Background of the Invention
This invention relates generally to the quantitative detection of
concentrations of gases, and more particularly to methods and apparatus for detecting
concentrations of a gas based on its reaction with mercuric oxide.
Reduction gas detectors operate on the principle of flowing a gas
stream to be analyzed through a heated bed of mercuric oxide (HgO). Gases in the
stream that can be oxidized (referred to as "reducing gases"), react with the mercuric
oxide to produce free mercury vapor as shown in the following general reaction:
X + HgO → XO + Hg
In this equation, X represents a reducing gas species and Hg is present
as free mercury vapor. The mercury vapor produced in this reaction can be detected
by its absorption of ultraviolet (UV) light within a sample cell forming a part
of an ultraviolet photometer. An example of a reduction gas detector can be found
in US Patent No. 4,411,867 of Ostrander, incorporated herein by reference.
Reactions with mercuric oxide are not specific to any particular gas
species and a large number of reducing gases can react with mercuric oxide to produce
mercury vapor. Gas measurement apparatus intended for quantitative measurements
of specific gas species must therefore incorporate some process for isolating the
gas species to be measured. One such apparatus is a gas chromatograph, which time-separates
the gas sample into individual species. More particularly, this separation is obtained
using a long tube or "column" through which flows a gas stream. The exit gas flow
from the column is connected to the reduction gas detector and an apparatus for
injecting a precise volume of sample gas into the gas stream is located upstream
of the column. The column itself is packed with a granular substance which has
the characteristic of separating the different gases comprising the sample based
on their molecular size or other chemical properties. In the case of columns containing
molecular sieve materials, small molecules such as H2 will flow through
the column faster than large molecules such as CO. It will therefore be appreciated
that the difference in such properties cause each species or element of the sample
to move through the column and into the detector at different times, and the gas
species are detected as a series of Gaussian-shaped concentration "peaks." Starting
from a single sample injection onto the column, each peak arrives at the detector
in a characteristic time and the peak itself is essentially comprised of a single
gas species. The height of each peak, or the integrated area under each peak, is
representative of the concentration of the gas species.
In the prior art, reduction gas detectors have typically been operated
at temperatures of 150-300°C in order to promote the desired reactions with mercuric
oxide. The sample cell as well as the mercuric oxide bed were heated in this temperature
range in order to prevent mercury from condensing on the interior surfaces of the
sample cell. As is well known to those skilled in the art, mercury vapor is quite
condensable and adheres to relatively cool surfaces. Mercury condensation within
the sample cell can result in slow equilibration of the sample cell to changing
mercury concentrations and therefore slow time response of reduction gas detectors.
Additionally, ultraviolet sample cells include quartz (i.e. pure SiO2)
windows which allow ultraviolet radiation to be transmitted through the cell. Mercury
condensation on the quartz windows reduces the optical transmission of the cell
due to absorption of the ultraviolet radiation by mercury condensation on the windows.
This results in reducing signals for UV light sensors in the photometer, and correspondingly
higher noise levels.
In general, gas detectors used in conjunction with gas chromatography
must have relatively fast response times in order to accurately follow the concentration
peaks created by the chromatography column. Additionally, typical gas chromatography
flow rates are in the range of 20-60 cc/minute which are much lower than the 500-2000
cc/min flow rates associated with other gas measurement techniques (e.g. continuous
analyzers). Gas chromatography detectors therefore preferably have small internal
volumes in order to minimize concentration equilibration times to rapidly changing
gas concentrations, and to reduce condensation of the flowing gas species as described
Sample cells of the prior art, when embodied as a continuous sampling
analyzer, were, of necessity, quite large in order to accommodate the large gas
flows through the detectors. The large diameters of the prior art continuous sampling
analyzer cells also transmitted relatively large quantities of ultraviolet radiation,
which was desirable to reduce noise levels in the detector output signal. Sample
cells of the prior art for chromatography detectors were smaller than those used
for continuous sampling detectors but were still limited to a minimum diameter of
0.15 cm and a maximum length of 10 cm which were the dimensions that could still
transmit adequate amounts of ultraviolet light through the passageway of the cell.
That is, the diameter of the passageway of the cell was kept fairly large and the
length of the cell was kept fairly short, so that a sufficient amount of light
from the ultraviolet source could travel through the cell and still be detected
by the ultraviolet (UV) sensor. This is because ultraviolet sources are non-coherent
and, therefore, the amount of light impinging upon the UV detector is directly
proportional to the diameter of the cell passageway and is inversely proportional
to the square of the length of the cell. Hence, short, large diameter cells were
the norm in the prior art.
The temperature of prior art chromatography detector cells were maintained
at the same temperature as the HgO beds which, in practice, was in the range of
265-285°C. Based on this relatively high temperature, the optical windows of the
cell were constructed of relatively long quartz rods (approximately 5 cm in length)
in order to isolate the hot cell from the temperature-sensitive ultraviolet lamp
and light sensor. The amount of UV light transmitted through these rods is also
quite dependent on temperature of the rod and, therefore, minor changes in rod
temperature affect the amount of light impinging on the UV sensor. Minor variations
in convective cooling of the rods of the prior art heated detector cells therefore
introduced variations in light transmitted through the cell which were not due
to mercury vapor concentration. The net affect of these variations was to increase
drift and noise in the output of the light sensor.
It will therefore be appreciated that the performance of the prior
art chromatography cell was limited by: a) the relatively large cell diameter and
short length required for transmission of suitable levels of UV light; b) the relatively
large condensation surface area of the cell due to its diameter and length; and
c) the relatively high cell temperature which necessitated the requirement for
optical windows comprised of quartz rods which added thermally-induced drift and
noise to the detector output.
Since the sensitivity of mercury detection is directly proportionate
to cell length, the ideal sample cell would be infinitely long and have zero diameter,
zero internal volume, and zero internal surface area when one ignores other factors
such as the amount of light in gas that could travel down the passage way of such
an ideal sample cell. Additionally, the optical cell windows, if heated, would
ideally be infinitely thin and therefore not prone to produce thermal convection
Summary of the Invention
A preferred embodiment of the present invention is directed to an
improved photometer for detecting mercury vapor in a low flow-rate carrier gas.
As such, it is well suited for gas chromatography for species that can be reduced
in a heated mercuric oxide bed.
The sample cell of the improved photometer of the present invention
is long and thin, as compared to sample cells of the prior art. The low internal
surface area has eliminated the need to heat the cell, which permits very thin
optical cell windows, which are essentially not prone to the production of thermal
convention errors. The present invention stabilizes the temperature of an intense
UV light source to provide sufficient, low noise UV light through the long, thin
sample cell. As such, a fast, highly sensitive, and reliable photometer is provided
by the method and apparatus of the present invention.
A preferred embodiment of the present invention therefore relates
to detecting small concentrations of gases by measuring the spectral absorption
of mercury vapor produced by those gases in a reduction process with a heated mercuric
oxide bed. The apparatus includes an elongated cylindrical sample cell preferably
operated at ambient temperatures and optimized to have a long passageway to increase
the sensitivity of the photometer. A quartz window assembly is provided at each
end of the sample cell such that ultraviolet light can be directed into a first
window assembly, through the passageway of the sample cell, and out of a second
window assembly to impinge upon an ultraviolet detector.
By providing a sample cell that is very long in proportion to the
diameter of the passageway the need for heating the sample cell has been eliminated.
Preferably, the sample cell is made from stainless steel, aluminum, or borosilicate
glass. Also preferably, the ratio of the length of the sample cell to the diameter
of the passageway through the sample cell is at least 100 to 1, which reduces internal
surface area upon which mercury can condense and which increases the sensitivity
of the cell.
The quartz window assemblies are preferably provided with individual
heaters to encourage the evaporation of condensates on the windows. The ultraviolet
lamp is also preferably provided with a heater, a heat sink, and closed loop control
system to maintain the temperature of the lamp within precise limits. The elongated
sample cell is preferably held in a V-block arrangement to provide a straight optical
path through the passageway of the sample cell.
It will therefore be appreciated that a photometer of the present
invention includes an elongated sample cell having a first end, a second end, and
an elongated passageway extending between the first end and the second end. Preferably,
a ratio of a length of the sample cell to a lateral dimension of the passageway
is at least 100 to 1. Furthermore, the cell is preferably maintained at about ambient
temperature. A first quartz window assembly is located at the first end of the
sample cell and has a first port communicating with the passageway proximate to
the first end, and a second quartz window assembly is located at the second end
of the sample cell and has a second port communicating with the passageway proximate
to the second end. A source of electromagnetic radiation (preferably UV radiation)
is positioned to emit electromagnetic radiation through the first quartz window,
the passageway, and the second quartz window, and a detector of electromagnetic
radiation (preferably a UV detector) positioned to receive electromagnetic radiation
emitted through the second quartz widow. Preferably, the sample cell is operated
at about ambient temperature, and the volume of the sample cell is no greater than
about 0.2 cc to provide fast transient response.
A method for measuring mercury vapor concentration in accordance with
the present invention includes flowing a carrier gas through a mercuric oxide bed
and then through a passageway of an elongated sample cell, where the sample cell
has a length and the passageway has a lateral dimension such that a ratio of the
length to the lateral dimension is at least 100 to 1. An ultraviolet light is directed
through the cell to impinge upon a detector, and an output signal of detector is
zeroed. Next, a gas sample is inserted into the flow of the carrier gas, where the
gas sample comprises one or more substances that can be reacted with a mercuric
oxide bed to form a mercury vapor. Finally, the output signal of the detector is
The present invention provides a number of advantageous features over
the prior art. For one, the sample cell is not heated, eliminating costly and potentially
unreliable heaters and heater control systems. Furthermore, by not heating the
cell, the quartz windows can be made much shorter than in the prior art, eliminating
the noise component caused by small localized variations in temperature due to
convention currents. Finally, the long sample cell of small diameter provides superior
sensitivity and faster response time than shorter, wider cells of the prior art.
These and other advantages of the present invention will become apparent
upon a reading of the following detailed descriptions and a study of the several
figures of the drawings.
Brief Description of the Drawings
Detailed Description of the Preferred Embodiments
- Fig. 1 is a side elevational view of the gas detector of the present invention;
- Fig. 1A is a cross-sectional view taken along line 1A-1A of Fig. 1;
- Fig. 2 is an enlarged, cross sectional view of the of the lamp end assembly
of the present invention;
- Fig. 3 is an enlarged cross sectional view of the quartz window assembly of
the present invention;
- Fig. 4 is an enlarged cross sectional view of the detector end assembly of the
- Fig. 5 is a block diagram that illustrates the functional elements and operation
of the present invention;
- Fig. 6 is a flow diagram illustrating the operations involved in detecting small
concentrations of gases in accordance with the present invention; and
- Figure 7 is a flow diagram illustrating the operations involved in zeroing the
photometer of the present invention.
In Fig. 1, a photometer 10 in accordance with the present invention
includes an elongated sample cell 12, a first quartz window assembly 14, a second
quartz window 16, a lamp assembly 18, and a detector assembly 20. The sample cell
12 is supported by a V-block unit 22 and is held in place by a clamp unit 24. The
various components are supported by a base 26, which can further support other
components such as a lamp inverter 28 and detector output electronics 30. During
operation, many of the components are covered with one or more covers 32, 34, and
As seen in Fig. 1, the sample cell is an elongated structure, preferably
formed as a tube, having a length L which, in a preferred embodiment of the present
invention, is about 30 cm. With additional reference to the cross-sectional view
of Fig. 1A, the sample cell is supported by a plurality of V-groove blocks 38 having
V-grooves 40 and is held in place by clamp 42. The long sample cell 12 requires
precise alignment to allow the UV light to shine down the internal passageway without
excessive loss. The V-groove blocks 38 provide this support and alignment. Similar
V-groove block arrangements have been used in the laser arts to precisely align
laser rods, as will be appreciated by those skilled in the art. The clamps 42 hold
the sample cell 12 firmly within the V-grooves 40 of the V-groove blocks 38.
As can also be seen in Fig. 1A, the cell 12 is provided with a passageway
44. Preferably, this passageway is a cylindrical bore or the like, such that the
lateral dimension "d" of the passageway is, essentially, the diameter of the bore.
Alternatively, the passageway may not be cylindrical, in which case a maximum lateral
dimension is defined as the maximum diameter of the bore taken perpendicularly
to an axis of the cell. However, the walls of the passageway should be smooth (e.g.
electropolished or hydraulically bored) to a finish of 20 RA or less to inhibit
mercury from adhering to the surface.
It should be noted that the sample cell 12 is very long in relation
to the lateral dimension of the passageway 44. In the present example, the passageway
is cylindrical (the cell 12 forming a tube) such that the lateral dimension d is
about .040 cm in diameter. Since the length L of the cell 12 is 30 cm in this example,
the ratio of the length L to the lateral dimension d is L/d = 750:1. This provides
very good sensitivity, quick response time, and minimal internal surfaces (wall
area) of the passageway 44 to which mercury can stick. However, if higher volumes
of sample gas and/or greater lamp intensity is desired at the detector, this ratio
can be reduced to as little as 100:1 in some instances, although it is preferable
that it is at least 250:1. It is desirable, nonetheless, to have a total cell volume
of no more than about 0.2 cc, in this preferred embodiment.
The material of the sample cell is preferably one or more of a borosilicate
glass, stainless steel, or aluminum. If the sample cell is made from a borosilicate
glass, it is preferably encased in a stainless steel tube for protection.
It has been found that by providing a ratio of cell length to passageway
diameter of less at least 100:1 that the heater required in prior art reduction
gas detectors can be eliminated. That is, the cell 12 can be operated below about
150°C, in contrast to sample cells of the prior art. In fact, the cell 12 can be
operated below 100°C and even at ambient temperatures (about 25°C) without creating
a substantial problem from the condensation of mercury vapor on the inner walls
of the passageway.
In Fig. 2, the first quartz window assembly 14 and the lamp assembly
18 are shown in greater detail. The first quartz window assembly includes a heater
block 46 provided with a resistive heater 48 in a bore 50. The heater block is
preferably made from a suitable metal such as aluminum, and serves to stabilize
the heat from the resistive heater 48. A window unit 52 is attached to the cell
12 by a fitting 54. If the sample cell 12 is stainless steel, the fitting 54 is
preferably brazed to the sample cell. If the sample cell is borosilicate glass,
the fitting is preferably glued to the sample cell with a suitable adhesive.
The lamp assembly 18 includes a heater block 56 made, again, preferably
from a good thermal storage material such as aluminum. A lamp 58 is preferably
positioned within a bore 60 in the heater block 56. In this preferred embodiment,
the lamp is an ultraviolet (UV) lamp having operating frequency centered at about
254 nanometers, and is available from a variety of sources. For example, such lamps
are commercially available from BHK, Inc. of Claremont, California. A resistive
heater 62 is coupled to the heater block, and a thermocouple 64 is disposed within
a bore 66 of the heater block 56. Heat sinks 68 couple the heater block 56 to the
base 26 to draw heat from the heater block. The heat sinks 68 are preferably made
from the same metal as the heater block 56.
It is desirable that the heater 62 and the heat sinks 68 have about
the same time constants. This makes it easier to maintain the temperature of the
lamp assembly 18 with a very tight tolerance (e.g. within about .05 degree centigrade)
when forming a part of closed-loop temperature controller, as will be discussed
in greater detail subsequently. It is important to maintain this accurate temperature
control since the present invention does not utilize a reference detector proximate
to the lamp 58, as was the case in the prior art. By maintaining a very accurate
fixed temperature on the lamp 58, the UV light output by the lamp will be a constant,
eliminating the need for such a reference detector. A tube 72 (preferably metal)
extends from the lamp 58 to the quartz window 70 to shade that portion of the optical
path from stray ambient light.
It is to be re-emphasized that, in the past with other instruments,
a reference detector was required to produce a signal Vref that was used in conjunction
with the output signal Vsig of the main detector to create the output signal Vout.
In practice, these two signals could not be accurately measured with any consistency.
In the prior art, the output signal Vout was calculated by the log(Vref/Vsig).
By stabilizing the temperature of the ultraviolet source, the reference signal
Vref becomes a constant and does not have to be measured. Also, as long as the mercury
vapor concentration is less that 50 parts per billion (ppb) in a 30 cm long cell,
the changes in the detected signal Vsig are so small that it is approximately linearly
proportional to logVsig. That is, when the mercury vapor concentration is less
than about 2ppb per centimeter of cell length, the output signal from the detector
is generally linearly proportional to the absorption of ultraviolet radiation by
mercury vapor in the cell. As a result, the output signal Vout becomes essentially
equivalent to the detected signal Vsig. Thus, with the present invention not only
is the need for a reference detector eliminated, but also costly logarithmic processing
of the signal is eliminated.
In Fig. 3, a window unit 52 is shown in cross-section. The window
unit includes a body that is coupled to the sample cell 12 by the fitting 54. A
small bore 76 communicates with the passageway of the cell 12. A feed tube 78A
forms a port 80 which communicates with the bore 76 and, therefore, the passageway
of the sample cell. In this embodiment, the port 80 is an outlet port for injecting
gas into the sample cell, although it could equally well be an inlet port releasing
gas from the sample cell. A disk-shaped quartz window 70 is sandwiched between
two Teflon washers 82 and 84. A more rigid washer 86 (e.g. a metal washer) forms
a bearing surface for a spring 88 which is held in place by a retainer bolt 90.
The window unit 52 therefore forms a gas-tight seal to the end of the sample tube
12 with only gas port 80 for the ingress or egress of gas. A bore 92 in the bolt
90 is receptive to the light guide 72.
It should be noted that the thickness "t" of the quartz window 70
is much less than that required in the prior art. This is because the sample cell
is operated at lower temperatures than in the prior art, and a thick window is
not required to dissipate the heat of the sample cell. As such, the quartz window
is much less susceptible to changes in the index of refraction due to large temperature
gradients along its length. In this preferred embodiment, the thickness t of the
window is about 2.5 millimeters. Preferably, the thickness is no greater than 1.25
millimeters, and preferably it is less than .625 millimeters, and most preferably
is no thicker than is required for structural integrity.
By heating the window 70 to a temperature of at least about 80°C with
the heater 48 and heater block 46, any mercury condensate on the window 70 can
be cause to evaporate over time. This "cleaning" feature enhances the operation
of the window 70 by permitting more light to enter the sample cell. While at least
about 80°C is one preferred temperature range in which to heat the windows, a temperature
of about 50°C or greater can also be used.
In Fig. 4, the window assembly 16 and detector assembly 20 are shown
in cross section. The window assembly 16 is of essentially the same construction
as the window assembly 14 described previously, but is oriented in the opposite
direction. Therefore, the construction of window assembly 16 can be considered
to be a mirror image of the construction of window assembly 14. The same numerals
have been used to indicate the same elements in window assemblies 14 and 16.
The detector assembly 20 includes a mounting block 94 having a first
bore 96 receptive to a tube 72 (which blocks stray ambient light from the light
path) and a second bore 98 receptive to a UV filter 100. The filter 100 is retained
by a washer 102 and a nut 104. An ultraviolet detector 106 can be mounted on a
printed circuit (PC) board 108, as will be appreciated by those skilled in the
art. UV detector 106 is commercially available from a number of sources such as
EG&G Electro-Optics Division of Salem Massachusetts and Hamamatsu Photonics,
K.K. of Hamamatsu City, Japan.
In Fig. 5, a functional block diagram of the photometer 10 will be
used to describe the operation of the present invention. Items previously described
are shown in a diagrammatic form and are referenced with the same numerals as previously
used. The UV lamp 58 produces UV light 110 which goes through quartz window 70
of window assembly 14, through the passageway of sample cell 12, through the quartz
window 70 of the window assembly 16, through optical filter 100, and impinges upon
detector 106. The detector 106 produces a signal Vsig, which is processed in a
signal processor (e.g. an analog-to-digital (A/D) converter) to produce a digital
signal Vout which represents the concentration of mercury vapor in the sample cell
12 and, therefore, the concentration of the reduced gas being measured in the sample.
A temperature controller 114 is used to maintain the temperature of
the UV lamp 58. More particularly, the controller 114 is responsive to an output
of the thermocouple 64 and controls the current flowing through heater 62. The
heat sink arrangement described previously aids in the precise maintenance of the
lamp temperature using this closed-loop feedback system.
A window temperature controller 116 likewise controls the temperature
of the quartz window 70 by controlling the current flowing through the heaters
48. The temperature of the windows are, in this example, maintained at about 80°C
to provide self-cleaning of deposited materials. Preferably, the temperature is
maintained at a constant level with a feedback loop type controller, as described
There may be one or more temperature controllers associated with the
heated mercury bed 118 and the feed tube 78A. A bed temperature controller 119
controls a resistive heater 121 to maintain the bed 118 within an operating temperature
range (e.g. 265-285°C), as is well known to those skilled it the art. An optional
feed tube temperature controller 79 controls a resistive heater 81 to inhibit condensation
of mercury vapor within feed tube 78A. Like the windows, the feed tube is preferably
heated to at least about 50°C, and more preferably about 80°C or more. These and
the other temperature controllers preferably under the control of a master system
controller (not shown).
In operation, a mercury oxide bed 118 is heated by heater 121, and
a sample gas is caused to flow through the mercury oxide bed. Gaseous components
that can be oxidized will be reduced by the mercury oxide bed, resulting in the
creation of mercury vapor which flows through tube 78A, through the sample cell
12, and out an outlet tube 78B along with the carrier gas. Since mercury vapor
strongly absorbs UV light, the detected light level will drop as the mercury vapor
level within the cell 12 rises. The resulting waveform can be analyzed to determine
the concentration levels of the reduced gases in the carrier gas.
Fig. 6 is a flow diagram illustrating the operations (process) 119
performed by the gas detection instrument to detect and analyze small concentrations
of gases in accordance with the present invention. First, an inert gas is caused
to flow through the sample cell in an operation 120. Next, in an operation 122,
the system heaters are turned on. For example, window heaters 48, lamp heater 64,
and bed heater 121 are turned on at this time. If there is a feed tube heater 81,
it is also turned on at this time. Controlled by the lamp temperature controller
114, the lamp heater 62 applies the proper amount of heat to the heater block in
order to precisely stabilize the temperature of the ultraviolet lamp, which in
turn stabilizes the operating frequency, reduces noise, and otherwise enhances
the performance of the ultraviolet lamp.
With the inert gas flowing and the heaters on, the gas detector is
then stabilized, e.g. for about 15 minutes or more, in an operation 124. This "Stabilize
Baseline" operation 124 is sometimes referred to as "baseline stabilization", since
it provides a baseline reference against which subsequent measurements can be compared.
After baseline stabilization, the detector goes through a zeroing
algorithm in an operation 126. This operation 126 is discussed in greater detail
below with reference to Fig. 7. Briefly, with the presence of a gas sample to be
analyzed flowing through the sample cell, the gas detector in operation 126 uses
a zeroing algorithm to establish a zero baseline output by the detector prior to
the injection of the sample.
In operation 128 it is determined whether another gas sample is to
be injected into the carrier gas stream. If there is, in an operation 130, the
detector is utilized to analyze the concentration of mercury vapor in the carrier
gas, as previously described. Process control then returns to operation 126 to
prepare for a possible additional sample. If there are no more samples to be injected,
the system is powered down as indicated at 132.
In Fig. 7, the operation 126 of performing the zeroing algorithm is
described in greater detail. To set the baseline signal, signal Vsig in operation
132 is measured. To insure that the signal is Vout is in a proper range, the zeroing
circuitry (typically an operational amplifier controlled by a D/A converter) is
adjusted in an operation 134, as will be appreciated by those skilled in the art.
The output signal Vout is then measured in an operation 136. If the signal Vout
is out of range, process control is returned to operation 134 to again adjust the
zeroing circuitry (not shown). When the signal Vout is within range as determined
by operation 138, the process is complete as indicated at 140.
Although the foregoing invention has been described in some details
for purposes of clarity and understanding, it will be apparent that certain changes
and modifications may be practiced within the scope of the appended claims. Accordingly,
the present embodiments are to be considered as illustrative and not restrictive,
and the invention is not to be limited to the details given herein, but may be
modified within the scope and equivalents of the appended claims.