Note: Descriptions are shown in the official language in which they were submitted.
CA 02244218 1998-07-24
PCTIUS9710I383
WO 97128435
- i -
IMPROVED OPTICAL ACTIVITY DETECTOR
FOR USE WITH OPTICALLY ACTIVE COMPOUNDS
FIELD OF THE INVENTIOIZ
This invention relates to an optical rotation detection device
and more particularly to an optical detection system for use in
measuring optical activity within a sample as it flows through a
detector consisting of a laser diode with a beam shaping means, a
polarizing prism, a flow cell, and a means for attenuating laser
fluctuation effects from the sensed signal.
BACKGROUISD OF fiHE =NVENTIOId
Chemical species called chiral compounds appear in two subtly
different molecular forms. These molecular forms are referred to as
dextrorotatory and levorotatory, because they cause the plar_e of
polarization of light passing through them to rotate clockwise (or to
the right) for dextrorotatory and counterclockwise (or to the left)
for levorotatory. The amount of rotation is proportional to the
specific rotation of the particular chemical and' the optical path
f
f
length of interaction between the polarized light and the chemical.
The geometric shape and relative positions of the constituent atoms
for the two types of chiral molecules are generally thought to be
mirror images of each other. As such, they have the same chemical
composition and are difficult to differentiate. Unfortunately, these
two molecular forms can effect human physiology in rather different
ways. In extreme cases, patients have been harmed or have died as a
result of ingesting a drug that contained too much of the wrong type
of chiral molecule.
The enantiomeric ratio is the ratio of dextrorotatory molecules
to levorotatory molecules in a particular chiral compound. In less
CA 02244218 2004-07-26
_ 2 _
severe cases, patients have failed to respond favourably to drug
treatment because the drug they ingested had an undesirable
enantiomeric ratio. Simply stated, depending on the particular
chiral drug and enantiomeric ratio of that specific batch of
chiral drug and to some extent the particular patient, ingest~_ng
a chiral drug may not have the desired positive effect and may
have a very negative effect. Such results are especially alarming
because, at present, there is no cost effective way to measure a.nd
control the enantiomeric ratio of a particular chiral drug.
U.S. Patent No. 4,498,774 discloses a liquid chromatography
detector using a gas laser as a radiation source. This high power
radiation source is employed to give a very high ultimate signal-
to-noise ratio (S/N). To use this power to its full advantage,
very high quality prism pairs are needed to give an extinction
ratio of around 10+l0. These prism pairs are hand selected by
trial and error techniques and as such, turn out to be very
expensive when the cost of labour is factored in. Typically .it
might take 5 - 25 hours to hand select and find the highe:~t
quality area to use on any given prism. Additionally, an
intensity stabilization unit is used on the laser output to reduce
the flicker noise contribution. To retain the high extinction
ratio, air-based modulators are used. Such air-based modulators
use a great deal of current and hence require added cooling
sources to the power source and the modulator units. This patent
was issued on February 12, 1985 and has since been abandoned.
This patent may be referred to for further details.
CA 02244218 1998-07-24 pCT~S97/0i383
fVO 97128435
- 3 -
U.S. Patent No. 5,012,101 discloses an optical-rotation detector
which employs various principles of U. S. Patent No. 4,498,774, but
instead utilizes a laser diode f.or improved stability and S/N ratio.
A modulation means is used to produce a carrier wave by modulating the
rotation of the polarization from a radiation source. Optically
active samples in the beam path produce a detectable amplitude
modulation in the carrier signal. However, since the overall
amplitude modulation can be quite small, improvements in the
sensitivity of detection schemes are needed.
Additionally, systems need to be more robust for easier set up,
construction, and generalized high volume end use. Such, enhancements
would come through the improved construction of~such_elements as the
flow. cell to reduce factors causing depolarization or misalignment.
Other improvements might include the machining of the system parts
into shapes which can be easily precision interfaced so that they can
be interchanged with minimal mounting fixture adjustments. Moreover,
the detection scheme could include multiple sensors~and compensation
circuitry to eliminate the effect of power fluctuations in the laser
source.
SUMMARY OF THE INVENTION
The instant invention is an improvement upon the prior art in
that an optical rotation detector is provided that is more compact,
more robust, and generally easier to set up and operate than previous
systems. Unlike prior systems, the present system uses a diode laser
whereby the beam is first passed through a length of fiber optic cable
to first insure a near-gaussian beam. Alternatively, the laser might
have a cylindrical focusing lens attached directly to the emitting
CA 02244218 1998-07-24
WO 97/28435
PCT/US97/OI383
'' r
- 4 -
.aperture of the laser diode; underneath the hermetically sealed
window:
Next, the system employs a plan-Thompson prism to improve the
linear polarization ratio of the laser output, but unlike previous
systems the' prisms need not be hand selected as described above. Also
unlike prior systems, a Faraday rotator is used which consists of a
rod of Terbium Gallium Garnet (TGG) placed inside a solenoid. The
polarization of the beam is then varied according to a reference
oscillator frea_uency (f). This TGG modulator uses significantly less
current than other Faraday rotators (e.g., air core) as suggested by
the prior art.
Additional improvements include a specially constructed flow
cell, with improved construction and sealing techniques, for-allowing
the beam to pass through a sample substance with minimal reflection
or dispersion. The sample substance alters the angle of the plane of
polarization of the beam if the sample is optically active and this
results in a periodic amplitude variation in the det,~cted signal. To
further enhance detection of this amplitude variation, the present
invention attenuates (or subtracts and filters out), via low cost
circuitry, laser amplitude fluctuations and the 2f signals relative
to the desired amplitude modulated f frequency signal to be detected.
Ti'iis allows for greater resolution in detection of chiral compounds
with the present invention which is immune to laser fluctuations.
To date, this invention has been used as. a detector in a High '
a Performance Liquid Chromatography (HPLC) system. In a typical HPLC
system a liquid sample is injected into a high pressure solvent stream
and is subsequently forced through a tightly packed column, A column
'~ W097/28435 CA 02244218 1998-07-24 pCT~S97/01383
1 n
consists of a cylindrical pressure vessel usually packed with very
small spheres of silica, where the silica are often coated. Because
of variances in chemical attraction between the different types of
molecules in the sample and the column packing material, and in the
presence of the proper solvents, the different chemical molecules of
the sample elute from the column separated in time.
HPLC is a separation technique that can present separated
chemical species from the sample to downstream detectors one at a
time. Thus, one can measure various physical parameters (absorbance
of light at various wavelengths, refractive ir_dex, conductivity,
optical rotation, etc.) of each constituent in a sample as that
constituent flows through a flow cell in the detector.- Typically,
each detector has its own flow cell and measures only .one type of
physical parameter_ It is common also to have two or more
detectors/flow cells 1n series in order to better quantify the sample
being measured.
A typical system using HPLC and the present. invention would
f
separate the sample and, for each elutant, measure the optical
absorbance with an optical absorbance detector to determine the total
amount of each sample constituent eluted. Then, an optical rotation
detector as described by the present invention would be used to
determine the net optical rotation. Knowing the specific rotation of
the elutants, one can calculate the enanticmeric ratio. This general
procedure is typical of analytical systems used in research and -
quality control. -
- Alternatively, if a special chiral separation column is used to
- separate the enantiomers, the present invention might be used to
'WO 97/28435
CA 02244218 1998-07-24 PCT/US9710I383
v
,i.. . ~ . _ 6 _
identify the direction of optical rotation and thus to identify the
u.
enantiomer. This approach is typical of preparative applications used
to collect and/or, concentrate a desire enantiomer. Additionally,
' process control systems might use the present invention to aid in
optimizing conditions for production of the preferred enantiomer.
Conversely, the present detector might be used to minimize production
of unwanted enantiomers.
In general practice, the analytical applications are the most
i difficult and stressing on instrument design because they require a
small flow cell volume, and yet must have enough sensitivity to detect
very dilute. samples. Preparative and process control applications
typically use much larger flow cell volumes and usually have the
advantage .of much higher concentrations of sample. The present
ir_vention provides a detector capable of being readily used in any one
of these suggested applications.
Hence, the present invention i s an optical instrument that passes
a narrow beam of laser light through the long narrow bore of a flow
r
cell. Analytical applications (for instance) might' typically use a
flow cell having an inside diameter of 0.030 inches and length of 2
inches. After passing through the glan-Thompson prism, the beam has
a high degree of linear polarization .(typically 100,000:1). The flow
cell is configured so as to continuously-flow liquid thrcugh the same
long narrow bore, thus allowing the liquid and beam to interact. The
rotation of the plane of linear polarization of tine beam is measured _ _
as liquid flows through the bore of the flow cell. The rotation due
to the flow of sample through the flow cell is typically much less
than +/- one degree. Given the disclosed techniques of the present
CA 02244218 2004-07-26
invention, a precise measurement of a sample can be made
approximately 30 - 45 minutes after the detector has been
delivered and turned on. With previous systems, the typical. time
needed from turn on to sampling is upwards of 25 hours in order to
make all the necessary alignments and adjustments.
Accordingly, the present invention seeks to provide an
improved optical rotation detector, particularly for use in
detecting, separating and measuring enantiomeric ratios of chi:ral
compounds and/or related pharmaceutical drugs in high performance
systems.
Further, the present invention seeks to provide an improtTed
optical detector which utilizes a length of optical fiber or a
lens for conditioning a laser diode beam to be near-gaussian in
distribution.
Still further, the present invention seeks to provide an
improved optical detector with an improved flow cell for passing
the laser beam through chiral substances with minimal mechanical
birefringence.
Further still, the present invention seeks to provide an
improved optical detector with compensation circuitry which
detects the reflected photodetector signal and subtracts it from
the primary photodetector signal to leave a resulting amplitude
modulation signal with greater detectable resolution and
resistance to laser fluctuations.
Yet further, the present invention seeks to provide an
improved optical detector with flow cell and mounting block parts
that have been machined to be more easily interchangeable with r:~o
CA 02244218 2005-03-02
significant loss of alignment integrity.
Moreover, the present invention seeks to provide an improved
optical detector which can be setup and used in significantly
less time than prior systems.
The invention in one broad aspect provides an improved
optical rotation detector comprised of a laser diode light source
system, a first polarizing means to improve the linear
polarization ratio of the laser beam, a Faraday rotator, and a
flow cell. A second polarizing means is oriented 90 degrees from
the first polarizing means in its ability to transmit polarized
light, wherein the second polarizing means attenuates
transmission of most of the polarized laser beam. A
photodetector is provided for detecting a signal consisting of
amplitude modulated light which is transmitted through the second
polarizing means, and a phase sensitive detector, or lock-in
amplifier, as driven by a frequency oscillator, senses the f
frequency signal. The improvement comprises a means attached to
the laser diode for producing a near-gaussian laser beam, a
cylindrically shaped flow cell means constructed to minimize
mechanical birefringence and dispersion of transmitted light, a
machined ~~V~~-block mount for precision alignment of the
cylindrically shaped flow cell, and a second photodetector
positioned to detect a signal consisting of laser light reflected
from an entrance of the second polarizing means. Compensator
circuitry means produces a compensated signal with attenuation
of the interference and noise contributions from frequencies
other than frequency f.
Another aspect of the invention provides an improved optical
rotation detector with components fixedly mounted and aligned on
an operational bench, the components comprised of a laser diode
based light source system with a means selected from the group
consisting of a fiber optic cable, a cylindrical focusing lens
or multiple lenses and prisms for producing a near-gaussian laser
beam. A first glan-Thompson polarizing prism is provided to
improve the linear polarization ratio of the laser beam to at
least 1000:1. A Faraday rotator utilizing a Terbium Gallium
Garnet rod is provided for affecting the polarization of the
CA 02244218 2005-03-02
- 8A -
beam, the rotator being driven by an alternating current source,
or reference oscillator, at frequency f. A replaceable,
cylindrically shaped flow cell means is mounted on a precisioned
mounting fixture, the flow cell means being used for directing
a solvent and sample flow through the linearly polarized beam,
the flow cell means utilizing a fused quartz tube through its
longitudinal center, the tube having a capillary as the main bore
with the capillary being sized sufficiently larger than the beam
so as to minimize scattering and subsequent depolarization of the
beam. A second glan-Thompson polarizing prism is oriented 90
degrees from the first polarizing prism in its ability to
transmit polarized light, wherein the second polarizing prism
attenuates transmission of most of the polarized laser beam. A
detector means detects the transmitted light, and compensator
means processes the detected signals.
Other aspects and advantages of this invention will become
apparent from the following description taken in conjunction with
the accompanying drawings wherein are set forth, by way of
illustration and example, certain embodiments of this invention.
The drawings constitute a part of this specification and include
exemplary embodiments of the present invention and illustrate
various objects and features thereof.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is a side view of an optical rotation detector
setup showing the bench and mounts with the laser, first prism,
Faraday rotator, flow cell, second prism and detector.
Figure 2 is an exploded side view of a flow cell assembly.
Figure 2A is an end view of the flow cell body.
Figure 2B is a side view of the flow cell body.
Figure 2C is an end view of the sealing washer from the flow
cell.
Figure 2D is an end view of the donut shaped end washer from
the flow cell.
Figure 2E is an end view of the endcap of the flow cell.
Figure 2F is a side view of the endcap of the flow cell.
Figure 2G is yet another embodiment of the flow cell with
an alternative arrangement for mounting the windows.
CA 02244218 1998-07-24
fVO 97/28435
PCTlIJS9710I383
_ g _
Figure 2H is an end view of the endcap used for the flow cell of
Figure 2G.
Figure 2I is a side view of the endcap of Figure 2H.
Figure 3A is a side view of a Faraday rotator in its mounting
fixture.
Figure 3B is a side view of a Faraday rotator showing the wound
solenoid core.
Figure 4 (4A-4F) is an engineering drawing showing the front
(aA) , top (4B) , and side (4C) views of a V-block mounting fixture, and
the front (aD) , top (4E) , and sides (4F) of the top mour~ting bar for
holding the flow cell on the mounting fixture.
._ . Figure 5 is an electrical block diagram showing the processing
of the detected signal to subtract out the amplitude fluctuations of
the laser, and the notch filtering of the 2f frequency signal to
enhance resolution and stability against non-sample introduced
fluctuations in the measured frequency signal.
Figure 5A shows a plot of the amplitude modulat"e°ri 2f signal, the
non-amplitude modulated 2f signal, and the resulting signal resulting
from the amplitude modulation at frequency f.
Figure 6 shows a plot of the lock-in amplifier output versus the
laser atr~plitude, but with the compensator circuit activated which
subtracts out the amplitude fluctuations of the laser so that changes
in the laser amplitude do not significantly affect the lock-in
amplifier results.
Figure 7 shows a strip chart recording of the detected results
from a sample containing a mixture of optically oriented molecules and
a:° '.~: ..... ~ -CA 02244218 1998-07-24- -. '. ._ _. . .
.. . ' WO 97/Z8435
PCT/US97/01383
- 10 -
the.detected results which allow calculation of the true enantiomeric
ratio.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
Although the invention is described in terms of a preferred
specific embodiment, it will be readily apparent to those skilled in
this art that various modifications, rearrangements and substitutions
can be made without departing from the spirit of the invention. The
scope of the invention is defined by the claims appended hereto.
Referring now to Figure 1, an optical bench 1o is shown which
rigidly supports the various components of the detector assembly and
is typically made from solid aluminum. Once the various mounts for.
the components are in place they are generally not to be--readjusted
by the user {e, g, laser, prisms, detector) . Prior systems al lowed for
too much adjustability and systems were often knocked cut of
alignment. The flow cell is unique in that it is often changed and/or
cleaned and interchangability is a necessary design feature without
the user having to realign the entire bench alter the cell is
remounted. The present system off ers significant advantages in that
a complete system can be constructed, shimmed, and aligned in much
less time than previously disclosed systems. The overall length Idl
of the mounted components is 18.91 inches. Length M2 of 7.87a inches
(200mm) shows the focal length of the laser assembly. Length h3 of
17.000 inches shows the length between mounting holes 8 and 9.
The detector assembly includes a laser unit 12 which contains a --
visible laser diode which typically operates at a wavelength of 690
nanometers with a power output of to milliwatts. The laser diode is -
coupled to a length of single mode optical fiber (not shown),
.:1- ;':".:- ;- s ~:
CA 02244218 1998-07-24
'BYO 97128435
PCTIUS97/01383
- 11 -
typically 2 maters in length. The output of the optical fiber is
coupled to a focusing lens 14 which is used for bringing the beam to
focus inside the flow cell 20. This arrangement is used because the
output of a laser diode. is generally astigmatic and asymmetric in both
divergence angle and shape . Af ter traveling the length of the optical
fiber, however, the beam is substantially uniformly symmetric and non-
astigmatic and can be focused into a well defined near-gaussian beam.
An alternative method of achieving a satisfactory beam shape is
to place a cylindrical lens under the hermetically sealed window and
directly on the emitting aperture of the laser diode. Another method
involves the use of multiple lenses and prisms to achieve a near-
gaussian beam. . Beam shape is important because ~ any portion of the
beam that is reflected from the walls of tine flow cell, for example,
becomes depolarized and adds to the system noise. Thus for maximum
sensitivity, it is important that almost all of the laser beam pass
through the flow cell without reflection.
A plan-Thcmpson prism 16 is used to im~ove the linear
polarization ratio of the laser output from 100:1 to 100,000:1. A
FaY-aday rotator 18 is used to rotate the plane of polarization about
the longitudinal axis of the rotator, which is coaxial with tine laser
beam. Figure 3A shows the Faraday rotator 18 on its mount 19 which
fixes the rotator to the optical bench 10. The Faraday rotator
consists of a rod of Terbium Gallium Garnet (TGG), typically. 5
millimeters in diameter and 30 millimeters long, placed inside a
solenoid. Referring to Figure 3B, the solenoid 21 typically is
comprised of 4000 turns of #30 magnet wire uniformly wound on a non-
magnetic spool (e. g. Delrin) of length 40 millimeters and inside wire
~'fV097128435 CA 02244218 1998-07-24 pCT/US97IOi383
1.
- 12 -
rdiameter of 7 millimeters. The solenoid of the Faraday rotator is
'driven by a sine wave from a reference oscillator (see Figure 5) at
a frequency f (typically S00 Hz).
Referring now to Figure 2, the flow cell 20 (in Figure 1) is .
shown in exploded fashion. As stated, the flow cell is used to direct
the solvent and sample flow through the linearly polarized beam. The
f low cell body 22 of the flow cell 20 typically consists of 316
'stainless. steel and the body is engineered to fit in a "V" block mount
(see Figure 4). Of course, other suitable materials might be. used.
In particular, the flow cell body 22 and the mount 24 are precision
machined so that different flow cells can be used on the same "V"
block without realignment. Given that cylindrical lathing is a very
precise operation, and the formation of a "V" mount can be similarly
precise, these two shapes used together create a unique interface
which allows such interchangability without realignment. This is an
extremely important feature for the flow cell mount because the flow
cell might need to be removed and cleaned periodica~.ly, or replaced
altogether. In prior disclosed systems, a difficult realignment
process occurs every time a flow cell is replaced or removed for
cleaning.
Referrir_g new to Figures 2A and 2B, various dimensions as used
in the flow cell of this embodiment are shown. The flow cell body
length L1 measures 1.969 with a tolerance of +/- 0.050 inches after
assembly and grinding. The distance R1 measures 0.550 inches from the
centerline 1 (CL1). The largest diameter D1 of the flow cell measures
1.2000 inches. A glue hole 46 (see flow cell assembly details below)
' is located 0.73 inches from end 34 with a diameter D3 of 0.05 inches.
.. W097I28d35 CA 02244218 1998-07-24 p~~S9710I383
. . ,
- 13 -
_~, A flat cutout 45 runs along the top of the flow cell body 22 from end
34 to end 36. As shown below, this flat cutout allows a top cross
member to be placed over the flow cell to hold it in place on its
mount.
The end view 2A of the flow cell shows a 0.177 inch through hole
52 which runs the length of the flow cell body 22. Additionally, two
partial alignment holes 54, 56 are shown (180 degrees apart along CL1)
with a nominal diameter of 0.125 inches, 0.3 inches deep, and centered
a radius R2 of 0.950 inches from the center of the flow cell. Four
other._partial through holes 58, 60, 62, and 64, are used for
attachment screws, are each 45 degrees from centerlines CL1 and.CL2,
threaded..#4-.4p X 0.3 inches deep ar_d centered along radius R2 of 0.950
inches from center.
Referring again to Figures 2, 2A, and 2B, a length of fused
quartz capillary tube 27 (4mm outside diameter OD x 0.75mm inside
diameter ID) is shown cemented in the center of the flow cell body
with epoxy. This thick capillary tube in the na~n bore reduces
scattering and subsequent depolarization because those portions of the
laser beam that do not travel through the bore (e.g. the portion of
the beam that hits the solid portions of the fused quartz capillary
tube) are less likely to scatter into the detector. Less scattering
of the beam leads directly to lower system noise.
Additionally, an inlet tube 26, as well as an outlet tube 26',
is wrapped around a groove 44 in the flow cell body 22. In this
embodiment, the groove measures 0.40 inches wide and 0.07 inches deep.
The depth and width of this groove might vary with the amount of
tubing desired for the particular application. Inlet tube 26
X~,-~,i:_:', :~T6.C ~.: ~ - -
- - CA 02244218 1998-07-24
VVO 97128435
PCTlUS97/Ox3&3
14 _
terminates in an inlet port 28 and outlet tube 26' terminates in
outlet port 30.
Additionally, ports are drilled at each end 34 and 36 of the flow
cell body 22. The port holes 48' and 50' start just outside the fused
quartz capillary tube 27 and continue~into and radially out of the
body at 45 degrees to the longitudinal axis exiting the flow cell at
countersink holes 48 (input) and 50 (output). Putting the inlet port
28 (entering the cell at point 48) on the bottom and the outlet port
30 texiting the cell at countersink hole 50) on the top makes it
easier to flush the cell of gas (e. g. air) bubbles. Additionally,
each 45 degree port hole 48', 50' is counterbored from the outside
diameter of- the flow cell body 22 to accept-tubes 26,_ 26'. The
preferred embodiment for analytical applications uses 1/16 inch
outside diameter tubing with a 0.010 inch inside diameter for the
inlet tube 26 and 0.030 inch inside diameter for the outlet tube 26'.
The 45 degree port holes 48' and 50' are drilled through with a
diameter of 0.030 incites and countersunk a lengtiz~~"4 of 0.25 incises
deep with a diameter D2 of 0.062 inches.
The tubing 26, 26' is wrapped around the midsection 32 of t2-.e
body and brazed at the counterbores of the 45 degree port holes 48',
50' and the midsection 32. The wrapping and brazing around the
midsection greatly reduce temperature gradients within the flow cell
20. Each flow cell end 34 and 36 is covered with a sealing washer 38
anti 48. These washers might be made from a variety of materials
including chlorotrifluoroethylene (Kel-F), fluorinated ethylene
propylene (FEP), tetrafluororethylene (TFE), or other suitable
materials. Referring now to Figure 2C, each sealing washer has a slot
- -. -... --..-.
- . CA 02244218 1998-07-24
PCT/US97/01383
WO 9'1118435
- 1~ -
_42.which.allows liquid to flow between the central bore of the fused
quartz capillary tube and the inlet or outlet port. In this
embodiment, the slot measures 0.190 inches in height, 0.060 inches in
width, and has a radius on the ends of 0.030 inches. Additionally,
each sealing washer 38, 48 is 0.030 inches thick and has a set of
alignment holes 55 (two) and screw holes 59 (four) corresponding to
the holes 58-64 in the flow cell body 22.
Referring again to Figure 2, each sealing washer is covered by
an optical window 66, 68. In this embodiment, this cylindrical window
is 0.492 inches in diameter and has a length L5 of 0.236 inches. It
is made from synthetic fused silica with 5 arc parallelism and AR V-
Coat at 680 nm (reflectivity < 0.5% from 670 to 690 nm .. one side
only). Such materials can be obtained from such laboratory supply
sources as 2delles Griot of Irvine, California.
.Each window is retained by an endcap 74 and 76 with a length L8
of 0.335 inches. As shown in Figures 2E and 2F, each endcap also has
a set of a?ignment holes 79 (two) and screw holesf 77 (four) which
correspond with holes 58-54 of the flow cell body 22 and sealing
washers 38, 48. The endcap has an outer diameter D6 of 1.200 incises
and a through hole with diameter D8 of 0.200 inches. The endcap steps
down to a section with a length L7 of 0.235 inches and a diameter D7
of 0.700 inches. Additionally, the endcap has an inner cylindrical
cavity 78 facing the flow cell with a diameter of 0.500 inches and a
depth L6 of 0.245 inches for receiving the protruding end of the _,-
optical window. Typically, this endcap is made from
polyetheretherketone (PEEK). Other suitable materials could similarly
be used.
,yVV097/28435 CA 02244218 1998-07-24 PCT/US97/01383
. 3. 6 -
__ As shown im Figures 2 and 2D, a donut shaped end washer 70, 72
is placed between each endcap 74, 76 and each optical window 66, 68.
As with the previous washers, these washers are also made from
materials such as chlorotrifluoroethylene (ICel~F), fluorinated
ethylene propylene (FEP), tetrafluororethylene (TFE), or other
suitable materials. In this embodiment, the washer is 0.020 inches
thick and has an outer diameter D4 of 0.490 inches and an inner hole
with a diameter DS of 0.200 inches.
While such end washers are shown in this embodiment, they are not
recruired. One reason for including them is to provide a smooth
surface for contacting the optical window 66, 68 and hence minimize
induced bimechanical refringence. In general,_it..is more difficult
to smooth the inside contacting surface of the endcap as~opposed to
forming the end washer from an already relatively smooth material.
Hence a washer can be used to provide a smooth contacting surface (as
shown in Figure 2), or alternatively the flow cell can be constructed
without the end washer provided the inside surface of the endcap has
r ~,
been sufficiently machined or smoothed. ,
tZeferring now to Figure 2G, an alternative arrangement of the
flow cell is shovrn. In this embodiment, the flow cell 200 utilizes,
as before, inlet tube 26 and outlet tube 26' wrapped alternately
around central cutout 32. Inlet tube 26 enters the cell at the bottom
has at its other er_d input port 28. Outlet tube 26' exits the cell
at the top and has at its other end output port 30. Again, this _.
. embodiment uses sealing washers 38 and 48 over the ends of tine cell
body 22. The optical windows 66 and 68, however, are mounted
differently onto the flow cell. A different form of endcap 202, 204
::,..:.
~~p97i28435 CA 02244218 1998-07-24 pCT/t1S97/01383
.' ~ ~ ,
- 17 _
is used wherein the endcap mounts directly onto sealing washers 38,
48 with no receiving cavity 78 for the optical window. The windows
66, 68 are instead adhered directly to the outside ends 204, 206 of
the endcaps.
Referring also to Figures 2H and 2I, a front and side view of
endcaps 202, 204 are shown. Each endcap has a diameter D6 of 1.200
inches and a length L13 of 0.200 inches. As shown, each endcap also
has a set of alignment holes 208 (two) and screw holes 210 (four)
which correspond with holes 58-64 of the flow cell body 22 and sealing
washers 38, 48. The center through hole has a diameter D9 of 0,.100
inches. On the outward facing side 216 of the endcap, a cylindrical
mote 212 has been formed which has an outside diameter of 0.40. inches
and an inside diameter of 0.20 inches, and is 0.10 inches deep. A
donut-shaped island 214 then exists around the center hole with an
outside diameter of 0.20 inches.
To mount the optical window, the window is clamped over the mote
and flush against the side 216. The mote 212 is'then filled with
~' f
silicone adhesive via a glue holes 218 which penetrate the side of the
endcap 202, 204 and enter the mote 212. In practice, glue might be
injected into one hole until the mote fills and the glue exits the
other hole. The outer diameter of the window 66, 68 is slightly
larger than the outside diameter of the mote. The window then mounts
against the donut shaped island 214, with the adhesive contacting the
wir_dow along the mote. When the adhesive dries, it shrinks and draws
the window tightly and securely against the island 214. This mounting
techniaue is advantageous in that no external stressing forces are
placed upon the optical window to cause mechanical birefringence and
W097/28435 CA 02244218 1998-07-24 PCT/US97/01383
_ 18 _
misalignment_ However, it is particularly important with this
arrangement to thoroughly rinse the flow cell after use to prevent
buildup tahich might possibly crystalize and loosen the mounted window. ~ .:
Referring now to Figure 4, an example "V" mount 24 is shown with
its top, front, bottom, and side projected views. For this
embodiment, the "V° block is precision formed from 316 stainless
steel. Both "V" sections in the center 131 are oriented at 45 degrees
on each side and should be precision ground to place the center of the
flow cell body -- at its 1.2000 outside diameter points -- at a
height H1 of 1.570 +/- 0.001 inches above the bottom of the mount.
Only the areas where the "V"'s contact the 1.2000 outside diameter of~
the flow cell body are important. For example,- the bottom of the
"V"'s do not need to be machined.
Figure 4A shows the front view of the "V" mount 24 of height H1.
The top view 4B shows the alignment and mounting holes 132 (tvao) along
CL4 and CL7, each alignment and mounting hole 132 having a diameter
of 0.280 incites. These holes are used to fasten the,."V" mount 24 to
optical bench 10 with 1/4-20 socket head cap screws. Four screw holes
134 along CL3, CL5, CL6, and CL8 are drilled 0.5 inches deep for
receiving ~8-32 screws. The length L9 of the mount is 1.90 inches
with each °V" block measuring a length L10 of 0.30 inches. The mount
has a width W1 of 2.70 inches, with the screw holes 134 centered about
a width W2 of 2.250 inches. Accordingly, each screw hole 134 is 1.125
inches from CL4. Figure 4C shows the front view of the mount with
leaigth L11 of 1.90 inches and height H2 of 0.70 inches. A top
mountir_g bar 25 is shown in Figures 4D, 4E, and 4F which fits over the
top and clamps down on the mounted flow cell. Two through holes 136,
~~V097128435 CA 02244218 1998-07-24. .PCTlUS9710I383
- 19 -
along CL3, CL5 and CL6 and having a diameter 0.177 inches, allow
screws to penetrate the top bar and screw into the screca holes 134 in
the mount 24. The top bar has dimensions H3 = 0.40 i nches, L12 = 0 .30
inches, and W4 - 0.30 inches. As shown this mounting fixture
incorporates two spaced apart "V"-blocks, with each using a separate
top mounting bar.
The embodied flow cell and mount configuration is beneficial
because it has been found that mechanically induced birefringence can
cause excessive system noise if the window is stressed. In other
words,.mechanical stresses in assembling the flow cell can cause the
beam to become depolarized and this shows up as detected noise. With
the disci osed pieces, the flow cell might be assembled as follo~,=s to
reduce or eliminate such birefringence problems:
('! ) Machine the flow cell body 22, but do not grind the end faces
until all steps are complete;
(2) Wrap inlet tube 25 and outlet tube 25' around the flow cell
body 22 in the groove 44. The tubes should be wrapped so that they
~, f
alternate inlet/outlet/inlet/outlet alor_g the length of the body.
Each tube should lay snugly against the contour and complete at least
1.5 wraps. Each tube should then be close dressed into the
appropriate countersink hole 48, 50. Again, as noted, the inlet is
on the side opposite the flat cutout 45, and the outlet is adjacent
the flat cutout;
(3) Stainless steel tubing should be brazed at both countersink -
holes and at the wraps in the groove, with appropriate measures being
taken to avoid plugging the tube ends during brazing. After brazing
~'V0 97/28435 CA 0 2 2 4 4218 19 9 8 - 0 7 - 2 4 PCT/US97/OI383
i
- 20 -
is complete, the assembler should conf firm that the tubing ends are not
plugged;
(4) Position a length of fused quartz capillary tube (0.75mm ID
x 4mm inch OD) inside the flow cell body 22 such that the tube inside
diameter is centered within the 1.2000 outside diameter of flow cell
body 22 to within +/- 0.001 inches. This can be more easily
accomplished by using a °V" block fixture to support the body and
tapered pins to support each protruding end of the fused quartz
capillary tube;
(5) Cover all exposed holes and cement the glass lined capillary
tube inside the body by forcing epoxy through the glue hole 46 until
excess epoxy drips out both ends and all internal voids-are filled;
(6) When the epoxy has fully cured, grind both end faces of the
body until they are smooth, perpendicular to the glass lined capillary
tube axis, and the 0.030 through holes break out with a gap of 0.008
+/- 0.004 inches from the epoxy ring surrounding the glass lined
capill a=-y tube;
r
{7) Press in four locating or alignment pins (0.125 incr~es in
diameter and 0.5 inches long) so that each protrudes 0.2 inches from
- the face of the endcaps. The endcap and sealing washer are aligned
and the endcap is then retained in position with four #4-40 screws.
Qther window retaining schemes have been tried, including the
alternative arrangement shown in :figure 2G. While this alternative
arrangement has certain advantages (e.g. low mechanical -
birefringence), such schemes do not prove as robust as the primary
embodiment. Additionally, prior systems fail to properly keep
- solvents from penetrating the silicone cement located between the
CA 02244218 1998-07-24
- ~ n ~'V0,97/28435
-- PCT/US97/OI383
- 21 -
window and its mounting surface. As a result, the cement fails and
the flow cell leaks. Under the present scheme, leaking problems --
as well as mechanical birefringence problems -- are minimized.
Referring again to Figure 1, a second glan-Thompson prism 80 is
used as an analyzer and is mounted on the optical bench 10 after the
flow cell 20. The mounting fixture 82 holds the prism 80 in fixed
alignment with the beam passing through the flow cell 20. The prism
8o is oriented so that its transmission axis is 90 degrees from the
transmission axis of the first glan-Thompson prism. As a result, very
little light is transmitted by this analyzer prism. The effect of the
oscillatory rotation of the Faraday rotator is to rotate the Mane of
polarization -away from the nominal null or minimum ~transmiss~ion
condition created when the axis of the two prisms are oriented 90
degrees from each other. The oscillation occurs at a reference
frequency f. Such oscillatory rotation is generally small, i.e. less
than +/- one degree. Moreover, while operation of the s ~~:netric
oscillation about the null or minimum transmission is the usual
condition of operation, it has been found that this symmetry is net
a requirement for operation.
Light passing through the ar_alyzer components falls on a silicon
photodetector 84 positioned in lire with the oscillating beam. A
mour~ting fixture 86 holds the detector 8a in a fixed position.
Accordingly, the predominant photodetector output signal is amplitude
modulated at 2f, or twice the reference oscillator frequency. This -
occurs because a rotation in either direction by the Faraday rotator
causes an increase in light passing through the second plan-Thompson
prism 80 and falling on the photodetector 84. This increase in light
CA 02244218 1998-07-24
PCT/US97lOI383
WO 97/28435
- 22 -
occurs because the rotation by the Faraday rotator causes the beam
polarization to move away from the minimum transmission relationship
between the two prism polarizers 16 and 80. When an optically active
sample flows through the flow cell, then another -- and usually much
smaller -- modulation component at the reference oscillator frequency
f is produced. The amplitude of this modulated signal can be sensed
and quantified to indicate the net rotation of the angle of the plane
of polarization cause by the sample. Such detection schemes have been
generally used in prior disclosures.
Referring now to the block diagram of Figure 5, the present
invention greatly improves upon tine robustness of this detection
scheme by introducing a compensator circuit. This 'circuit requires
the placement of a second photodetector (similar or identical to the
first) so as to catch light reflected from the entrance of the
analyzer prism 80. This reflected beam will not be amplitude
modulated by polarization changes caused by an optically active sample
flowing through the flow cell 20 and affecting the po'iari zation of the
beam, nor by the rotation caused by the Faraday rotator (reseriflling
plot 90 with amplitude range 88 in Figure 5A). The reflected beam
will, however, be modulated by any variation in laser amplitude.
The block diagram shows the compensator circuit 101 and its
related processing of the signals. The signal from the primary
photodetector 100 (resembling plot 92 in Figure 5A) and the signal
from the reflected photodetector 102 are both fed through active low __
pass filters 104 and 106. In this embodiment, these filters are
implemented using operational amplifier (op-amp) components which
first changes the input current to a corresponding voltage and then
P CTIUS97/01383
~V097I28435 CA 02244218 1998-07-24
( 1
1
- 23 -
actively filters the signal, typically with a roll-off frequency of
750 Hz. The low pass filters remove unwanted noise prevalent at
higher frequencies and provide some attenuation of the 2f signal.
Compensator~circuit 101 also uses op-amps wired as an inverter
108 and as a summing inverter 110 to invert the primary signal and sum
the result with the reflected signal. Variable resistor 109 can be
adjusted to balance the primary signal and reflected signal so that
they evenly cancel out. This can be accomplished during factory
adjustment by introducing a modulation into the laser output and
adjusting the resistor 109 until a flat signal is detected. This '.
"subtraction" of signals serves to attenuate any amplitude
fluctuations that might occur due to laser beam pov,=er fluctuations.
Since both the primary signal and the reflected signal will
correspondingly fluctuate, a subtraction of the two yields a robust
signal. This robust signal will show only the amplitude modulation
as attributable to an optically active sample flowing through the flcw
cell and affecting the polarization of the beam.
f ~,
As an additional measure, a multi-stage active notch filter 112
is used to further attenuate the 2f signal relative to the f signal.
This notch filter is similarly comprised of lcw cost op-amp
components. Hence, as a whole, the low cost op-amp compensator
circuit 101 (components indicated by dotted the line) is designed to
attenuate the influence of laser power fluctuations and also to
attenuate the 2f frequency.
The resulting f frequency signal (resembling plot 94 in Figure
5A, with an amplitude range of 96) is then processed by a lock-in
_ amplifier 114 which is driven by reference oscillator 116. This
fV097/28435 CA 02244218 1998-07-24 PCT/IlS9710I383
- 2 a_ _
amplifier, sometimes called a phase sensitive detector, is used to
provide a DC output voltage that is linearly proportional to the
amplitude and phase .modulation at frequency f in this compensated
signal. -A lock-in amplifier is extremely selective and is primarily
affected by the amplitude and phase modulation at exactly the same
frequency as the supplied reference oscillator. This reference
oscillator is the same device used to drive the Faraday rotator 118
at frequency f. The output from the lock-in amplifier might be used
to drive a strip chart recorder pen or a computerized data collection
system.
' Yet another alternative embodiment would include the use of a
compensator circuit, but without the second photodetector 102. In
Figure 5, this would be represented by eliminating elements 102, 106,
108, 109, and 110. As a result, the primary photodetector signal
would still pass through the active low pass filter 104 and the active
notch filter 112 to attenuate the effect of the 2f sign-al. Fence,
even systems with relatively "quiet" lasers would sill benefit from
compensator circuitry which attenuates undesired signal components.
Referring now to Figure 6, an example of the benefits of the
compensator circuit .are shown. The lock-in amplifier output is
plotted 140 in parallel with the laser amplitude 142. As shown, at
time 144 the fluctuations in the laser amplitude increase and become
more erratic. The relative noise level 141 in signal 140, however,
is unaffected by the spikes 145 in the laser amplitude output. If
such variations showed up in the signal 140, then the detection limit
would be degraded.
CA 02244218 1998-07-24
PCTlUS9710I383
WO 97128435
- 25 -
,Accordingly, the compensator circuit is important because all
lasers have "flicker" noise, or short term amplitude variations.
Today, it is difficult to build or buy a laser with "flicker" noise
that is less than O.lo. The output of the lock-in amplifier (phase
sensitive detector) has been shown 'to be sensitive to "flicker" noise.
In order to achieve a system noise level that is below a few hundred
microdegrees of rotation, "flicker" noise must be better than 0.o5%,
or some means of compensation must be used. This compensation has
traditionally been applied to the laser beam. The present invention
discloses a technique that removes the "flicker" noise from the
signal. The advantages include, for instance, reduced cost and
complexity. - - -
In summary, as an optically active sample passes through the fl ow
cell, the output from the lock-in amplifier will deflect from the
nominal solvent baseline in a positive or negative direction,
depending on the direction and amount of rotation of the plane of
_ "
polarization due to that sample. The amount and direction of
deflection (e.g. 95 in Figure 5A) of the output from the lock-in
amplifier is linearly proportional to the rotation caused by the
sample. From this, the presence, and/or amounts, and/or net optical
rotations of chiral compounds can be detected. As stated above, such
a robust, yet sensitive, detector might be used fox analytical,
preparative, or process control applications. It is also possible to
use.an optically active solvent and .detect the reduction of rotation -
caused by a sample that is not optically active. Other analysis or
processing scenarios should suggest themselves for the current
application at hand.
x;'097/28435 CA 02244218 1998-07-24 PCTIUS97101383
- 26 -
Figure 7 shows an example of the output from the described
system. Raffinose (R) and fructose (F) were diluted to a
concentration of 0.1 micrograms/microliter and injected directly into
the solvent flow just before the flow cell (no column). With a 2
centimeter path length and the stated concentraticn of 0.1
micrograms/microliter the rotation is +526,000 microdegrees for
raffinose and -447,500 microdegrees for fructose. The appropriate
teaks are marked R and F. Then a 50:50 mixture of these two solutions
was injected and marked R+F. As shown, the resulting peak is equal
to the net rotation of 526,000-447,500 = 78,500 microdegrees. Further
the laser amplitude fluctuations, indicated on .the upper trace 160;
did not unduly affect the output of this system. _Accordingly,;the
system of the present invention has a much lower noise level and has
a more sensitive detection limit. The chart speed = 1 cm/min and the
detector uses a 20 microliter flow cell with a 20 microliter
injection. The lock-in amplifier output is charted as 162 and the
corresponding laser amplitude is plotted as 160.
In practice, the optical activity detector is often used with a
mass detector (e.g. absorbance detector). S-dhen both the mass and net
optical rotation are known, the enantiomeric ratio can be calculated.
As stated above, imprecise ratios can have disastrous effects in drug
and pharmaceutical applications.
It is to be understood that while a certain form of the i nvention
is illustrated, it is not to be limited to the specific form .or _.
arrangement of parts herein described and shown. It will be apparent
to those skilled in the art that various changes may be made without
departing from the scope of the invention and the invention is not to
,~"-a, p ~. .
CA 02244218 1998-07-24
pCTlUS97101383
? ~~p 97/28435
7 _
be considered limited to what is shown in the drawings and described
in the specification.
~TT
