Joost Lo¨tters
R&D Manager, Bronkhorst High-Tech B.V., Nijverheidsstraat
1A, 7261 AK Ruurlo, The Netherlands
Abstract
Purpose – To present a new
generation of liquid flow sensors that is capable of
meeting the requirements as imposed by the life science,
analysis, biotech and other markets.
Design/methodology/approach – A
description of the design and development of low flow
rate measuring system and typical applications.
Findings – The system described
uses tubes made of silica, stainless steel or PEEK,
and either constant power or constant temperature methods
in conjunction with a heater and temperature sensor.
The tested instruments were capable of measuring flow
ranges between 25-500 nl/min (smallest flow range) and
100-2000 µl/min
(largest flow range) water, with operating pressures
up to 100 bar (up to 400 bar for flow meters with flow
ranges below 100 µl/min).
Originality/value – Presents
information on a new generation of liquid flow sensors.
Keywords Liquid flow, Sensors, Measurement
Paper type General review
Accurate measurement and control of tiny liquid flows
of the order of nanolitres through millilitres per minute
is becoming more and more important for a lot of applications
in the life science, analysis (e.g. HPLC), biotech,
synthesis (of e.g. pharmaceuticals) and nanotechnology
markets.
Accompanying demands to flow sensors suited for this
low flow range are an extremely small internal volume
and the use of for instance PEEK or fused silica as
wetted material for the flow sensor tube as alternative
to stainless steel. Furthermore the instruments should
have a modular set-up, so they can be easily exchanged
and adapted to a new need.
For example, the separation column at the detector
side of analytical equipment is sometimes made of the
material fused silica with an internal diameter typically
of the order of 100 µm.
In some cases, it is necessary to measure the –
very small – flow at the detector side to improve
the accuracy of the analysis. In this application, the
internal diameter and wetted material of the flow sensor
tube should preferably be the same as those of the separation
column, to avoid disturbances in the flow and to minimise
the internal volume.
Until recently, none of today’s commercially
available flow sensors were equipped with the above-mentioned
features. In this paper, a new generation of liquid
flow sensors is presented that is capable of meeting
the requirements as imposed by the life science, analysis,
biotech and other markets.
Sensor structure and basic operating principle
The actual flow sensor consists of a straight flow
tube with two active elements around it, as shown in
Figure 1(a) and (b). The wetted material of the flow
tube is stainless steel, or as an option, fused silica
or PEEK. The internal diameter of the flow tube may
vary between 20 and 200 µm,
depending on flow range. The corresponding internal
volume of the mass flow meter is 1.5-20 µl.
Two measurement principles can be distinguished, namely,
the constant power (CP) and the constant temperature
(CT) method. The CP measurement principle is used for
the flow ranges below circa 100 µl/min.
In this case, the two elements are used as both heater
and as temperature sensor, as shown in Figure 1(a).
Both elements are provided with an equal amount of constant
power, the temperature difference (∆T)
between them is a measure for the flow.
The CT measurement principle is used for the flow
ranges above circa 100 µl/min.
In this case, the first element acts as temperature
sensor, and the second element acts as a heater, as
shown in Figure 1(b). The heater is heated to a certain
constant temperature difference (∆T)
over the medium temperature. The heater power Pheater
necessary to keep ∆T
constant is a measure for the flow.
The above-mentioned flow sensor structures and operating
principles have the following innovative features and
advantages. . The flow sensor comprises a short straight
flow tube, with an internal diameter varying between
20 and 200 µm,
thus having an extremely small internal volume, varying
between 15 nl and 1.5 µl.
- The smallest measurable flow range is 25-500 nl/min,
the largest measurable flow range is 100-2,000 µl/min;
the response time t98% is of the order
of 1 s.
- The measurable flow range can easily be adjusted,
by varying the internal diameter, material and wall
thickness of the flow tube and the measurement principle,
so a wide flow range can be covered with the same
type of instrument.
- The material of the flow tube can be either stainless
steel, PEEK or fused silica; other materials may also
prove to be feasible.
- The length of the flow sensor tube is the same
for all flow ranges. This enables a modular set-up
and exchangeability of the instruments.
- The active elements are placed outside the flow
tube, so all wetted parts are either stainless steel,
PEEK, or fused silica.
Electronic circuitry
Both measurement principles, CP and CT, need their
own specific electronic circuitry, which is based upon
a wheatstone bridge configuration. The electronic circuitry
converts the output signals, namely ∆T
for the CP and Pheater for the CT measurement principle,
into an output voltage, showing a linear relation with
the mass flow.
Flow control
Flow control is achieved by integrating a control
valve onto the body of the liquid flow metre. This control
valve has a purge connection on top of the sleeve that
enables easy elimination of air or gas when starting
up the system. The electronic control function forms
part of the standard circuitry in the liquid flow metre,
so the need for an external controller is eliminated.
Test procedure
For the experiments several liquid flow sensors were
used, with flow tubes made out of stainless steel, PEEK
and fused silica, and with internal diameters varying
between 20 and 200 µm.
Furthermore a liquid flow controller was used, comprising
one of the sensors working according to the constant
temperature measurement principle. The output signal
of the flow sensors was measured for flow ranges varying
between 25-500 nl/min and 100-2,000 µl/min
water. The mass flow controller was used for measurements
of the dynamic behaviour. The following three stepwise
variations in set-point were performed:
1 0% → 100% →
0%;
2 20%→ 80%→
20%;
3 20%→ 40%→
60% → 80% →
100%:
A digital oscilloscope measured the resulting response
of the mass flow sensor.
Test results
The measured output signals as a function of the mass
flow are shown in Figures 2-4. The measured dynamic
behaviour of the mass flow controller is shown in Figure
5.
Figure 2 Measurement results obtained
with two liquid flow sensors for H2O with flow range:
(a) 500 nl/min¼100%; and (b) 10 µl/min
¼ 100% both working according to the CP measurement
principle; the flow tubes are made of stainless steel
Figure 3 Measurement results obtained
with two liquid flow sensors for H2O both working according
to the CP measurement principle with flow range 2,000
nl/min¼100%; the flow tubes are made of: (a)
fused silica; and (b) PEEK
Figure 4 Measurement results obtained
with a liquid flow sensor for H2O working according
to the CT measurement principle with flow range 2,000
µl/min¼100%;
the flow tube is made of stainless steel
The measured curves, as shown in Figures 2-4, correspond
well with the theoretically expected values. The measured
response times, as shown in Figure 5, are all within
the value of t98%≈
2 s.
Conclusions
The new generation of liquid flow sensors presented
in this paper is capable of meeting the requirements
as imposed by the life science, analysis, biotech and
other markets. The actual flow sensor consists of a
straight flow tube with two active elements around it.
The internal diameter of the flow tube may vary between
20 and 200 µm.
The corresponding internal volume of the sensor tube
is 15 nl and 1.5 µl,
respectively, which is extremely small.
Instruments with a flow tube made of stainless steel,
fused silica, or PEEK, were tested successfully. The
feasibility of other materials suitable for the above-mentioned
markets will be further investigated. The flow sensors
have been driven with two different measurement principles,
namely the CP and the CT method. The CP measurement
principle proved to be useful for flow ranges below
circa 100 µl/min,
the CT measurement principle was suited for flow ranges
above circa 100 µl/min.
The tested instruments were capable of measuring flow
ranges between 25. . .500 nl/min (smallest flow range)
and 100. . .2,000 µl/min
(largest flow range) water, with operating pressures
up to 100 bar (up to 400 bar for flow metres with flow
ranges below 100 µl/min).
The measured response time of the flow sensors is of
the order of t98% ≈
2 s: The feasibility of the new measurement concept
for low liquid flows has also been field-proven. Products
based upon this new technology have been introduced
on the market in 2003 and were found to be a reliable
solution for many low liquid flow applications, for
example in HPLC systems (Figures 6 and 7).
Figure 5 Measured response times of
a liquid flow sensor adjusted for H2O: 2,000 µl/min¼5
V; the X-axis shows the time [s] with 1 s/div., the
Y-axis displays the output voltage [V] with 1 V/div.;
the flow tube is made of stainless steel; the sensor
is working according to the CT measurement principle
Figure 6 Example of application: liquid flow
control in HPLC
Figure 7 The Bronkhorst m-flow metre and control
unit
|
