DETECTOR FOR DETECTING ANALYTES IN GAS PHASE COMPRISING POROUS DIELECTRIC OR SEMICONDUCTING SORBENT AND CORRESPONDING DETECTION METHOD

Abstract

A detector for, and a method of, detecting analytes in gases in described. The detector comprises a sorbent for sorbing therein and/or thereon and/or desorbing therefrom, an analyte included in a gas exposed thereto, at a zeroth temperature, pressure (T.sub.0,P.sub.0), a controller arranged to change the zeroth temperature, pressure (T.sub.0,P.sub.0) to a first temperature, pressure (T.sub.1,P.sub.1) according to a first equation, to desorb and/or sorb at least some of the analyte; and a sensor arranged to sense at least some of the analyte and to output a response corresponding to the sensed analyte. The response comprises and/or is a characteristic response of the analyte. The first response is modified based on a first baseline response at the zeroth temperature, pressure (T.sub.0,P.sub.0).

Claims

1-17. (canceled)

18. A detector for detecting analytes in gas-phase, comprising: a first sorbent of a set of sorbents, wherein the first sorbent comprises and/or is a microporous and/or mesoporous dielectric or semiconducting material such as silica or silicon, a zeolite, activated carbon and/or a metal organic framework, MOF, for sorbing therein and/or thereon and/or desorbing therefrom, a first analyte of a set of analytes included in a first gas of a set of gases exposed thereto, at a zeroth temperature, pressure (T.sub.0,P.sub.0) of a set of temperatures, pressures (T,P), wherein the first gas is ambient air; a cooler configured to cool the first sorbent to the zeroth temperature (T.sub.0); a controller configured to change the zeroth temperature, pressure (T.sub.0,P.sub.0) to a first temperature, pressure (T.sub.1,P.sub.1) of the set of temperatures, pressures (T,P) according to a first equation of a set of equations, to desorb and/or sorb at least some of the first analyte; and a sensor configured to sense at least some of the first analyte and to output a first response of a set of responses corresponding to the sensed first analyte, wherein the sensor comprises and/or is a broadband analyte detector such as a photoionization detector, PID; wherein the detector is configured to obtain a first baseline response of a set of baseline responses at the zeroth temperature, pressure (T.sub.0,P.sub.0) and wherein the controller is configured to modify the first response based, at least in part, on the obtained first baseline response; and wherein the first response comprises and/or is a first characteristic response of a set of characteristic responses of the first analyte.

19. The detector according to claim 18, wherein the controller is configured to compare the first response with a set of reference responses.

20. The detector according to claim 19, wherein the controller is configured to match the first response with a first reference response of the set of reference responses.

21. The detector according to claim 18, wherein the controller is configured to change only the zeroth temperature (T.sub.0) to the first temperature (T.sub.1).

22. The detector according to claim 18, wherein the controller is configured to change only the zeroth pressure (P.sub.0) to the first pressure (P.sub.1).

23. The detector according to claim 18, comprising a heater and/or cooler configured to heat and/or cool the first sorbent and/or the sorbed first analyte to the first temperature, pressure (T.sub.1,P.sub.1).

24. The detector according claim 23, wherein the heater is thermally coupled to the first sorbent.

25. The detector according claim 23, wherein the heater is in contact with the first sorbent.

26. The detector according to claim 18, wherein the controller is configured to change the zeroth temperature, pressure (T.sub.0,P.sub.0) to the first temperature, pressure (T.sub.1,P.sub.1) according to a first survey equation of a set of survey equations, before changing the zeroth temperature, pressure (T.sub.0,P.sub.0) to the first temperature, pressure (T.sub.1,P.sub.1) according to the first equation.

27. The detector according to claim 18, comprising a pump configured to flow the first gas across and/or through the first sorbent.

28. The detector according to claim 18, wherein: the first sorbent is for sorbing therein and/or thereon, a second analyte of the set of analytes included in the first gas of a set of gases exposed thereto, at the zeroth temperature, pressure (T.sub.0,P.sub.0); the controller is configured to change the first temperature, pressure (T.sub.1,P.sub.1) to a second temperature, pressure (T.sub.2,P.sub.2) of the set of temperatures, pressures (T,P) according to a second equation of the set of equations, to desorb at least some of the sorbed second analyte from the first sorbent; the detector is configured to detect at least some of the desorbed second analyte and to output a second response of the set of responses corresponding to the sensed second analyte; and the second response comprises and/or is a first characteristic response of a set of characteristic responses of the second analyte.

29. The detector according to claim 28, wherein the sorbing is adsorbing and the sorbed second analyte is adsorbed second analyte.

30. A method of detecting analytes in gas-phase, comprising: exposing a first sorb ent of a set of sorbents, wherein the first sorbent comprises and/or is a microporous and/or mesoporous dielectric or semiconducting material such as silica, a zeolite, activated carbon and/or a metal organic framework, MOF, to a first gas of a set of gases, at a zeroth temperature, pressure (T.sub.0,P.sub.0) of a set of temperatures, pressures (T,P), wherein the first gas is ambient air; cooling the first sorbent to the zeroth temperature (T.sub.0) and sorbing by the first sorbent, therein and/or thereon, a first analyte of a set of analytes included in the first gas; desorbing and/or sorbing at least some of the first analyte by controlling a change from the zeroth temperature, pressure (T.sub.0,P.sub.0) to a first temperature, pressure (T.sub.1,P.sub.1) of the set of temperatures, pressures (T,P) according to a first equation of a set of equations; sensing, by a sensor, at least some of the first analyte and outputting a first response of a set of responses corresponding to the sensed first analyte, wherein the sensor comprises and/or is a broadband analyte detector such as a photoionization detector, PID; and obtaining a first baseline response of a set of baseline responses at the zeroth temperature, pressure (T.sub.0,P.sub.0) and modifying the first response based, at least in part, on the obtained first baseline response; wherein the first response comprises and/or is a first characteristic response of a set of characteristic responses of the first analyte.

31. The method according to claim 30, comprising comparing the first response with a set of reference responses.

32. The method according to claim 31, comprising matching the first response with a first reference response of the set of reference responses.

33. The method according to claim 30, comprising controlling only the change the zeroth temperature (T.sub.0) to the first temperature, pressure (T.sub.1).

34. The method according to claim 30, comprising, controlling only the change from the zeroth pressure (P.sub.0) to the first pressure (P.sub.1).

35. The method according to claim 30, comprising changing the zeroth temperature, pressure (T.sub.0,P.sub.0) to the first temperature, pressure (T.sub.1,P.sub.1) according to a first survey equation of a set of survey equations, before changing the zeroth temperature, pressure (T.sub.0,P.sub.0) to the first temperature, pressure (T.sub.1,P.sub.1) according to the first equation.

36. The method according to claim 30, comprising flowing the first gas across and/or through the first sorbent.

37. The method according to claim 30, comprising: sorbing, preferably adsorbing, by the first sorbent, therein and/or thereon, a second analyte of the set of analytes included in the first gas of a set of gases exposed thereto, at the zeroth temperature, pressure (T.sub.0,P.sub.0); desorbing at least some of the sorbed, preferably adsorbed, second analyte from the first sorbent by changing the first temperature, pressure (T.sub.1,P.sub.1) to a second temperature, pressure (T.sub.2,P.sub.2) of the set of temperatures, pressures (T,P) according to a second equation of the set of equations; and sensing at least some of the desorbed second analyte and outputting a second response of the set of responses corresponding to the sensed second analyte; wherein the second response comprises and/or is a first characteristic response of a set of characteristic responses of the second analyte.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0136] For a better understanding of the invention, and to show how exemplary embodiments of the same may be brought into effect, reference will be made, by way of example only, to the accompanying diagrammatic Figures, in which:

[0137] FIG. 1 schematically depicts a detector according to an exemplary embodiment;

[0138] FIG. 2 schematically depicts a method according to an exemplary embodiment;

[0139] FIG. 3 schematically depicts a detector according to an exemplary embodiment;

[0140] FIG. 4 shows a photograph of a detector according to an exemplary embodiment;

[0141] FIG. 5A shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a detector and a method according to an exemplary embodiment; FIG. 5B shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a detector and a method according to an exemplary embodiment; and FIG. 5C shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a detector and a method according to an exemplary embodiment;

[0142] FIG. 6 shows a graph of first response as a function of time for a detector and a method according to an exemplary embodiment;

[0143] FIG. 7 shows a graph of first response as a function of time for a detector and a method according to an exemplary embodiment;

[0144] FIG. 8 shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a first sorbent for a detector and a method according to an exemplary embodiment;

[0145] FIG. 9 shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a first sorbent for a detector and a method according to an exemplary embodiment;

[0146] FIG. 10 shows a graph of first response as a function of temperature fora sensor fora detector and a method according to an exemplary embodiment;

[0147] FIG. 11 shows a graph of temperature as a function of time and corresponding graphs of concentration of the first analyte, first gas and first response as a function of time for a detector and a method according to an exemplary embodiment;

[0148] FIG. 12 shows a graph of first response as a function of time in absence of a first analyte and during exposure of the first analyte;

[0149] FIG. 13 shows a graph of first response as a function of time for a detector and a method according to an exemplary embodiment;

[0150] FIG. 14 shows a graph of first response as a function of time for three different concentrations of a first analyte for a detector and a method according to an exemplary embodiment;

[0151] FIG. 15 shows a graph of first response as a function of time for three different concentrations of a first analyte for a detector and a method according to an exemplary embodiment;

[0152] FIG. 16 shows a graph of sensitivity increase for a detector and a method according to an exemplary embodiment;

[0153] FIG. 17 shows a graph of sensitivity increase for a detector and a method according to an exemplary embodiment;

[0154] FIG. 18 shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a detector and a method according to an exemplary embodiment;

[0155] FIG. 19 shows a graph of first response as a function of time for a detector and a method according to an exemplary embodiment;

[0156] FIG. 20 shows a graph of first response as a function of dwell time for a detector and a method according to an exemplary embodiment;

[0157] FIG. 21 shows a graph of first response as a function of time for N.sub.2, IPA, limonene and benzene for a detector and a method according to an exemplary embodiment;

[0158] FIG. 22 shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for IPA at four different concentrations for a detector and a method according to an exemplary embodiment;

[0159] FIG. 23 shows a graph of first response as a function of concentration for IPA;

[0160] FIG. 24 schematically depicts alternative arrangements of the detector, according to exemplary embodiments;

[0161] FIG. 25 schematically depicts alternative arrangements of the detector, according to exemplary embodiments;

[0162] FIG. 26 shows a graph of equivalent concentration (ppb) as a function of time (s) for IPA, MEK, benzene, toluene, limonene and o-Xylene;

[0163] FIG. 27 shows a graph of peak time (s) as a function of boiling point (° C.) for IPA, MEK, benzene, toluene, limonene and o-Xylene; and

[0164] FIG. 28A shows a graph of temperature as a function of time; and FIG. 28B shows a corresponding graph of first response as a function of time for a gaseous mixture of octene and IPA.

DETAILED DESCRIPTION OF THE DRAWINGS

[0165] FIG. 1 schematically depicts a detector 10 according to an exemplary embodiment. The detector 10 is for detecting analytes in gases. The detector 10 comprises a first sorbent 110A of a set of sorbents 100, preferably wherein the first sorbent 110A comprises and/or is an adsorbent, for sorbing, preferably adsorbing, therein and/or thereon and/or desorbing therefrom, a first analyte of a set of analytes included in a first gas of a set of gases exposed thereto, at a zeroth temperature, pressure (T.sub.0,P.sub.0) of a set of temperatures, pressures (T,P). The detector 10 comprises a controller 120 arranged to change the zeroth temperature, pressure (T.sub.0,P.sub.0) to a first temperature, pressure (T.sub.1,P.sub.1) of the set of temperatures, pressures (T,P) according to a first equation of a set of equations, to desorb and/or sorb at least some of the first analyte. The detector 10 comprises a sensor 130 arranged to sense at least some of the first analyte and to output a first response of a set of responses corresponding to the sensed first analyte; wherein the first response comprises and/or is a first characteristic response of a set of characteristic responses of the first analyte.

[0166] The detector 10 may be as described above with respect to the first aspect.

[0167] FIG. 2 schematically depicts a method according to an exemplary embodiment. The method is of detecting analytes in gases.

[0168] At S201, a first sorbent of a set of sorbents, preferably wherein the first sorbent comprises and/or is an adsorbent, is exposed to a first gas of a set of gases, at a zeroth temperature, pressure (T.sub.0,P.sub.0) of a set of temperatures, pressures (T,P).

[0169] At S202, the first sorbent sorbs, preferably adsorbs, therein and/or thereon, and/or desorbs therefrom, a first analyte of a set of analytes included in the first gas.

[0170] At S203, at least some of the first analyte is desorbed and/or sorbed by controlling a change from the zeroth temperature, pressure (T.sub.0,P.sub.0) to a first temperature, pressure (T.sub.1,P.sub.1) of the set of temperatures, pressures (T,P) according to a first equation of a set of equations.

[0171] At S204, at least some of the first analyte is sensed and a first response of a set of responses corresponding to the sensed first analyte is output. The first response comprises and/or is a first characteristic response of a set of characteristic responses of the first analyte.

[0172] The method may be as described above with respect to the second aspect.

EXPERIMENTAL

Introduction

[0173] The challenge of many sensors for airborne explosive is the accurate detection of extremely low concentrations. A goal of this work is to test nanoporous material (i.e. a first sorbent) as a pre-concentrator for a commercially available sensor and to measure the improvement in the sensitivity. For this purpose, a PID sensor is used; however, similar improvements could be assumed for other sensors.

[0174] This work details a method of using a detector comprising the nanoporous material to enhance the sensitivity of a PID sensor, which is a broadband VOC sensor. The temperature of the porous material is lowered to enhance the VOCs adsorption from a gas flow onto the material. The temperature is then raised up to desorb all the adsorbed VOCs at the same time, leading to a peak (i.e. a first response) in the concentration in the proximity of the nanoporous material. The size and/or shape of this peak can then be analysed to determine the concentration and/or identity of the VOC.

[0175] This method is found to give a gain of 50 to 70 times the steady-state change in signal of the sensor. This method has a further advantage that the zero point of the sensor is not required. This means that drifts in the zero point over the lifetime of the sensor do not lead to measurement errors. The shape of the peak also depends on the particular VOC, indicating that it should be possible to perform some degree of classification.

[0176] The detector comprises the porous material that acts as a VOC storage medium, the PID sensor, a Peltier heater/cooler and Peltier controller to set the temperature of the porous material. The temperature control is coordinated by a controller with the measurements from the sensor which also relates the update rate of the sensor to the time constants associated with adsorbing/desorbing the VOC and storing enough VOC to enhance the gain.

[0177] This work starts with an overview of the method, the analysis of the data and the detector. The effects of concentration for two different VOCs (limonene and IPA) are compared to a more traditional use of the sensor to demonstrate the gain in sensitivity. The effect of the time spent gathering VOCs is also tested showing that the technique does eventually saturate at long enough dwell times. A slower increase in temperature demonstrates that VOCs have different response characteristics which might be used for VOC classification.

Description of Measurement Process

[0178] A constant flow of gas containing a VOC (i.e. a first analyte of a set of analytes included in a first gas of a set of gases) is passed through a chamber containing a porous media on a Peltier heater/cooler and a broadband VOC PID sensor. In this report the first sorbent is porous silica etched into a silicon wafer and the VOC sensor is an Alphasense PID-AH, available from Alphasense Limited (UK).

Charge Phase

[0179] During the charge phase, the first sorbent is exposed to the first analyte of the set of analytes included in the first gas of the set of gases, at a zeroth temperature, pressure (T.sub.0,P.sub.0) of a set of temperatures, pressures (T,P), thereby sorbing therein the first analyte.

[0180] In more detail, the temperature of the Peltier is dropped so that the VOCs can deposit (particularly, adsorb) on the porous media. It is expected that VOCs will continue to deposit until an equilibrium with the VOC concentration in the gas is reached. The constant flow through the chamber keeps the source concentration of VOC (nearly) constant. The quantity of VOC which deposits on the porous media should be related to the surface area of the porous media, the dwell time at this temperature, the concentration of the VOC in the carrier gas and the surface chemistry between the porous media and the VOC. In this report the same sample of porous material is used throughout, so the surface area and the surface chemistry (with respect to the different VOCs) is fixed.

Measure Phase

[0181] During the measure phase, the zeroth temperature, pressure (T.sub.0,P.sub.0) is changed to a first temperature, pressure (T.sub.1,P.sub.1) of the set of temperatures, pressures (T,P) according to a first equation of a set of equations, to desorb at least some of the first analyte.

[0182] The temperature is then increased to evaporate the VOCs from the porous media. If this evaporation takes place over a short period of time then the concentration of VOC in the chamber should peak well above the concentration of VOC in the carrier gas. This excess concentration will gradually be removed by the constant gas flow. The size and shape of the peak in concentration should depend on the quantity of VOC released from the porous media and the sensitivity of the VOC sensor to the particular VOC. The change in temperature should be as fast as possible so as to increase the peak concentration, as the continuous gas flow will start to remove the excess VOCs.

[0183] Ideally the VOC is entirely evaporated (desorbed from the nanoporous material) at this temperature to gain the maximum response. It is possible to check that this is the case by having a further increase in temperature—if the VOC has been completely evaporated then this second increase in temperature will not release any further VOCs and there will be no resulting peak in concentration. If classification of VOCs is desired rather than the absolute sensitivity then a series of levels could be used to determine the temperatures at which the VOC evaporates. An alternative approach to classification is to ramp the temperature gradually to measure at which temperatures the peak occurs.

[0184] FIG. 3 schematically depicts a detector 30 according to an exemplary embodiment. The detector 30 is for detecting analytes in gases. The detector 30 comprises a first sorbent 310A of a set of sorbents 300 (not shown), preferably wherein the first sorbent 310A comprises and/or is an adsorbent, for sorbing, preferably adsorbing, therein and/or thereon and/or desorbing therefrom, a first analyte of a set of analytes included in a first gas of a set of gases exposed thereto, at a zeroth temperature, pressure (T.sub.0,P.sub.0) of a set of temperatures, pressures (T,P) The detector 30 comprises a controller 320 (not shown) arranged to change the zeroth temperature, pressure (T.sub.0,P.sub.0) to a first temperature, pressure (T.sub.1,P.sub.1) of the set of temperatures, pressures (T,P) according to a first equation of a set of equations, to desorb and/or sorb at least some of the first analyte. The detector 30 comprises a sensor 330 arranged to sense at least some of the first analyte and to output a first response of a set of responses corresponding to the sensed first analyte; wherein the first response comprises and/or is a first characteristic response of a set of characteristic responses of the first analyte.

[0185] In more detail, the detector 30 comprises: a silicon wafer 311A with an etched and oxidised porous silica layer 310A to act as the concentrator (i.e. the first sorbent 310A), having a thickness of 50 μm and a porosity of about 50%; an Alphasense PID-AH sensor (i.e. the sensor 330) as the broadband VOC sensor; the controller 320 (as described below); a Peltier 340 (i.e. a heater/cooler) and heatsink 350 to control the temperature of the porous silica 310A; and a PEEK cell (i.e. a housing 360) to contain the gas flow and mount the PID 330, silicon 311A/porous silica wafer 310A and the Peltier 340 (FIGS. 3 and 4).

[0186] The PEEK cell comprises a gas inlet 362 in fluid communication with a cylindrical chamber 364 via an inlet passageway 363, extending radially therefrom, at a first end of the housing 360, and a gas outlet 366 in fluid communication with the chamber 364 via an outlet passageway 365, extending radially therefrom, at a second end of the housing 360, diametrically opposed to the first end of the housing 360. The first sorbent 310A and the sensor 330 are arranged at opposed ends of the cylindrical chamber 364, in fluid communication therewith. The housing 360 has a diameter of 1.5 cm and a height of 0.5 cm, such that a volume between the first sorbent 310A and the sensor 330 is about 0.88 ml. Seals, for example O-rings, seal the first sorbent 310A and the sensor 330 against the housing 360 such that gas flow is restricted to a path defined by the gas inlet 362 and the gas outlet 366 via the chamber 364.

[0187] The porous silica was fabricated from a boron-doped silicon wafer with a resistivity of 0.01-0.02 Ωcm and (100) crystal orientation. Silicon was electrochemically etched in a 1:1 mixture of 48% hydrofluoric acid and ethanol under a current density of 120 mA/cm.sup.2, 288 cycles of 2 sec on 3 sec off. The sample was then thermally oxidised in a furnace under 50 sccm O.sub.2 flow for 16 hours at a temperature of 800° C. The final sample was 50 μm thick and had about 50% porosity. The gas was generated using an Owlstone Oven Vapour Generator, available from Owlstone Limited (UK), with a gas dilution rig to control the concentration and flowrate of the VOCs. This also provides a purge gas (99.99% N.sub.2) containing no VOCs to help clear the sensor and to provide background readings. The gas flow was controlled to be 0.050 SLM.

[0188] The system is controlled by the controller via software which coordinates the detector, logs the measurements (i.e. responses) and analyses the results.

[0189] FIG. 3 schematically depicts a detector according to an exemplary embodiment. In more detail, FIG. 3 shows a cross-section of the detector comprising the porous media, Peltier and broadband VOC sensor.

[0190] FIG. 4 shows a photograph of a detector according to an exemplary embodiment. In more detail, FIG. 4 shows a photograph of the detector housing. The PID has been replaced with a glass window allowing an internal region of the cell to be seen.

[0191] FIG. 5A shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a detector and a method according to an exemplary embodiment. In more detail, FIG. 5A is an illustration of the enhancement effect, as described above with respect to the first mode of operation. VOCs deposit from the gas flow onto the porous media during the charge phase when the temperature is low. During the measure phase the temperature is increased, releasing the VOCs and causing a peak in the VOC concentration (i.e. the first characteristic response), during the measure phase. This extra concentration is gradually removed by the gas flow.

[0192] FIG. 5B shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a detector and a method according to an exemplary embodiment. In more detail, FIG. 5B is an illustration of the sorption effect, as described above with respect to the second mode of operation. Particularly, the first characteristic response is measured during the measure phase, during which the temperature is ramped linearly downwards relatively slowly. During the clear phase, the temperature is heated relatively quickly and a relatively small positive response is due to desorption of the relatively small amount of the first analyte sorbed in and/or on the first sorbent.

[0193] FIG. 5C shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a detector and a method according to an exemplary embodiment. FIG. 5C includes both the second mode of operation and the first mode of operation, successively, as described with respect to FIGS. 5B and 5A, respectively. A hold period at a relatively low temperature after the second mode of operation and before the first mode of operation results in sorption of a relatively large amount of the first analyte sorbed in and/or on the first sorbent, such that a relatively large positive response is observed during the second measure phase.

Analysis of Measurements

[0194] The standard method of using the PID-AH VOC sensor is to determine in advance the output voltage without any VOC present and compare this to the output voltage in the presence of the VOC. The difference in voltage is linearly related to the concentration of the VOC, though the constant of proportionality varies between different VOCs. A more sensitive measurement (lower detection limit) may be achieved by taking multiple readings to more accurately determine the output voltage, thus the longer the measurement period the more sensitive the overall measurement. This depends on the zero-VOC voltage being stable and known to a sufficient accuracy. In practice, this is not the case, so for the purposes of this work the zero-VOC voltage was determined each time a traditional measurement was taken. This is possible because a VOC-free carrier gas is available. If such a carrier gas is not available, a different approach is required, as described below.

[0195] Two algorithms for analysing the response are presented.

[0196] The first algorithm, as shown in FIG. 6, involves measuring the height of the peak which occurs when the temperature is increased. While conceptually simple and easy to implement, the transient nature of the peak makes it hard to measure at the highest value and means that multiple measurements may not be taken to reduce noise. The measurement may be taken as the difference between the value at the lower temperature (i.e. T.sub.0) to the value at the peak at the higher temperature (i.e. T.sub.1) or from the peak to the subsequent equilibrium value. The latter would avoid temperature dependant shifts in the PID output voltage, while the former should give more significant response (the VOC concentration at the lower temperature should be slightly suppressed due to VOCs depositing on the porous media, so the range of concentration should be higher). The latter would also give a more immediate response. In this work, the former is the definition of peak height used.

[0197] FIG. 6 shows a graph of first response as a function of time for a detector and a method according to an exemplary embodiment. In more detail, FIG. 6 shows an example of a VOC peak at the measure and check phases. The charge phase ends at time=0 s when the temperature is increased leading to a peak in the VOC concentration and the PID voltage. The temperature is then held constant until the check phase at time=300 s, when it is increased again. The double-headed arrow shows the definition of the peak height.

[0198] The second algorithm, as shown in FIG. 7, is to take the area under the PID voltage curve due to the evaporating VOCs. The area should be more closely related to the quantity of VOC evaporated than the peak height because it does not depend on all of the VOC being released at the same point in time. It also allows measurement over the duration of the peak which reduces the measurement noise compared to the single point in the height. It is better to use a linear fit for the baseline than a constant as there is a slow response to the temperature step which can cause an error at the low concentrations.

[0199] FIG. 7 shows a graph of first response as a function of time for a detector and a method according to an exemplary embodiment. In more detail, FIG. 7 shows measurement of VOC peak area at the measure phase. A baseline is fitted to the second half of the period between two increases in temperature at time=0 s and time=300 s. The baseline is extrapolated under the peak at time=0 s. The area between this baseline and the peak is the peak area.

Temperature Effects

[0200] The PID voltage has a dependence on temperature (FIG. 8 and FIG. 9). As the temperature increases, the PID voltage increases with approximately a quartic relationship with respect to temperature (FIG. 10).

[0201] There appear to be two time-constants involved in this relationship. First is a fast response with a time constant of the order of a few seconds. This can be seen in the mirroring of the overshoot and the oscillations which occur in the temperature in the PID voltage. The second is a slow response, of the order 5 minutes, where the PID voltage gradually shifts to a new equilibrium value following a step change in temperature.

[0202] The cause and the coupling mechanism are not known yet.

[0203] The temperature control is limited to about 80° C. at the top, so to ensure that the highest temperature was repeatable a maximum temperature of 70° C. was used. At temperatures above about 50° C. oscillations in the temperature with a time constant of a few seconds and an amplitude of about 1° C. were observed. To avoid seeing these in the measurement phase a measurement temperature of 40° C. was used. 40° C. is also lower than the bulk of the temperature dependent change in PID output voltage. The lower temperature of 5° C. was chosen as being low enough to get good deposition of the VOCs while being comfortably within the achievable range.

[0204] FIG. 8 shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a first sorbent for a detector and a method according to an exemplary embodiment. In more detail, FIG. 8 shows a time-series showing porous silica temperature and PID voltage under a constant purge flow.

[0205] FIG. 9 shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a first sorbent for a detector and a method according to an exemplary embodiment. In more detail, FIG. 9 shows an excerpt from FIG. 8 showing the PID voltage mirroring the overshoot and oscillations in the porous silica temperature. The slow time-constant response of a gradual shift to a new equilibrium voltage can also be seen, particularly at higher temperatures.

[0206] FIG. 10 shows a graph of first response as a function of temperature fora sensor fora detector and a method according to an exemplary embodiment. In more detail, FIG. 10 shows steady-state PID voltage V and temperature T under a constant purge flow. The fitted line is a quartic of the form V=aT.sup.4+b.

Results

Concentration of VOC

[0207] The effect of concentration on the system was measured for two different VOCs (isopropyl alcohol (IPA) and limonene) at 3 concentrations each. For each VOC and concentration, the steady state PID voltage at a fixed temperature was measured followed by temperature cycling of the porous media to measure the dynamic response. Because the zero-VOC voltage of the PID-AH was found to vary these measurements were also performed on the VOC-free carrier gas to establish the baseline performance.

[0208] Details of the measurement steps, as shown in FIG. 11, are: [0209] 1. Change to a new gas mixture. The sensor is in the purge gas (99.99% N.sub.2), so this gives time (15 minutes) for the new gas mixture to settle down. [0210] 2. Steady state (purge). Measure the PID voltage under the purge gas (99.99% N.sub.2) at 40° C. 10-minute settling time followed by 5 minutes of measurements. [0211] 3. Steady state (sample). Measure the PID voltage under the sample gas at 40° C. 5-minute settling time followed by 5 minutes of measurements. The first analyte is carried in 99.99% N.sub.2. [0212] 4. Clear. 10 minutes at 70° C. to get clear any residual VOCs from the porous silica, using the first analyte carried in 99.99% N.sub.2. The aim of this step is to provide a consistent starting point for each test. [0213] 5. Charge. 10 minutes at 5° C. to absorb VOCs from the gas, using the first analyte carried in 99.99% N.sub.2. The PID voltage will drop during this step for two reasons—first because the temperature is lower, and second because the porous silica is absorbing VOCs from the gas flow and hence lowering the VOC concentration in the chamber. Potentially, given long enough, the latter effect might decrease as the porous silica becomes saturated and stops absorbing further VOCs. [0214] 6. Measure. 5 minutes at 40° C. to see the response of the sensor, using the first analyte carried in 99.99% N.sub.2. The porous silica releases the VOCs absorbed at the lower temperature leading to an increase in the VOC concentration in the chamber. This effect causes a peak in the PID voltage above the steady state level. The flow through the chamber gradually clears the excess concentration and the PID voltage returns to the steady state value. [0215] 7. Check. 70° C. for 5 minutes to see if any VOCs remained in the porous silica, using the first analyte carried in 99.99% N.sub.2. Any remaining absorbed VOCs should be released at this point, so a pattern of a peak and relapse in the PID voltage (as seen in step 6) would indicate that the release of VOCs in step 6 was incomplete. [0216] 8. Repeat of 4-7 under purge gas (99.99% N.sub.2) to check the background response of the system.

[0217] FIG. 11 shows a graph of temperature as a function of time and corresponding graphs of concentration of the first analyte, first gas and first response as a function of time for a detector and a method according to an exemplary embodiment. In more detail, FIG. 11 shows system parameters and PID voltage illustrating the steps used to enhance the response of a PID with a porous media.

[0218] The steady-state PID voltage is shown in FIG. 12 for IPA at 0.0033 ppm. This concentration can be detected directly by the PID, though the measurement is approaching the quantisation of the ADC used to read the PID. Note that the difference between the sample reading and the VOC-free purge reading is less than the day-to-day variability in the VOC-free purge reading, so without the capability to compare to a purge gas this concentration would be hard to detect.

[0219] FIG. 12 shows a graph of first response as a function of time in absence of a first analyte and during exposure of the first analyte. In more detail, FIG. 12 shows steady-state PID voltage for IPA at 0.0033 ppm. The difference in voltage is only a few quantisation steps of the ADC.

[0220] The transient voltage curves during the measure and check phases is shown in FIG. 13 for the same gas. Again, the response in the VOC-free purge gas is also shown. At the start of the measure phase (time=0 s) there is a step change in temperature from 5° C. to 40° C. and a corresponding peak in the PID voltage. This peak is higher (0.012 V) than the steady state change in voltage (0.002 V) and occurring at a known time is easier to detect. There is also a peak in the purge gas (0.005 V), probably due to residual VOCs not cleared from the porous silica in the clear phase. This gives a contribution to the peak height due to the VOCs of 0.008 V, which is 4 times the steady state response.

[0221] FIG. 13 shows a graph of first response as a function of time for a detector and a method according to an exemplary embodiment. In more detail, FIG. 11 shows the transient PID voltage for the measure (time=0 s) and check (time=300 s) phases for IPA at 0.033 ppm.

[0222] The baselines for calculating the peak area are also shown as dotted lines. Of note here are the slow time-constant effects which mean that the voltage is not in a steady state during any of these phases. This causes the baseline to be sloped rather than constant and reduces the error in measuring the peak area. This is typical of measurements at very low concentrations, while at higher concentrations the peak is much larger and these details are lost. The peak area here is 0.147 with the VOC, while the purge peak area is 0.042 (the units are proportional to volt-seconds with some scale factors associated with the measurement times), so a difference of 0.105, which is approximately 50 times the steady state response.

[0223] The signal to noise ratio of these measurements may depend, at least in part, on how the detector is used. The noise level for the measurement of the steady state voltage will be proportional to √{square root over (1/t)} where t is the measurement duration. The noise level for the peak height will be constant as this depends on just the one measurement point. The noise level for the peak area will depend on √{square root over (1/t_s)} where t_s is the duration for which the peak exists. While this scaling initially seems to favour the steady state response, both peak measurements benefit from the increased signal magnitude compared to the steady state change. If the measurement duration for the steady state measurement is the same order of magnitude as the peak duration, then the noise levels for the peak area and the steady state are similar and the sensitivity or detection limit for these two methods can be taken approximately as the ratio of their signal strengths.

[0224] All three concentrations of IPA and a purge sample are shown in FIG. 14. It is clear, that the peak at the measure phase grows with increasing concentration, and that it grows faster than the steady state voltage. It can also be seen that not all the VOC is being evaporated during the measure phase, and at the higher concentrations there is a noticeable peak developing at the check phase (at time=300 s) when the temperature is increased to 70° C.

[0225] FIG. 14 shows a graph of first response as a function of time for three different concentrations of a first analyte for a detector and a method according to an exemplary embodiment. In more detail, FIG. 14 shows the transient PID voltage curves for IPA at three concentrations and under purge. The voltage offset at t=0 s is removed from all traces so that the peaks may be compared.

[0226] The transient PID voltage curves for limonene are shown in FIG. 15. The ratio between the three concentrations is the same as between the concentrations of IPA, though the absolute levels are approximately 4 times lower. The same pattern of increasing peak height and area with concentration is present, and of a peak developing in the check phase at higher concentrations. The peak has a longer tail for limonene than for IPA, and at the lowest concentration (0.8 ppb) this makes fitting the baseline difficult.

[0227] FIG. 15 shows a graph of first response as a function of time for three different concentrations of a first analyte for a detector and a method according to an exemplary embodiment. In more detail, FIG. 15 shows the transient PID voltage curves for Limonene at three concentrations and under purge. The voltage offset at t=0 s is removed from all traces so that the peaks may be compared.

[0228] The gain for peak area and peak height compared to the steady state voltages are shown in FIG. 16 and FIG. 17 for two series of tests with IPA and one with limonene. Both measurements show the same trends—a high gain at the higher concentrations of IPA falling off at the lowest IPA concentration, and a gain reducing with reducing concentrations of limonene. While the gain is not constant as a function of concentration, calibration curves, optionally for each VOC of interest, may be generated for quantitative measurements. That the gain decreases at the lower concentrations means that the effectiveness in improving the detection threshold may be relatively limited.

[0229] FIG. 16 shows a graph of sensitivity increase for a detector and a method according to an exemplary embodiment. In more detail, FIG. 16 shows peak area gain as a function of concentration of the first analyte.

[0230] FIG. 17 shows a graph of sensitivity increase for a detector and a method according to an exemplary embodiment. In more detail, FIG. 17 shows peak height gain as a function of concentration of the first analyte.

Dwell Time

[0231] The effect of the dwell time during the charge phase was studied using IPA at a relatively high concentration (0.16 ppm). For each dwell time there was a clear phase (5 minutes), charge phase (for the required duration), measure phase (5 minutes) and check phase (5 minutes), as shown in FIG. 18.

[0232] FIG. 18 shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for a detector and a method according to an exemplary embodiment. In more detail, FIG. 18 shows Temperature profile and PID response for the dwell time tests.

[0233] The transient PID voltage curves are shown in FIG. 19 for a range of dwell times. Both the peak height and the peak area can be seen to increase with longer dwell times. Interestingly, the peak for the check phase does not seem to vary in size with dwell time, suggesting that this may be more complicated than residual VOCs being driven off at the higher temperature.

[0234] FIG. 19 shows a graph of first response as a function of time for a detector and a method according to an exemplary embodiment. In more detail, FIG. 19 shows the transient PID voltage curves for different dwell times. The start of the measure phase is at t=0 s, the start of the check phase is at t=300 s.

[0235] Plotting the peak area as a function of dwell time shows that the measure phase peak does saturate as expected, with the knee in its response being at about 900 s (FIG. 20). This would therefore be a good dwell time to use for getting the maximum sensitivity in the shortest time. It is likely that the saturation time will depend on the concentration so that lower concentrations will benefit from a longer dwell time. In this case the dwell time should be thought of as setting the measurement range.

[0236] FIG. 20 shows a graph of first response as a function of dwell time for a detector and a method according to an exemplary embodiment. In more detail, FIG. 20 shows peak area plotted against dwell time.

Temperature Ramps (VOC Classification)

[0237] Generally, in conventional detectors, the temperature is changed as quickly as possible, for example increased at 150° C./min or more, to achieve the largest possible peak response. In this work, the temperature going from the charge to the measure phase was ramped at a very much lower rate of just 10° C./min to allow the profile of the VOC evaporation to be measured. In these experiments a single cycle of clear (5 minutes), charge (10 minutes), measure and check (5 minutes) was performed. The measure cycle was extended by ramping the temperature from 5° C. to 40° C. at 10° C./min before holding for 5 minutes.

[0238] FIG. 21 shows a graph of first response as a function of time for N.sub.2, IPA, limonene and benzene for a detector and a method according to an exemplary embodiment. In more detail, FIG. 21 shows the temperature profiles under purge (i.e. N.sub.2), 0.16 ppm IPA, 0.039 ppm limonene and 1 ppm benzene. The dotted line shows temperature, the solid line shows PID voltage. The temperature is ramped from 5° C. to 40° C. in 3.5 minutes starting at time=0 s. At time=510 s, the temperature is stepped to 70° C. to see how much VOC remains on the sample. The voltage offset at t=0 s is removed from all traces. It is clear that IPA, limonene and benzene have very different traces compared to the purge gas. There are also some notable differences between IPA, limonene and benzene, in particular the initial peak in voltage at the start of the ramp for benzene and the sharp peak in voltage towards the end of the ramp for IPA. The PID responses have a large broad peak and smaller sharper peaks. The sharper peaks are due to the unstable temperature control (as indicated by the dashed line). Nevertheless, the results indicate that it should be possible to differentiate between these three VOCs from the shape of their response. That is, the respective first responses of IPA, limonene and benzene are characteristic first responses, respectively.

[0239] FIG. 22 shows a graph of temperature as a function of time and a corresponding graph of first response as a function of time for IPA at four different concentrations (0.03996 ppm, 0.05327 ppm, 0.02991 ppm and 0.1599 ppm) for a detector and a method according to an exemplary embodiment. The temperature is ramped from 5° C. to 40° C. in 3.5 minutes starting at time=0 s. At time=510 s, the temperature is stepped to 70° C. to see how much IPA remains on the sample. The voltage offset at t=0 s is removed from all traces. The shapes of the respective responses are similar (i.e. characteristic of IPA).

[0240] FIG. 23 shows a graph of first response as a function of concentration for IPA, for the peak areas of FIG. 22. The relationship between peak area and concentration is linear, thereby allowing quantitative determination of the concentration for a given peak area.

[0241] FIG. 26 shows a graph of equivalent concentration (ppb) as a function of time (s) for IPA, MEK, benzene, toluene, limonene and o-Xylene. Particularly, FIG. 26 shows a graph of PID output voltage signal converted to concentration as a function of time. During desorption, the temperature was increased gradually from 5° C. to 70° C. Particularly, sorbing of the analytes was at 5° C. for 300 s while desorption was by heating to 70° C. over 150 s. The flow rate of gas was kept constant at 50 mL/min. The different analytes desorb at different times and hence temperatures (i.e. the first response comprises and/or is a first characteristic response of a set of characteristic responses of the first analyte).

[0242] FIG. 27 shows a graph of peak time (s) as a function of boiling point (° C.) for IPA, MEK, benzene, toluene, limonene and o-Xylene. Particularly, the four non-polar analytes (benzene, toluene, o-xylene and limonene) have similar affinities for the surface (i.e. the sorbent), and their desorption times and hence temperatures correlate, for example directly, linearly, with their respective boiling points. Methyl ethyl ketone (MEK) and isopropyl alcohol (IPA) more strongly interact with the surface, so desorb at a higher temperatures i.e. at later times.

[0243] FIG. 28A shows a graph of temperature as a function of time; and FIG. 28B shows a corresponding graph of first response as a function of time for a gaseous mixture of octene and IPA. Here the temperature rate of 0.5 C/sec raising temperature from 5 to 70 C over 130 seconds. Particularly, FIG. 28B shows the PID response as a function of time after the detector has been exposed to a gaseous mixture of IPA and octene. The two peaks at lower and higher time correspond to octene and IPA, respectively. Three replicates are shown, demonstrating the reproducibility of the detector.

Experimental Summary

[0244] This technique has the ability to increase the response of a broadband VOC sensor by a factor up to 70 for the tested VOCs. While this does not translate to a 70 times improvement in the detection limit, since the gain appears to drop off at lower concentrations, even at very low concentrations, a gain of 20 times was observed. It is likely that the low concentration performance could be improved by a more detailed algorithm. Using the peak area rather than the peak height yields a factor 10 increase in the response, and use of a linear rather than a constant baseline in the peak area baseline improves the low concentration performance noticeably. This technique also removes the requirement for long-term stability of the sensor as the zero-VOC voltage is not required. Different VOCs have also been shown to have different temperature profiles for desorption. This enables classification (i.e. identification) of the analytes based on their respective profiles. It also highlights an issue (also seen during the concentration testing of limonene) that the choice of temperatures is not entirely arbitrary. To get the best performance out of the system the temperatures should be chosen to achieve as full desorption as possible.

Alternatives

[0245] FIG. 24 schematically depicts alternative arrangements of the detector, according to exemplary embodiments. The same schematic representations are used to represent the same features. FIG. 24A shows a first sorbent 110A and a sensor 130 arranged in a housing having one inlet (i.e. open end) only, thereby restricting flow of the first gas therein such that diffusional transport of the first analyte in the first gas dominates movement thereof, rather than flow of the first gas. FIG. 24B shows a first sorbent 110A and a sensor 130 having an open housing. FIG. 24C shows a first sorbent 110A and a sensor 130 closely bound (i.e. coupled) or fabricated on the same substrate. FIG. 24D shows a first sorbent 110A and a sensor 130 arranged in a semi-open housing oriented, in use, to promote convectional flow of the first gas. Note that if orientated correctly, when actively cooled air flow will draw sample from the environment to the first sorbent 110A, while when actively heated, the air flow will carry concentrated sample from the first sorbent 110A to the sensor 130. FIG. 24E shows a first sorbent 110A and a sensor 130 arranged in a housing 160 having a closeable inlet 162. This is a relatively simple arrangement which extends the high concentration measurement duration, since the inlet may be closed during measurement (i.e. sensing).

[0246] FIG. 25 schematically depicts alternative arrangements of the detector, according to exemplary embodiments. The same schematic representations are used to represent the same features, as described with respect to FIG. 24.

[0247] FIG. 25A shows a first sorbent 110A and a sensor 130 arranged in a housing 160 having an inlet and an outlet, in fluid communication via a single chamber. The first sorbent 110A and the sensor 130 are in fluid communication with the chamber. Active flow of the first gas is through the chamber. Measurement can benefit from a non-constant flow rate if flow is controlled. FIG. 25B shows a first sorbent 110A and a sensor 130 arranged in a housing 160 having an inlet and an outlet, in fluid communication serially via a first chamber and a second chamber. The first sorbent 110A and the sensor 130 are arranged in the first chamber and in the second chamber, respectively. FIG. 25C shows a first sorbent 110A and a sensor 130 arranged in a housing 160 having an inlet and an outlet, in fluid communication via a single chamber. The first sorbent 110A and the sensor 130 are in fluid communication with the chamber. An inlet valve and an outlet valve are provided for the inlet and the outlet, respectively. In this way, repeat analysis may be performed and/or under different conditions.

[0248] Although a preferred embodiment has been shown and described, it will be appreciated by those skilled in the art that various changes and modifications might be made without departing from the scope of the invention, as defined in the appended claims and as described above.

[0249] In summary, the invention provides a detector for, and a method of, detecting analytes in gases. The detector comprises a sorbent for sorbing therein and/or thereon and/or desorbing therefrom, an analyte included in a gas exposed thereto, at a zeroth temperature, pressure (T.sub.0,P.sub.0), a controller arranged to change the zeroth temperature, pressure (T.sub.0,P.sub.0) to a first temperature, pressure (T.sub.1,P.sub.1) according to a first equation, to desorb and/or sorb at least some of the analyte; and a sensor arranged to sense at least some of the analyte and to output a response corresponding to the sensed analyte. The response comprises and/or is a characteristic response of the analyte.

[0250] Attention is directed to all papers and documents which are filed concurrently with or previous to this specification in connection with this application and which are open to public inspection with this specification, and the contents of all such papers and documents are incorporated herein by reference.

[0251] All of the features disclosed in this specification (including any accompanying claims and drawings), and/or all of the steps of any method or process so disclosed, may be combined in any combination, except combinations where at most some of such features and/or steps are mutually exclusive.

[0252] Each feature disclosed in this specification (including any accompanying claims, and drawings) may 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.

[0253] The invention is not restricted to the details of the foregoing embodiment(s). The invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including any accompanying claims and drawings), or to any novel one, or any novel combination, of the steps of any method or process so disclosed.