180. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/rsio/rsicr.jsp
1328
Rev. Sci. Instrum., Vol. 73, No. 3, March 2002
glycerol 共Fig. 3兲. Figure 3共a兲 clearly illustrates the advantage
of moving the Rhodamine dye/glycerol sample during the
data collection. The echo profile, at zero-population time,
collected while the sample was in motion demonstrates a
clear, well resolved peak at a nonzero coherence time, whilst
the echo profile for the stationary sample exhibits multiple
peaks with a poorly resolved maximum. The initial peak shift
values for the moving and stationary echo profiles were 13 fs
and 23 fs, respectively, and were determined by fitting the
integrated echo signals to Gaussian functions. The error bars
for the moving and stationary samples were 1 fs and 26 fs,
respectively, at a 95% confidence 共3␴兲.
The constructed 3PEPS profiles for the Rhodamine dye/
glycerol sample measured at both liquid phase 共293 K兲 and
below the glass transition temperatures 共170 K兲 are shown in
Fig. 3共b兲. These data corroborate predicted temperature dependent 3PEPS profiles and previous temperature dependent
3PEPS data.11 The uncertainties in the peak shift values 共not
shown兲 for the data collected with the motion device are
estimated at around ⫾500– 800 attoseconds at a 95% confidence, whilst the uncertainties for the signals collected without the motion device were ⬃10–20 fs 共a signal to noise
ratio of ⬃0.5兲. It is clear that measuring photon echo traces
and 3PEPS profiles for the Rhodamine 640/glycerol system
requires the use of sample motion that is accomplished with
the motion device. The data presented in Fig. 3 were collected with a 100 ␮m pathlength sample cell. The echo signals can only be produced when all three laser beams are
spatially overlapped 共focused to 50 ␮m for the data shown
Delmar S. Larsen
in Fig. 3兲 and since no noticeable decrease in the signal
intensity was observed during the sample translation, the precision of the motion was smaller than the cell pathlength,
which is accurate enough for most time-resolved measurements.
Sincere gratitude goes to Professor Graham R. Fleming
共UC Berkeley兲 for providing the financial support for constructing the initial version of this translating cell cryostat
insert. Additional thanks goes to Dr. Jeffery Musiak 共Boeing
Company兲 for useful and informative discussions concerning
the initial design.
K. K. Rebane, J. Lumin. 86, 167 共2000兲.
S. Mukamel, Principles of Nonlinear Optical Spectroscopy 共Oxford University Press, New York, 1995兲.
3
H. J. Eichler, P. Günter, and D. W. Pohl, Laser-induced Dynamic Gratings
共Springer, New York, 1986兲.
4
J. P. Simons, Photochemistry and Spectroscopy 共Wiley, New York, 1971兲.
5
W. Hage, M. Kim, H. Frei, and R. A. Mathies, J. Phys. Chem. 100, 16026
共1996兲.
6
K. J. Hellingwerf, Photochem. Photobiol. 54, 94 共2000兲.
7
D. M. Jonas, M. J. Lang, Y. Nagasawa, T. Joo, and G. R. Fleming, J. Phys.
Chem. 100, 12660 共1996兲.
8
R. Agarwal, B. P. Krueger, G. D. Scholes, M. Yang, J. Yom, L. Mets, and
G. R. Fleming, J. Phys. Chem. 104, 2908 共2000兲.
9
X. J. Jordanides, M. J. Lang, X. Y. Song, and G. R. Fleming, J. Phys.
Chem. B 103, 7995 共1999兲.
10
T. Joo, Y. Jia, J.-Y. Yu, M. J. Lang, and G. R. Fleming, J. Chem. Phys.
104, 6089 共1996兲.
11
Y. Nagasawa, S. A. Passino, T. Joo, and G. R. Fleming, J. Chem. Phys.
106, 4840 共1997兲.
1
2
Downloaded 03 Apr 2002 to 130.37.36.180. Redistribution subject to AIP license or copyright, see http://ojps.aip.org/rsio/rsicr.jsp