U.S. patent application number 11/171854 was filed with the patent office on 2005-10-27 for variable flow rate injector.
Invention is credited to Neyer, David W., Rakestraw, David J., Rehm, Jason E..
Application Number | 20050236314 11/171854 |
Document ID | / |
Family ID | 33450037 |
Filed Date | 2005-10-27 |
United States Patent
Application |
20050236314 |
Kind Code |
A1 |
Neyer, David W. ; et
al. |
October 27, 2005 |
Variable flow rate injector
Abstract
A variable flow rate injector provides accurate, precise, and
reproducible injection volumes that have low dispersion. The
invention is particularly well-suited for HPLC injection volumes
<500 nL but can be used to inject larger volumes and in
different applications as well. Injections are performed at a first
flow rate and separations are performed at a second flow rate. For
improved HPLC system performance, the first flow rate is less than
the second flow rate. The injector uses a variable flow rate fluid
supply that allows rapid switching between flow rates desired for
injection and flow rates desired for separations.
Inventors: |
Neyer, David W.; (Castro
Valley, CA) ; Rakestraw, David J.; (Livermore,
CA) ; Rehm, Jason E.; (Alameda, CA) |
Correspondence
Address: |
SHELDON & MAK, INC
225 SOUTH LAKE AVENUE
9TH FLOOR
PASADENA
CA
91101
US
|
Family ID: |
33450037 |
Appl. No.: |
11/171854 |
Filed: |
June 29, 2005 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
11171854 |
Jun 29, 2005 |
|
|
|
10441640 |
May 20, 2003 |
|
|
|
Current U.S.
Class: |
210/198.2 ;
210/137; 422/70; 73/61.56 |
Current CPC
Class: |
Y10T 436/2575 20150115;
G01N 30/16 20130101; G01N 2030/324 20130101 |
Class at
Publication: |
210/198.2 ;
210/137; 073/061.56; 422/070 |
International
Class: |
B01D 015/08 |
Claims
1. (canceled)
2. An injector system for injecting a sample into a device
comprising: (a) a variable flow rate source of a working fluid; (b)
a sample source; (c) a valve in fluid communication with the sample
source and with the working fluid source, wherein the valve has (i)
a sample load position, and (ii) a sample inject position, wherein
in the sample inject position sample from the sample source can be
placed in a sample injection zone for injection by the working
fluid into the device; and (d) a controller in communication with
the working fluid source for controlling the flow rate of the
working fluid, wherein the working fluid flows at a first flow rate
when the working fluid is injecting placed sample into the device
and the working fluid flows at a second flow rate after the working
fluid injects the sample into the device; wherein the first flow
rate is less than the second flow rate.
3. A system of claim 2 where the device is a microcolumn of a HPLC
system.
4-12. (canceled)
13. The system of claim 2 wherein the variable flow rate source
comprises an electroosmotic flow controller.
14. The system of claim 2 wherein the variable flow rate working
source comprises an electrokinetic pump.
15. The system of claim 2 wherein the variable flow rate working
source comprises a variable pressure source.
16. The system of claim 2 comprising a plurality of variable flow
rate working fluid sources.
17. The system of claim 16 further comprising flow detection means
for detecting the flow rate of working fluid from each variable
flow rate working fluid source, the flow detector means being in
communication with the controller; wherein the controller can
compare the detected flow against a target flow rate and adjust a
respective variable flow rate working fluid source so the
respective working fluid flows at about the target flow rate.
18-26. (canceled)
27. A system for injecting a sample into a device comprising the
steps of: (a) means for placing the sample into a remote injection
zone; (b) means for moving at least a portion of the placed sample
from the sample injection zone into the device with a working fluid
at a first flow rate; (c) means for introducing the working fluid
into the device at a second flow rate, wherein the first flow rate
is at least 25% less than the second flow rate; and (d) a
controller for controlling the flow rates.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application is related to U.S. patent application Ser.
No. 10/246,284, filed Sep. 17, 2002, which is a
continuation-in-part of U.S. patent application serial number U.S.
2002/01953444 filed May 24, 2002, with a continuation-in-part of
U.S. patent application number U.S. 2002/0189947 filed Aug. 29,
2001 that claims the benefit of U.S. Provisional application No.
60/298,147 filed Jun. 13, 2001, the entire disclosures of which are
incorporated by reference in their entirety for any and all
purposes.
BACKGROUND
[0002] High performance liquid chromatography (HPLC) is a technique
that has been used for many years as a means of separating,
identifying, purifying and quantifying components of often complex
mixtures. HPLC is an important tool used by biotechnological,
biomedical, and biochemical research as well as in the
pharmaceutical, cosmetics, energy, food, and environmental
industries.
[0003] Conventional HPLC typically is performed using
chromatographic columns with inside diameters (I.D.'s) in the range
of about 2-10 mm, 4.6 mm columns being a common standard.
Microcolumn liquid chromatography ("LC"), which is the most widely
accepted term to describe liquid chromatography using packed
columns having I.D.'s of 2 mm or less, is gaining in popularity.
Advantages of microcolumn LC include the ability to analyze smaller
sample volumes, reduction of solvent usage, and enhanced mass
sensitivity.
[0004] Due to their relatively large volume, sample injection
systems developed for conventional HPLC systems are inadequate for
use in microcolumn LC systems. Sample volume requirements for
microcolumn LC relative to conventional HPLC can be determined by
considering a constant sample volume to column volume ratio. For
example, a direct scaling of injection volumes indicates that a 10
.mu.L sample injection into a 4.6 mm i.d. conventional HPLC column
would be equivalent to 43 nL, 4.7 nL and 1.2 nL sample injections
into 300 .mu.m, 100 .mu.m and 50 .mu.m i.d. nanobore HPLC columns,
respectively.
[0005] Injection valves and injection methods have been developed
in an attempt to meet the demands of volume injections of 500 nL
and less. See, for example, those disclosed in Vissers, J. P. C.,
Arnoud, H. R., Ursem, M. and Chervet, J. P., "Optimised injection
techniques for micro and capillary liquid chromatography", J. of
Chrom A, 746, p 1, (1996); Bakalyar, S. R., Phipps, C., Spruce, B.
and Olsen, K., "Choosing sample volume to achieve maximum detection
sensitivity and resolution with high-performance liquid
chromatography columns of 1.0, 2.1 and 4.6 mm I.D.", J. Chrom. A,
762, p 167, (1997); Foster, M. D., Arnold, M. A., Nichols, J. A.
and Bakalyar, S. R., "Performance of experimental sample injectors
for high-performance liquid chromatography microcolumns", J. Chrom.
A, 869, p 231, (2000). Others include commercially available valves
from companies such as VICI Valco Instruments, Rheodyne and
Upchurch Scientific. Valve designs include both external and
internal sample loops. Injection volumes of less than 100 nL are
typically achieved using valves with internal sample loops where a
groove in the rotor serves as the loop. Larger injection volumes
can be achieved with either internal loops or external loops
connected to the valve ports.
[0006] In conventional HPLC systems, the resolution and efficiency
of the separation have been primarily determined by the performance
of the column itself. In contrast to conventional HPLC systems, the
resolution and separation efficiency of microscale HPLC systems is
often determined by band-broadening from the sample injector,
connection tubing and detector cell.
[0007] The band-broadening due to the instrumental components as
well as the column is called dispersion and is characterized by the
variance of the peak shape. The ideal injection (square pulse) will
introduce a variance, .sigma..sup.2, of .about.V.sup.2/12, where V
is the injection volume. In practice, this ideal injection
performance has not been demonstrated for injection volumes of
<500 nL and the sample is contained in a volume much larger than
the ideal square pulse.
[0008] In addition to minimizing the variance, microcolumn LC
systems must have the ability to quantitatively inject small sample
volumes. Although internal loop injectors can directly provide for
injection volumes as small as 10 nL, the absolute accuracy and
consistency from loop to loop are poor.
[0009] As the injection volumes become smaller, quantitative
injections become more difficult. This difficulty is reflected in
the literature as well as product specifications of commercial
instruments. For example, the specifications for the relative
standard deviation on the Agilent 1100 HPLC system using a
conventional injection size of 5 .mu.L is 6 times better than for
the Agilent 1100 CapLC using injection volumes of 0.2-1.0 .mu.L. No
specification is given for smaller injection volumes. Typical
repeatability for peak areas in microscale LC are in the 2-6% range
compared to 0.5% for conventional size systems.
[0010] Accordingly, there is a need in the art for an HPLC injector
and a method for injecting a sample fluid into a separation column
that provides the ability to reduce injection variance and maintain
very accurate and precise injection volumes.
SUMMARY
[0011] The present invention provides an injection system
particularly suitable for HPLC systems, that satisfies this need.
According to the present invention, a method for injecting a sample
into a device such as a microcolumn comprises placing the sample
into a sample injection zone, moving at least a portion of the
placed sample from the sample injection zone into the device with a
working fluid flowing at a first flow rate, and thereafter,
introducing the working fluid into the device at a second flow
rate. The first flow rate is at least 25% less than the second flow
rate. Thus, sample is introduced at a rate lower than the rate at
which the working fluid is introduced during the working portion of
the process. This lower rate reduces sample dispersion.
[0012] Typically the volume of the sample injected is less than
about 500 nL. Preferably the time delay between injecting the
sample into the device and increasing the flow rate of the working
fluid is less than about five seconds, and more preferably less
than about one second. The first flow rate is typically from 0.01
to about 0.75 of the second flow rate. Typically, the second flow
rate for a microcolumn is less than about 100 .mu.L per minute.
[0013] When the sample is injected into the device, the working
fluid can move substantially all of the sample, or only a portion
of the sample.
[0014] Hardware for practicing this method can comprise a variable
flow rate source for the working fluid, a sample source, a valve in
fluid communication with the sample source, and a controller. The
valve has a sample load position and a sample inject position. In
the sample load position, sample is loaded for later injection. In
the sample inject position, a coded sample is injected by the
working fluid into the device. The controller is in communication
with the working fluid source for controlling the flow rate of the
working fluid.
[0015] The variable flow rate working source can be an
electrokinetic pump, and can include an electroosmotic flow
controller or a variable pressure source. More than one working
fluid can be used. A control system can be provided that maintains
the working fluid flow rate set against a target flow rate.
DRAWINGS
[0016] These and other features, aspects and advantages of the
present invention will become better understood with regard to the
following description, appended claims, and accompanying drawings
where:
[0017] FIG. 1 schematically illustrates a first HPLC system having
a variable flow rate injector in accordance with the present
invention.
[0018] FIG. 2 schematically illustrates a second HPLC system having
a variable flow rate injector in accordance with the present
invention.
[0019] FIG. 3 is a graph illustrating the relationship between the
flow rate during injection and variance for a system of the type of
FIG. 1.
[0020] FIG. 4 is a graph illustrating the variance of timed
injections at a flow rate of 540 nL/min for a system of the type of
FIG. 1.
[0021] FIG. 5 is a graph of chromatographic data obtained using a
system of the type illustrated in FIG. 2.
DESCRIPTION
[0022] The present invention is directed towards an injector and
method that can provide accurate, precise, and reproducible
injection volumes that have low dispersion. The invention is
particularly well-suited for HPLC injection volumes <500 nL but
can be used to inject larger volumes and in different applications
as well.
[0023] A system 10 having features of the present invention is
shown in FIG. 1. An injection valve 12 has a sample load position
14 and a sample inject position 16. In FIG. 1 the injection valve
12 is in the sample load position 14. The valve 12 has six ports,
ports A-F, as shown in FIG. 1, with port A at about a 10 o'clock
position, and the other ports being substantially equally spaced
apart from each other. The valve 12 has three internal flow paths,
also referred to as channels, 12A-B, 12C-D, and 12E-F between ports
A and B, C and D, and E and F, respectively.
[0024] A device such as a separation column 18 is in fluid
communication with the injection valve 12. A variable flow rate
working fluid source 20 providing a working fluid that is in fluid
communication with the injection valve 12 and the separation column
18. When the injection valve 12 is in the sample load position 14,
the working fluid flows through flow path 12 A-B into the
separation column 18. A sample source provides a sample to be
injected into the column 18. The sample can be injected by a
syringe 28 into a sample injection zone such as a loop 29 that
includes a sample retention coil 30. The loop 29 is external to the
valve 12.
[0025] When the injection valve is in the sample injection position
16, as shown in FIG. 2, the working fluid displaces a volume of
sample solution from the injection valve 12 so that the volume of
sample solution enters the separation column. As shown in FIG. 2,
working fluid flows through valve channel 12E-F, into the loop 29
and through coil 30, thereby displacing at least a portion of any
placed sample, and then through valve channel 12A-B and into the
column 18.
[0026] While a typical six-port injection valve is shown in FIGS. 1
and 2, many other configurations of injection valves which have
different numbers and positions of ports and channels are common
and can be used. Rather than using a loop external to the valve as
the sample injection zone, one or more of the internal valve
channels (similar to channel 12E-F and/or channel 12A-B) can be
used.
[0027] A controller 22 is in communication with the variable flow
rate pressure source 20 and the injection valve 12 for controlling
the flow rate of the working fluid and controlling the position of
the valve 12. When the working fluid displaces the sample solution
from the injection valve 12, the working fluid flows at a first
flow rate. After the working fluid displaces at least a portion of
the sample, the working fluid flows at a second flow rate. The
first flow rate is different from the second flow rate.
[0028] FIG. 2 illustrates an HPLC system embodying features of the
invention. The variable flow rate working fluid supply 20 comprises
two variable flow rate working fluid supplies 20a and 20b. The flow
rate of each fluid supply is measured by a respective flowmeter 24a
or 24b that is in communication with the controller 22. The
controller 22 compares the measured flow rates of each of the
working fluids with a respective desired or target flow and adjusts
the respective variable flow rate working fluid supply 20a or 20b
so that the respective working fluid flows at about the target flow
rate and so that the working fluid flows through the injection
valve 12 at about the target flow rate. The HPLC system 11 shown in
FIG. 2 also includes a detector 26 for detecting analytes in fluid
after passing through the separation column 18.
[0029] The working fluids are mixed after exiting the flowmeters
24a and 24b and before passing through the injection valve 12.
Mixing can occur via diffusion or via passive or active devices.
Ideally, the mixed fluids only need to flow through a minimal
volume, the delay volume, 100 nL for example, to the injection
valve 12 so that changes in mixture composition are accurately
represented, with little time delay, in the fluid passing through
the injection valve. If desired, one or both fluids can be used for
sample injection, and one or both can be used for running the
column 18.
[0030] Preferably, the controller 22 can compensate for changes in
the working fluids or mixtures that affect the flow rates. These
may include, for example, viscosity changes or volumetric changes
upon mixing. Hence, preferably, the controllers 22 can obtain the
physical properties of the fluid such as composition, temperature
and pressure. For example, the composition and mixing ratio of both
fluids can be input to the controller; the flowmeter can measure
the pressure; and a thermocouple in communication with the
controller can take the temperature measurement. Alternatively, the
system can be temperature controlled and the temperature
communicated to the controller.
[0031] Because the flow rate is measured and the measured flow rate
is used to adjust the variable pressure fluid supply as opposed to
adjusting the mechanical displacement of a pump element, e.g., a
lead-screw driven piston, so that it is proportional to a desired
flow rate, the system is capable of delivering fluid predictably
and reproducibly even at low flow rates and even if there is check
valve leakage, pump seal leakage, flexing and creep of mechanical
seals, thermal expansion of components and compression of the
working fluid.
[0032] Preferably the system has a response time of less than one
second so that when the measured flow rate does not substantially
equal the target flow rate, the actual flow rate is quickly
adjusted to substantially equal the desired flow rate within one
second, wherein substantially equal means within 5%.
[0033] The injection valve 12 can be any injection valve known in
the art, for example commercial rotary valves such as those
available from Rheodyne, Valco Instruments, or Upchurch Scientific,
or microfluidic injectors such as those disclosed in U.S. Pat. No.
6,290,909 (which is incorporated hereby reference) and U.S. patent
publication No. 2002/0194909 (incorporated herein by reference).
Sample loading is effected with the syringe introducing sample into
valve port E for flow through valve flow path 12E-F, and into the
sample loop 29 including into the coil 30. To be certain that
sufficient sample is provided, there is a waste receptacle at valve
port D, wherein valve channel 12C-D is in communication with the
waste receptacle. By introducing sufficient sample into the sample
injection zone that sample appears at waste insures that sufficient
sample is provided for injection into the column 18.
[0034] The variable flow rate working fluid supply 20 can be of any
type known in the arts or developed in the future including but not
limited to: direct electrokinetic pumps, such as those disclosed in
U.S. Pat. No. 5,942,093, which is incorporated herein by reference
for any and all purposes; electrokinetic flow controllers, such as
those disclosed in U.S. patent application Ser. Nos. 09/942,884 and
10/155,474 which are incorporated herein by reference for any and
all purposes; electropneumatic pumps with and without hydraulic
amplifiers, such as those described in U.S. patent application Ser.
No. 10/246,284, which is incorporated herein by reference for any
and all purposes; and mechanically actuated pumps. Although many
current designs of positive displacement pumps, such as lead-screw
driven pumps, do not have the performance to address the precision
at the low flow rate ranges, they may be used in active flow rate
feedback in future designs. When more than one variable flow rate
working fluid supplies are used, they need not be the same.
Preferably the variable flow rate working fluid supply 30 is
continuously variable, can provide flow rates in the range of 1
nL/minute to 100 .mu.l/minute into back pressures of up to 5000 psi
or higher, and have a response time of seconds or less, thus
allowing rapid changes in flow rates.
[0035] The controller 22 can be any controller known in the art,
for example, a PID servo-loop controller, and can be constructed
using discrete analog circuits, discrete digital circuits,
dedicated microprocessors or a computer, for example.
[0036] The working fluid can be the mobile phase in an HPLC system,
for example or any other fluid. The sample solution can contain
analytes or be any fluid. The term "sample" is used herein broadly
to refer to any material which it may be desired to inject into a
device. For example for a column such as a microcolumn the sample
can contain one or more compounds that are to be separated,
analyzed, and/or reacted, where the compound(s) are known or
unknown.
[0037] The flowmeter 24 may be of any type known in the art
including but not limited to: determining flow rate from measured
pressure difference across a known flow conductance; a Coriolis
flowmeter as disclosed in P. Enoksson, G. Stemme and E. Stemme, "A
silicon resonant sensor structure for Coriolis mass flow
measurements," J. MEMS vol. 6 pp. 119-125 (1997); a thermal
mass-flowmeter; a thermal heat tracer as disclosed in U.S. Pat. No.
6,386,050; and an optical flowmeter, for example, a Sagnac
interferometer as disclosed in R. T. de Carvalho and J. Blake,
"Slow-flow measurements and fluid dynamics analysis using the
Fresnel drag effect," Appl. Opt. vol. 33, pp. 6073-6077 (1994).
Preferably the flowmeter 24 provides accurate and precise
measurements of flow rates in the range of 100 .mu.L/min to 10
nL/min, typical flow rates for microcolumns. It is further
preferable that the flowmeter 24 provide a signal that is
continuous over all desired flow rates including fluid flow in both
directions. It is further preferable that the signal bandwidth of
the flowmeter 24, i.e. the frequency corresponding to the minimum
time between meaningful readings, is faster than one Hertz, and
more preferable faster than 10 Hertz.
[0038] These objectives can be accomplished by a flowmeter
comprising a metering capillary having a sufficiently long length
and a sufficiently small inner diameter so that the pressure drop
across the metering capillary is at least 5% of the input pressure
to the capillary at the desired flow rate, and a pressure sensor
for measuring the pressure drop across the metering capillary,
wherein the input pressure is the pressure of the fluid as it
enters the capillary. One or more pressure sensors can be used to
measure the pressure drop across the capillary directly or by
measuring the pressure at both ends of the capillary and
subtracting one pressure measurement from the other. The pressure
sensor can be a pressure transducer. Minimizing the volume and size
of the pressure transducers to 10 microliters, for example, allows
for rapid response of the flowmeter since the compressibility of
the fluid and the deflection of the pressure transducer membrane
contribute to the time response.
[0039] For example, a pressure drop of about 450 psi through a 10
cm long and 10 micron ID capillary indicates a flow rate of about
500 nL/min for water at room temperature. Similar relations can be
determined for other fluids, geometries, pressure differences, and
lengths of tubing using the well known Darcy's law for pressure
driven flow. Accurate flow rate measurements also require knowledge
of the fluid viscosity.
[0040] There are at least two fill methods of using an HPLC system
having features of the present invention, a complete fill method
and a partial fill method. In the complete fill method, while the
injection valve 12 is in the load position 14, an amount of sample
in excess of what is needed for injection is inserted into the
sample loop 29. The amount of sample needed to uniformly and
completely fill the sample loop 29 depends on many parameters
including, the geometry of the sample loop, connection ports,
volumetric flow rate during loading, viscosity of the sample
solution and the diffusion constants of the components in the next
sample. The first flow rate of the working fluid is established.
This is the flow rate at which the sample solution is displaced
from the sample loop. The valve is actuated into the run position
for a time great enough to displace a desired amount of the placed
sample. Timed injections (also referred to as moving, temporary, or
time slice injections) take advantage of either pneumatic or
electronic valve actuation, for example, that switches the
injection valve 12 into the run position 16 for a desired period of
time to transfer a desired volume of the sample solution out of the
sample loop 29. When the injection valve 12 is in the run position
16, the desired volume of the sample solution is displaced from the
sample loop 29 by the working fluid. This injection method requires
the first flow rate to be stable and precise enough to allow a
known and repeatable volume of sample to be injected. The desired
volume is usually only a portion of the placed sample 12. The
injection valve 12 then switches back to the load position 14. This
method often allows the dispersion in the trailing boundary of the
sample to be reduced.
[0041] The flow rate of the working fluid (second flow rate)
desired for separation is preferably established before the sample
enters the separation column. Because the connection volume between
the injector and the separation column is preferably kept to a
minimum, the second flow rate is preferably established rapidly,
preferably in less than about five seconds, and more preferably in
less than about one second. After passing through the injection
valve 12, the sample solution can enter the separation column 18 at
the second flow rate. After the injection valve 12 is switched back
to the load position 14, more sample solution can be inserted into
the sample loop to flush out the sample loop 29, forcing all of the
working fluid out of the sample loop and through the fluid outlet
32 into waste and refilling the sample loop with the sample
solution.
[0042] In the partial fill method, less than the amount of sample
needed to fill the loop 29 is used. While the injection valve 12 is
in the load position 14, a predetermined amount of sample solution
is metered into the sample loop 29. It is preferable that none of
the sample solution reaches the far end of the sample loop 29
during loading to ensure that the metered sample volume is fully
contained in the sample loop. When the injection valve 12 is
switched into the run position 16, the entire volume of the sample
loop 29 is displaced by the working fluid. After passing through
the injection valve 12, the sample solution can enter the
separation column 18. The injection valve 12 can be switched back
to the load position 14 and the process can be repeated.
[0043] These methods are applicable to any sample volume, including
when the volume of the sample solution displaced from the injection
valve 12 is less than approximately 500 nL. When either method is
used, one flow rate of the working fluid is used to inject the
sample solution and a second greater flow rate of the working fluid
is used during separation of the sample solution. When the sample
solution is being neither injected nor separated, the flow rate of
the working fluid is preferably changed from the first flow rate to
the second flow rate in a prescribed and reproducible method so
that the elution times of the separated sample components are
reproducible. This may be a rapid step change in the flow rate or a
more gradual ramp from the first flow rate to the second flow rate
flow rate. The flow rate of the working fluid can switch from the
first flow rate to the second flow rate at any time between
injection and separation and can switch from the second flow rate
to the first flow rate at any time after separation of the sample
solution and before injection of the next volume of sample
solution. Both flow rates are preferably less than approximately
100 .mu.L/min. For reduced dispersion and optimal separation the
first flow rate is preferably less than the second flow rate.
[0044] The device 18 need not be a separation column, but can be
another component, such as a microfluidic reaction chamber or
detector.
[0045] The controller 22 can cause the flow rate to switch from the
first flow rate to the second flow rate in less than approximately
five seconds from the time that the injection valve 12 switches
from the run position 16 to the load position 14. Alternatively,
the switching of the flow rates can be delayed from the valve
switching by a predetermined time or have other triggers, such as
an optical, electronic, or electrochemical sensor.
[0046] Typically the first flow rate is about 25% less than the
second flow rate for minimizing dispersion of the injected sample.
For example, the first flow rate can be from about 1 to about 75%
of the second rate, and preferably from about 10 to about 75% of
the second flow rate. As an example, for an HPLC system, the first
flow rate use for injection can be about 100 nanoliter per minute,
and the second flow rate of working fluid used for separation can
be 10 microliters per minute.
[0047] The ability to use different flow rates for injection and
separation allows both processes to be optimized independently. The
benefits of this invention can be illustrated by considering the
case of optimizing injection and separation flow rates for a 300
.mu.m i.d. separation column 18. A flow rate of approximately 3-4
.mu.L/minute is typical for optimized separation performance.
Injection volumes of less than about 50 nL (assuming ideal
injections, i.e. a square pulse) may be desired to minimize
dispersion effects for certain applications. In particular HPLC
separations with weakly retained compounds including size exclusion
chromatography and other isocratic separations may require small
injection volumes with minimal dispersion. As an example, the total
extra column variance, including the injector, detector and
connecting tubing, for a 150 mm long column expected to yield
10,000 theoretical plates should be kept below about 370 nL.sup.2
(or 10% of the variance produced by the column). The amount of
dispersion introduced by increased flow velocity depends on
detailed geometries of the loop and valve ports, solvents and
analytes.
EXAMPLE 1
[0048] The lowering of flow rates during injection of the sample
solution results in a significant improvement in reducing the
variance of the injection, as demonstrated in FIG. 3. For this
example, an injection valve with a 20 nL internal sample loop was
used. The fluid flow rate for this experiment was controlled using
a constant pressure pump (Jasco PU-1580) connected to a small
diameter capillary that restricted the flow rate prior to the
injection valve. The flow rates were determined by directly
measuring the volume of fluid displaced over a fixed period of
time. The injection valve used was a Valco C4 Cheminert valve with
an internal sample loop volume of 20 nL (nominal specification).
The variance of the injection was measured by connecting the output
of the valve to a short piece of 25 micron i.d. capillary (15 cm
long). The sample plug was detected optically by measuring the
absorbance versus time as the sample passed through a short section
of the capillary (measurement volume of .about.1 nL). The variance
of the connection capillary and detection volume provided minimal
dispersion and allowed a direct measurement of the variance
introduced by the valve. The variance was estimated by converting
the time varying absorption to volume and then fitting the
distribution to a Gaussian profile. The variance of a Gaussian peak
was determined as the square of its standard deviation, .sigma..
The true variance is larger than this estimate because the fit to a
Gaussian underestimates the contribution from tailing.
[0049] The variances determined for flow rates of 3.0, 1.3 and 0.54
.mu.L/min were 1080, 530 and 280 nL.sup.2 respectively. As shown in
FIG. 3, the tail of the injection shows considerable dispersion and
adds significantly to the variance of the injection. Further
reduction in variance can be made in the system of Example 1 by
using timed injections that remove much of this dispersive
tail.
[0050] To achieve injection volumes of .about.40 nL at 4
.mu.L/minute would requires injection times of 0.6 seconds. In
practice, even shorter times may be required to achieve desirable
injection variances. Since the actuation times for most valves are
in the range of 100 ms, making very reproducible injections where
the valve position is changed on the 500 ms time scale is
challenging. However, if the flow rate can be reduced by a factor
5-10, injection times of 2-5 seconds can be used and increased
precision is possible.
EXAMPLE 2
[0051] The test results for this experiment are presented in FIG.
4, which illustrates the low variance of timed injections at 540
nL/min flow rate where the injection plugs were measured as
described above. The measurements used the same experimental
configuration as for Example 1, except the 20 nL internal loop of
the valve was changed to a .about.250 nL internal loop. The valve
was actuated between the load and injection positions with a
standard electronic actuator available from the manufacturer
(Valco).
[0052] The injection plug shapes for 2, 4, 6, 8, 12 and 40 seconds
are shown and result in Full Width Half Maximum (FWHM) widths of
29, 41, 58, 112 and 270 nL. The dispersion of the measured
injection plugs was characterized by the FWHM because the
distributions were not well characterized by a Gaussian. The 40
second injection allowed injection of the entire loop volume and
illustrates the dispersive tail from a complete loop injection. The
reproducibility of peak heights was also measured. A relative
standard deviation of 0.5% was measured for a series of 3 second
injections at 500 nL/min followed by switching flow rates to 3
.mu.L/min.
EXAMPLE 3
[0053] The performance of the embodiment of the invention shown in
FIG. 2 has been demonstrated and the resulting test results are
presented in FIG. 3. The injection valve used for this example was
a Valco CN2 with a 250 nL external sample loop. The valve was
pneumatically actuated under computer control. The separation
column was 150 long by 0.3 mm i.d. packed with Phenomenex Luna C18
stationary phase (3 micron diameter). Detection was accomplished
using a microfabricated detection cell of .about.45 nL volume and a
path length of .about.4 mm. The relative standard deviation of peak
height of <1% has been measured for 3 second injections
conducted at 500 nL/min followed by switching to 4 .mu.L/min for
separation. A mixture of uracil, acetophenone, propiophenone and
butyrophenone were run under isocratic conditions with a buffer of
55% methanol and 45% water. Exemplary chromatographic data is
presented in FIG. 5, where the results from 9 separate separations
are overlaid.
[0054] Although the present invention has been described in
considerable detail with reference to certain preferred versions
thereof, other versions are possible. For example, three or more
variable flow rate fluid supplies can be used in a single HPLC
system. In addition the working fluid used for sample injection
need not be the same as the working fluid introduced into the
device 18 after the sample is injected. Therefore, the spirit and
scope of the appended claims should not be limited to the
description of the preferred versions contained herein.
[0055] All features disclosed in the specification, including the
claims, abstracts, and drawings, and all the steps in any method or
process disclosed, may be combined in any combination, except
combinations where at least some of such features and/or steps are
mutually exclusive. Each feature disclosed in the specification,
including the claims, abstract, and drawings, can be replaced by
alternative features serving the same, equivalent or similar
purpose, unless expressly stated otherwise. Thus, unless expressly
stated otherwise, each feature disclosed is one example only of a
generic series of equivalent or similar features.
[0056] Any element in a claim that does not explicitly state
"means" for performing a specified function or "step" for
performing a specified function should not be interpreted as a
"means" for "step" clause as specified in 35 U.S.C. .sctn. 112.
* * * * *