DE4342624C1 - Ozone-contg. gas mixt. prepn. appts. - Google Patents

Abstract

An appts. for producing a defined ozone/impurity gas mixt. includes a vacuum system (1) and a closed container (2) with an associated temp. sensor (22), the vacuum system having (a) a vacuum pump (10) connected via a vacuum valve (121) to a first cold trap (13) which connects with a main pipeline (19) via a vacuum valve (122); (b) a first bypass line (1910) which branches between the pump (10) and the vacuum valve (121) and reconnects with the main pipeline (19) via a further vacuum valve (123); (c) pressure recorders (111,112), a vacuum valve (124)-controlled outlet (18) and a vacuum valve (128)-controlled impurity gas inlet (17) connected to the main pipeline (19); (d) an ozone inlet (16) connected via a vacuum valve (125) to the inlet of a second cold trap (14), which is filled with an ozone-storing medium and which has its outlet connected via a vacuum valve (127) to the main pipeline (19); (e) a second bypass line (1914) connecting the ozone inlet (16) directly via a vacuum valve (126) to the main pipeline (19); (f) a third cold trap (15) for ozone condensation and purification, connected by its inlet to the main pipeline (19) via a vacuum valve (129) and by its outlet to a vacuum valve (21) of the container (2) via a vacuum valve (121); and (g) a third bypass line (1915) connecting the main pipeline (19) directly via a vacuum valve (1210) to the vacuum valve (21) associated with the container (2). Also claimed is a process for preparing a defined ozone/impurity gas mixt. in a closed container using the above appts. Pref. all the connections are in the form of polyethylene hoses.

Description

The invention relates to a device for generating a de Finished ozone-foreign gas mixture and a process for loading providing a defined mixture of ozone and foreign gas in a closed container.

A quantitative determination of the ozone content in the atmosphere re is of increasing importance. This applies to both Troposphere as well as for the stratosphere. To measure the Different devices are used that contain ozone roughly divided into two groups, namely In-situ Senso sensors that measure ozone "on site" and remote sensing sensors Ren, which the ozone content in more distant air volumes voices. Remote sensing sensors are based exclusively on spectroscopic methods, while in-situ sensors all Processes can be used in which ozone in one Sensor produces a measurement signal that changes with the content different. However, all of these devices must be calibrated.

While spectroscopic methods are mostly due to the  characteristic, spectral signature of the Spe zies are characterized by high specificity, occur in others Methods, for example chemical in-situ methods, ver increases cross-sensitivity to other species that are sometimes difficult and time-consuming to determine. Further is Especially with in-situ sensors, the measurement signal often also in com plexer a priori unknown way of external influences like Pressure, temperature and extraneous gas, depending. Therefore, there should be the gas mixture used for calibration to avoid sy structural errors in composition and with regard to physi calic conditions to the mixture to be measured.

Calibrating all sensors is problematic because of it So far there is no primary ozone standard and even Knowledge of the ozone content of a mixture through the local Zu composition of the ozone at the location of the sensor systematic error can occur. Under primary ozone standard here is na to understand pure ozone, the content of which from measurements of Sizes are determined, which with the ozone content in one sufficiently well-known, theoretical connection. Pressure and temperature measurements are in the case of stable molecules usually sufficient to create a primary calibration standard dards, but so far with ozone due to its fragility insufficient.

For the calibration of remote sensing sensors and on spectro In-situ sensors based on scospic methods can include: U. to be dispensed with a primary standard. In this case however, the spectroscopic parameters, absorption coefficients or line weights as secondary standards whose determination is a primary one in most cases Standard assumes. When using a secondary standard is however one because of the necessary photometric accuracy elaborate sensor requirement.

To determine the ozone content of a mixture of almost pure There are several methods for ozone:  

1. Simple absolute pressure measurements

Neglecting the decomposition, the particles can number density of the ozone by simple absolute pressure measurements votes (K. Yoshino, J.R. Esmond, D.E. Freemann and W.H. Par kinson, J. Geophys. Res., 98, pp. 5205 to 5211 (1993); L.T. Mo lina and M.J. Molina, J. Geophys. Res., 91, pp. 14501 to 14508 (1986); M. Griggs, J. Chem. Phys., 49, pp. 857 to 859 (1968); A.G. Hearn, Proc. Phys. Soc., 78, pp. 932 to 940 (1961)). The This procedure corresponds to the usual evaporation procedures, the determination of the non-decomposable compounds Particle density from a single absolute pressure measurement lauben (J.Nami´snik, J. of Chromatography, 300, pp. 79 to 108 (1984)). There is also a combined mass method spectroscopic determination and absolute pressure measurement (K. Mauersberger, D. Hanson, J. Barnes and J. Morton, J. Geophys. Res., 92, pp. 8480 to 8482 (1987)).

All particle number density determinations given above were used to determine UV absorption cross sections det. The validity is based on different ozone pressures of the Lambert-Beer law. A comparison of the UV absorption cross sections determined from the 253.7 nm Hg line shows a match for the pre suggested solutions within 2%. The agreement of the Er results, however, is due to diverse, possible sy thematic sources of error using two principles no similar validation. Also valid The Lambert-Beer law becomes systematic error due to ozone decomposition proportional to the ozone pressure detected. In particular, the possible decomposition of ozone during Flow through a corresponding vacuum system and due to systematic errors associated with this have not yet been used considered. The UV absorption cross sections at 253.7 nm who currently used as a non-validated secondary standard. Although these values were only determined for pure ozone, it also applied to gas mixtures. However, this presupposes  that the UV absorption cross sections within the specified Errors are independent of the total pressure. Measurements in UV-Be can get rich due to possible scattered radiation and a Zer Systematic setting of ozone by UV radiation table falsified.

2. Titration

This takes advantage of the fact that ozone is stoichiometric, i. H. in de financed shares, is implemented chemically. The source material Fe or end products are stable compounds, their content is measurable with standard methods.

• a) Iodometric determination of ozone:
  The mixture to be examined is passed through a potassium iodide solution (G. Bergshoeff, RW Lanting, JMG Prop and HFR Reynders, Anal.Chem. 52, pp. 541 to 546 (1980))
• b) Gas phase titration:
  The mixture is reacted with nitrogen monoxide, and the decrease in the NO content is measured (WB DeMore and M. Patapoff, Enviromental Science & Technology, 10, pp. 897 to 899 (1976)).

The titration methods have various systematic errors sources. On the one hand, it is a prerequisite that the ozone at Transport from sampling to the place of chemical conversion not decomposed and no further chemical reactions falsify the titration result. Corresponding systemati It is almost impossible to characterize errors.

3. Combination of pressure measurement and mass spectroscopic Methods (K. Mauersberger, D. Hanson and J. Morton, Geophysical Research Letters, 12, pp. 89 to 92 (1985))

Here, the oxygen content formed by decomposition of the ozone with the help of mass spectroscopic measurements averages. With this, vapor pressures of ozone as a function of  Temperature determined and thus apparently an ozone standard create. The maximum error is given as 1.5%. Here is the systematic falsification of the oxygen mood through the decomposition of ozone when flowing through the Connection between the ozone reservoir and the calibration volu men, as well as the calibration volume and the mass spectrometer not considered. Measurements by the applicant according to zer ozone settles especially when flowing through narrow connecting areas le with unfavorable volume / surface ratio.

As an operational procedure for the calibration of ozone sensors the iodometric method (K. Pötzl, pure dust. Air, 45, pp. 181-183 (1985)) and the measurement of UV absorption (K.H. Becker, A. Heinrichs and U. Schurath, Clean Dust. Air, 35, pp. 326 to 329 (1975)), including VDI guidelines (VDI 2468 sheet 1 05.78 and 2 10.78) exist.

Because the systematic sources of error of all known methods are poorly characterized, the comparison is the one only possibility of review. The comparison between gas phase titration and UV absorption cross sections at 253.7 nm (W.B. DeMore and M. Patapoff, Enviromental Science & Technology, 10, pp. 897 to 899 (1976)) brought in synthetic air (Ozone content <1 ppm) a match within ± 5%. A further study showed a deviation of 1% (± 3%). (H. J. Van de Wiel, H. F.R. Reÿnders, R. Seifert, R.W. Lanting and W. Rudolf. Atmospheric Environment, 13, pp. 555 to 557 (1979)). The iodo metric determination deviates between -6% depending on the method (± 3%) and + 15% (± 3%) from the gas phase titration. Will the based on the specified error, the correspondence is of the different methods 4 to 18%. Change the deviations drastically when moist air is used. The before standing description of the various methods of ozone mood clearly shows that so far there is no primary potash brierstandard there.

The object of the invention is therefore an improved Vorrich device and a method for generating or providing to create a defined mixture of ozone and foreign gas.

This task is accomplished by a device with the features of claim 1 and by a method with the features of Claim 4 solved. Advantageous further developments are Subject of the subclaims.

With this device and the mentioned method can be a defined gas mixture of ozone and foreign gases, as with for example air, in a closed container. An essential feature of the device is that the container for a defined gas mixture metal surfaces avoided and due to the fact that with a simple Shape, for example a spherical, cylindrical or cuboid shape, which has a favorable volume / surface ratio, only one connection for connection to a vacuum system consists of decomposing the ozone at the best possible level homogeneity of the mixture is minimized.

In the method according to the invention, the ozone content can be reduced two absolute pressures and the temperature can be determined. Off Going from pure liquid ozone, both the decomposition during evaporation and during transport of the ozone into the load container as well as the subsequent decomposition in the container be taken into account. The ozone provided with less Contamination is therefore a primary calibration standard with an accuracy of 1%. By adding a foreign gas mixture and measuring the relative line thickness or absorption A possible change in the ozone is then possible stops are quantified during or after accelerating; a defined mixture of ozone and foreign gas is ready poses.

To provide a defined ozone mixture is one previous conditioning and characterization of the system  conducive. For this purpose, evaporation / condensation cycles performed and it is the degree of decomposition Calculate ozone from two pressures using two methods net. As a result, the conditioning is carried out as well also checked and at the same time characterized the system.

With the help of the not complex device, connected with absolute pressure measurements, the ozone content of almost pure Ozone measured with an accuracy better than 1% is. These absolute pressure measurements are a direct method which has been validated by an independent process. This Validation procedure enables the determination of particles number density without calibration and pressure measurement. The beginning mentioned ozone determinations from UV absorption cross sections so like the elaborate mass spectroscopic determination are in direct methods, which in turn are based on results of not vali based pressure measurements.

Advantageous compared to the previously used pressure measurements To determine the ozone content is to record the total Decomposition of the ozone both stationary in the gas phase as well when flowing through the vacuum system. By a spectroscopy determining the relative line width of individual, spectroscopic transitions or relative absorption coefficient a possible change in the ozone content in the Container and also when and after adding a foreign gas mixture quantify.

It is therefore a primary calibration standard for ozone manufacturing bar; Furthermore, the determination of the ozone content of ozone Foreign gas mixtures feasible. This is the calibration of sensors to avoid systematic errors under measurement conditions possible, which is particularly important in the case of non-spectroscopic in-situ sensors, since these are in the opposite sentence on the selective spectroscopic methods for Queremp tenderness. When using a calibration standard dards by a device according to claim 1 and with  a method according to claim 4 was produced de, ozone sensors can be calibrated, which are at rest Measure the gas mixture if it does not locally decompose ozone. Has the calibration standard used is roughly the same together Settlement like the gas mixture to be measured, then also arise no systematic errors when the sensor breaks the ozone puts.

To improve the spectroscopic database of ozone with respect to line widths or absorption cross sections an absorption cell can also be used as a container. This improvement in turn enables greater accuracy ozone determinations by remote sensing sensors. The precise spectroscopic data are secondary their calibration standard wise sensors calibrate in continuous flow of the gas mixture to be measured are operated.

With the help of a primary standard, remote sensing Sensors or based on spectroscopic methods situ sensors can be calibrated directly. This is beneficial because devices calibrated in this way only have low demands on the pho Set tometric accuracy and therefore considerably expensive can be manufactured.

The invention based on a preferred Aus example of a device for generating a defi nated ozone-foreign gas mixture with reference to the Drawing explained.

The device comprises two main components, namely a Va kuumsystem 1 and an associated container 2nd The vacuum system 1 has a vacuum pump 10 with a final vacuum of <10 ^-3 mbar, two pressure transducers 11 ₁ and 11 ₂ with various measuring ranges, three cold traps 13 to 15 , an ozone inlet 16 , a foreign gas inlet 17 , an outlet 18 , a main pipe line 19 , a detour line 19 ₁₀, numerous pipes not specified in detail and vacuum valves 12 ₁ to 12 ₁₁. The cold traps 14 and 15 also each have a detour line 19 ₁₄ and 19 ₁₅, in which vacuum valves 12 ₆ and 12 ₁₀ are provided, so that the vacuum system 1 can be operated even with closed cold traps. The cold trap 13 is surrounded with a cooling bath 13a, containing liquid at a temperature of 77 K N₂. A cooling bath 14 a surrounding the cold trap 14 is kept at a temperature of 190 to 220 K and filled with silica gel. A cold bath 15 a under the cold trap 15 is filled with liquid N₂ with a temperature of 77 K and easy to remove.

The container 2 has a glass body 20 , which is provided when using the container as a cuvette with opposing Fen star, a vacuum valve 21 and a temperature sensor 22 .

The following advantages are realized in this exemplary embodiment of the device for generating defined ozone / foreign gas mixtures:
Metal parts and rough surfaces and substances that can be oxidized by ozone are avoided as far as possible.

Vacuum valves sealed with ground grease should not either be used.

Possibly used polyethylene hoses and the connection from the cold trap 15 to the container 2 are kept as short as possible.

The container 2 has no metal surfaces.

The container 2 has a simple shape, preferably egg ner spherical, cuboid or cylindrical shape, in which a favorable volume / surface ratio is given, only a single connection for a connection to the vacuum system. 1 The vacuum valve 21 for separating the container 2 from the vacuum system 1 is mounted as close as possible to the container 2 .

Calibrated, thermostated, capacitive pressure transducers are used as pressure transducers 11 ₁ and 11 ₂, the absolute errors of which are usually in the range of 0.3%. The materials used for the vacuum system and the container are preferably used: Duran glass, nylon Swagelok connections, PTFE-sealed te vacuum valves from Young, as well as polyethylene hoses. These materials have the conditionability required in the invention; if other materials are used, their conditionability must be checked. Since ozone decomposition occurs increasingly on metal, ground grease, rough surfaces and oxidizable materials, these substances should be avoided as far as possible.

The subsequent calculation of degrees of decomposition is based on the assumption of a purely catalytic decomposition of ozone to oxygen. When using substances that can be oxidized by ozone, the method according to the invention cannot be used. If necessary, polyethylene hoses must be kept as short as possible due to their ozone decomposition and high outgassing rate. In order to transport pure ozone in the container 2 without decomposing it, its distance from the cold trap 15 is minimized.

Furthermore, the use of the contents of the container 2 as a calibration standard presupposes an extensive homogeneity of the gas mixture. Surfaces on which ozone is increasingly decomposed, such as metal, cause inhomogeneities. The container 2 is therefore constructed so that the ozone mixture in it does not come into contact with metal surfaces. Consequently, a pressure measurement in the container 2 is avoided because of the metal surfaces present in the pressure transducers. The decomposition in the container 2 is minimized by the largest possible and thus advantageous volume / surface ratio. Furthermore, Rohran rates, cold fingers, etc. are largely avoided. The vacuum valve 21 for separating the container 2 from the vacuum system 1 is as close as possible to the container 2 . This is guaranteed with the addition of gas mixtures that no longer pipe sections, due to a laminae flow, are initially only filled with the foreign gas mixture and then ozone diffuses only slowly into these pipe sections. An exact determination of the particle number density from the pressure presupposes an exact knowledge of the temperature in the container 2 . The temperature in the container 2 is preferably measured using calibrated Pt-100 sensors.

Ozone production, storage and purification

The ozone is produced from dry oxygen in a continuous flow in a gas discharge, for example with a commercially available ozonizer, at a pressure which is somewhat higher than the external pressure. The ozone produced in this way is introduced at the entrance 16 . Here, the vacuum system is first flushed with dry oxygen with the cold trap 14 closed, the bypass line 19 ₁₄ of the cold trap 14 , in which the vacuum valve 12 ₆ is provided, and the outlet 18 are open. Thereafter, at open cold trap 14 and closed vacuum valve 12 is ₆ the ozone when flowing through the cold trap 14 at -50 ° C to -80 ° C of commercially usual silica gel, for example in bead form or in the form of granules adsorbed. Saturation is indicated by an even blue coloration of the silica gel.

The ozonizer is then separated from the inlet 16 , the remaining oxygen is largely pumped out and the cold trap 14 is separated from the vacuum system 1 . Before removing ozone, it is advantageous to pump out the oxygen which has in the meantime been formed by decomposition of ozone and outgassing from silica gel. In order to keep ozone for a long time, the cold trap 14 should be cooled with liquid nitrogen.

To provide the defined ozone-foreign gas mixtures, who initially condenses small amounts of ozone from the cold trap 14 in the cold trap 15 . For this purpose, the vacuum system 1 is evacuated. The cold trap 15 is cooled to 77 K by the cold bath 15 a and the cold trap 14 is connected to the vacuum system 1 and the cold trap 15 for a certain time. The temperature of the cold trap 14 is between -50 ° C and -80 ° C. The amount of ozone which passes per unit time of the cold trap 14 to the cold trap 15 depends on the degree of saturation of the silica gel in the cold trap 14, the temperature of the cold trap 14 and the amount of remaining in the cold trap 14 oxygen from, and by tests to determine. After the desired amount of ozone has been condensed in the cold trap 15 , the oxygen is removed by pumping, the large vapor pressure difference between oxygen (about 150 mbar) and ozone (0.002 mbar) at 77 K being used. This means that almost 100% pure liquid ozone is obtained with a small amount of water due to outgassing.

Leak and outgas rate

The outgassing rate is minimized by longer pumping on the vacuum system and tank. If the vacuum system and the container have not been previously evacuated, this takes about one to two days. The leak and outgassing rates are then determined as follows. The cold trap 13 is separated from the plant, it is measured that the rise of the total pressure with time, and then the compound of the cold trap 13 is interrupted to the pump 10 and the connection to the vacuum system 1 and container 2 is Herge again. The remaining residual pressure provides the leak rate; the difference between the residual pressure and the total pressure then provides the outgassing rate.

Preconditioning

The vacuum system 1 and the container 2 are filled several times for several hours with a few millibars of ozone by, as stated above, correspondingly cleaned ozone is evaporated directly from the cold trap 15 in the vacuum system 1 and the container 2 . By repeated condensation and evaporation, all parts of the vacuum system 1 and the container 2 are repeatedly flowed through by ozone and thus preconditioned.

Conditioning and characterization

One of the steps a) to c) described below  This method is used for pre-conditioning averaging the degree of decomposition in an evaporation / condensate cycle. This is based on pure liquid ozone gene:

• a) First, purified ozone is evaporated from the cold trap 15 into the vacuum system 1 and the container 2 , for this purpose the cooling is removed and about 25 s after the evaporation of the ozone, the cold trap 15 is brought to room temperature with the aid of a water bath. The entire process takes about 1 min and must be carried out as quickly as possible to minimize the dwell time of ozone in the gas phase and the associated decomposition on the surface. Then the absolute gas pressure p _{a of} the gas mixture of ozone and of small amounts of oxygen and water is determined by means of the pressure transducers 11 ₁ or 11 ₂.
• b) Immediately after the pressure measurement by means of the pressure transducer 11 1 or 11 2, the ozone is condensed again by cooling the cold trap 15 at a temperature of 77 K. The immersion depth of the cold trap 15 in the liquid nitrogen should be minimal. The absolute pressure p₀₂ of the oxygen produced by decomposition is measured after complete condensation of the ozone, which can be recognized from the constant pressure.
• c) The oxygen is pumped out with the cold trap 15 initially closed and with the vacuum valve 12 ₁₀ open. At closing is then with the cold trap 15 until he reaches the vapor pressure of ozone, ie from 0.002 mbar, pumped further. The same state as before step a) is thus reached again and the cycle is ended.

By repeating the steps several times as quickly as possible a) to c) the system is conditioned, the status of Conditioning according to the below Equations (1) and (2) is tested. This will be the decomposition  degree z determined according to the two equations (1) and (2):

where the deleted sizes refer to the following cycle. If the degrees of decomposition determined from equations (1) and (2) match within 10% and if they are constant during successive cycles, the system is conditioned and characterized. The following should also be noted here:
When applying equation (1), u. U. the measured pressures are corrected using the previously determined outgassing rate, and if the pressure p _a decreases during the measurement due to the enrichment of outgassed water, a new ozone filling from the cold trap 14 should be used.

As a reactive gas, ozone decomposes on contact with un conditioned surfaces. The rate of decomposition however, the surfaces generally take longer when the clocks through passivation. For narrow pipe sections, which one Having a large surface area to volume ratio is condi a repeated flow of ozone is important because when standing still ozone is decomposed and by the slow Diffusion rate is supplied only a little ozone. However, after some time the conditioning will be without ozone lost contact.

To determine the initial ozone pressure in tank 2 and to check the conditioning, the quantitative knowledge of the decomposition of the ozone in the system is important. Furthermore, leakage and outgassing rates must be minimized and their numerical values must be known. Pressure changes in the overall system due to leakage and outgassing amounted to approx. 0.002 mbar / min and 0.003 mbar / min, respectively, in the measurements carried out by the applicant. If necessary, the measured pressures must be corrected using these values. The ozone pressures used in the applicant's experiments were between 4 and 10 mbar.

After several evaporation / condensation cycles, water in the ozone had accumulated after outgassing and the measurement of the pressure p _a became difficult. The pressure drops rapidly after the cold trap 15 is heated due to the adsorption of water on the walls of the cold trap 15 .

The condensation of the ozone takes some time, since the ozone must diffuse through the oxygen in the system to the cold trap 15 . The lower the oxygen pressure, the faster the condensation takes place. At residual oxygen pressures below 0.1 mbar, a time of 3 minutes to constant pressure (so that ozone was no longer present in the gas phase) was sufficient within the measurement uncertainty in the applicant's system. In order to minimize systematic errors in the pressure measurement of oxygen, the immersion depth of the cold trap 14 in the liquid nitrogen should be minimal. Calculating the particle number density from the pressure according to the general gas equation presupposes a constant temperature in the entire volume. With a low immersion depth, the volume fraction of the gas is negligible when the temperature differs from the rest of the system. No decomposition of ozone in the condensed state was observed. Otherwise the pressure should have risen steadily. The partial pressure of 0.002 mbar of the ozone at 77 K is negligible for the measurement of the oxygen pressure.

It should be noted that due to the finite vapor pressure of the ozone, ozone is also removed when pumping for a long time, and systematic errors occur. This can be avoided by first pumping out the oxygen in the system when the cold trap 15 is closed and then quickly removing the remaining oxygen after opening the cold trap 15 .

The amounts of decomposed ozone and oxygen produced by decomposition are linked by the stoichiometry in the following way: 2 O₃ ⇒ 3 O₂ ie 3 parts of oxygen are formed from two decomposed parts of ozone.
So that applies

p _a = po _3a + po _2a = po₃ (1 - z₁) + po₃ (1.5z₁) = po₃ (1 + 0.5z₁) (3)

p ′ _a = p′o _3a = p′o _2a = p′o₃ (1 - z′₁) + p′o₃ (1.5z′₁) = p′o₃ (1 + 0.5z′₁) (4)

in which
p _{a is} the pressure after thawing in step a,
po _3a = po₃ (1 - z₁) the ozone partial pressure p _a ,
po _2a = po₃ (0.5 z₁) the oxygen partial pressure of the pressure p _a ,
po₃ the pressure corresponding to the amount of liquid ozone and
z₁ (0 z₁ 1) are the degree of decomposition in step a.

The deleted sizes also apply above back to the following cycle.

From equations (3) and (4) it can be seen that the decomposition of gaseous ozone in a closed volume can be followed quantitatively on the basis of the increase in the total pressure. The determination of the degree of decomposition from this pressure increase in the entire system presupposes that after warming up the cold trap 15 (step a) pure ozone is present and ozone is only decomposed if it is in the gas phase. However, ozone can decompose as soon as it flows through the supply lines between the cold trap 15 and the vacuum system 1 and during the heating of the cold trap 15 in the gas phase if, despite complete evaporation of the ozone due to incomplete heating, no pressure reading is yet possible. Therefore, complete evaporation / condensation cycles were carried out and the degree of decomposition z between two cycles was determined according to the following equation (5):

If the vacuum system 1 and the container 2 are conditioned, the degrees of decomposition do not change in successive cycles (z₁ = z'₁) and it follows

The degree of decomposition between two cycles can also be determined calculate the oxygen pressure p₀:

The approximation is permissible since the degree of decomposition z 1 for a conditioned vacuum system and the container no larger than is one percent (1%) and the resulting systematic Degradation degree z errors compared to other sources of error is negligible.  

The degrees of decomposition indicate the average decomposition in the total volume that is formed from the vacuum system 1 and the container 2 . In areas with a small volume / surface area ratio, such as in pipelines or metal surfaces, locally higher degrees of decomposition are to be expected. Once conditioning has taken place, the degrees of decomposition calculated according to equations (6) and (7) should be the same and should not change for successive cycles. As soon as agreement is reached, the method according to equation (7) should be used to calculate the average degree of decomposition, because the method according to equation (6) is based on the difference between two similar pressures and is therefore more susceptible to systematic errors, such as temperature changes, Outputs etc. are due.

The time for a complete cycle in the applicant's vacuum system 1 and container 2 was about 5 minutes. When preconditioning was carried out or a conditioning was carried out not more than 24 hours ago, about one or two successive cycles for conditioning were required. The observed degrees of decomposition z were in the range of 0.3% <z <0.8%. The agreement of the degrees of decomposition determined according to equations (6) and (7) by both methods was better than 10%. This ensures the determination of the ozone partial pressure with the desired accuracy of 1%. The agreement of the degrees of decomposition calculated by both methods shows that the applicability of equations (6) and (7) is guaranteed and the average particle number density in the vacuum system 1 and container 2 can be determined.

Systematic sources of error

The pressure p₀ measured in step b is systematically too low because part of the system is at a lower temperature sits and also low oxygen on cold glass surfaces is readily adsorbed. Based on the match the decomposition degrees was verified that this error for the gefor accuracy was not important. Furthermore, can  be concluded that ozone without the formation of stoichiometric Amounts of surface oxygen react when, for example have an oxygen atom in the surface during the reaction of an ozone molecule is installed. In this case it would be Degree of decomposition according to equation (7) too small.

However, would be a constant proportion of oxygen in the ozone solved, none of the methods described here would sy point out static errors. However, is not an azeotrope, d. H. a mixture boiling with a constant composition, known between oxygen and ozone; is also due to the strongly different vapor pressures as well as the differ The chemical nature of the substances would not lead to an azeotrope The spectroscopic validation of the method closes an azeotrope with less than 99% ozone.

Determination of the ozone content in the container before mixing with a foreign gas mixture

The procedure described below requires an immediately preceding conditioning and characterization of the system. To determine the degree of decomposition in container 2 , proceed as follows:

• a ′) evaporating ozone and determining the total pressure, P _a (see step a) and the temperature of the container;
• b ′) closing the container;
• c ′) evacuating the vacuum system;
• d ') After a predetermined dwell time in the container 2 , the container volume is expanded into the vacuum system 1 with simultaneous condensation of the ozone in the cold trap 15 with the vacuum valve 12 ₁₀ closed and the oxygen p _{b0 is} measured after complete condensation of the ozone (constant pressure )

The degree of decomposition is then calculated according to  Equation (8):

With

where V _b denotes the volume of the container 2 and V _v the volume of the vacuum system 1 . The volume ratio q _v can be determined by expanding gases from the container 2 into the previously evacuated vacuum system 1 and by measuring the pressure before and after the expansion on the basis of Boyle-Mariotts law. The ozone partial pressure after a corresponding residence time in the container is calculated by p _a (1 - z _b ). The particle number density n of ozone is calculated according to the general equation: n = p _a (1 - z _b ) / kT, where p _a denotes the ozone partial pressure, T the temperature and k the Boltzmann constant. If necessary, a decomposition-time curve as well as the reproducibility can be determined by varying the residence time of the ozone in the container 2 . This curve can be used for directly subsequent fillings for calibration purposes to calculate the ozone content.

Since the container 2 and the vacuum system 1 differ both with respect to the surface / volume ratios and the materials, different degrees of decomposition can also be expected. However, the ozone content in container 2 should be determined. Therefore, the degree of decomposition in the container separated from the vacuum system 1 is relevant, which must be determined separately according to the method described above. By varying the time between steps b 'and d', the time course of the decomposition in the container 2 can be followed.

The determined degree of decomposition represents the upper limit of the actual decomposition, because an additional decomposition during the flow through the pipeline connection between the container 2 and the cold trap 15 cannot be ruled out during the condensation. In the worst case, in which there is only one decomposition when flowing and no decomposition when staying in container 2 , the actual degree of decomposition in container 2 would be half as large. However, if the degree of decomposition increases with increasing dwell time, a decomposition must also occur when ozone is "at rest" in the gas phase. Then the maximum possible systematic error only applies to the shortest dwell times in container 2 . Degrees of decomposition of between 0.2 and 0.7% were determined by the applicant with residence times of up to 3 minutes in different containers. Even with a maximum error in the degree of decomposition of 50%, an accuracy of the ozone pressure better than 1% is guaranteed. For a conditioned system, the degrees of decomposition for the respective ozone pressure can be reproduced at any time.

A direct falsification of the pressure measurement before the container 2 was closed by water, which gets into the gas phase due to outgassing and together with ozone in the cold trap 15 , was not observed; this was ensured by spectroscopic determination of the water content in container 2 , which in this case was a cuvette. This observation can be justified by the fact that first ozone and then water evaporate when the cold trap 15 is heated in accordance with the vapor pressures. As a result, the gaseous water first accumulates in the cold trap 15 . The rate of diffusion is too low to allow appreciable amounts of water to get into the container 2 .

Mixing with foreign gases and determining the change in ozone content after mixing

To mix with foreign gases and to determine the change in gas content after the mixing, the container 2 must be a cuvette with appropriate windows. To produce ozone mixtures with other gases, the system must be conditioned and characterized in accordance with the information given above; the degree of decomposition z _b in the cuvette must also be known at the time the gas mixture is added. The filling with the ozone and the gas mixture should follow the determination of the degree of decomposition z _b as quickly as possible. To carry out the method, a spectrometer in a spectral range suitable for ozone is required, the resolution of which is so high that the spectral signature is only slightly falsified.

The following method is used here:

• a '') First, the cuvette is filled with ozone to the desired pressure p _a according to the procedure described above and then closed. The ozone content before the addition of the gas mixture can be determined from the pressure p _a and the previously determined degree of decomposition z _b .
• b ′ ′) Then an absorption spectrum of the ozone in the cuvette measured.
• c '') Then a reservoir volume connected to the vacuum system 1 is filled with the desired gas mixture.
• d '') The gas mixture is first expanded into the vacuum system 1 and then into the cuvette. The pressure of the gas mixture must be greater than the ozone pressure.
• e ′ ′) The exact pressure measurement takes place when a minimum of the pressure is reached; the cuvette is then separated from the vacuum system 1 .
• f ′ ′) Then there are absorption spectra of the ozone in the kitchen vette in the required time intervals measure up.
• g ′ ′) α) If relative line widths from the spectra with ent speaking evaluation programs can be obtained, is the percentage change in the ozone content le change of these values given.
• β) If only relative absorption coefficients are available change, their percentage change corresponds to of the ozone content only after adding the gas mixture.

Statements about a possible change in the ozone content tes during the addition of the gas mixture can additional be obtained if

• I. the absorption coefficient of extraneous gas pressure un are dependent, or
• II. Correspond to the absorption coefficients the accuracy with the help of a "line by line" Calculation from spectroscopic parameters model can be liert or
• III. the integral over the relative absorption coefficient do not make significant contributions from third parties contains molecules and the integration limits so choose that no significant falsification integral by adding the gas mixed occurs.

In case I. the change in ozone content is proportional tional of the change in the absorption coefficient.

In case II. The change in ozone content is proportional tional of changing the part used in the model number densities for ozone.

In case III. is the change in ozone content per proportional to the change of the integral over the relati ven absorption coefficients.

To determine the change in ozone content after mixing with foreign gases is the knowledge of the decomposition behavior of the Pure ozone in the cuvette is insufficient to make the change the ozone content during and after the addition of foreign gases to estimate. When adding extraneous gas can be caused by turbulence zen and the associated better transport a more decomposition take place on the surface. The decomposition behavior in the presence of extraneous gases can drastically change distinguish that of almost pure ozone, what by under different diffusion coefficients and one caused by extraneous gas Change of surfaces as well as certain foreign gases, such as  for example with nitrogen monoxide (NO), by chemical re action can be caused.

The method for determining the Zer The degree of settlement in the cuvette cannot be added when extraneous gas is added be applied because the complete condensation of the ozone due to the slow diffusion through the foreign gas takes a long time and the time of the complete condensation from systematic sources of error, such as leakage and outgassing rate cannot be determined. Even the knows The ozone content after adding the foreign gas can be insufficient be appropriate if inhomogeneities occur in the gas mixture In this case, the suitability of the cell contents for the calibration of devices with selective measurement in question posed.

However, spectroscopic methods can be used to monitor the change in ozone content during and after the addition of foreign gases. However, inhomogeneities in the mixture can only be ruled out to a limited extent, since an averaging over the volume of the measuring beam takes place in the cuvette during the measurement. Spectral ranges with intensive spectroscopic transitions are suitable for this: far infrared (FIR) (10 to 150 cm ^-1 with pure rotation transitions), medium infrared (MIR) (580 to 800 cm ^-1 , a fundamental vibration band ν₂, 980 to 1080 cm ^-1 , Fundamental vibration bands ν₁, ν₃), UV (175 to 240 nm with electronic transition). In the far and middle infrared, if the external gas pressure is not too high, the rotation structure can be resolved so that relative line widths of individual transitions can be determined. Otherwise, relative absorption coefficients can be measured. In the far and middle infrared, the absorption coefficients are strongly dependent on the pressure of the foreign gas, while in the UV range the influence of the foreign gas is low due to the large Doppler line width and the large number of transitions.

The spectroscopic method was so by the applicant probably used in the FIR as well as in the MIR, whereby in the FIR at Ozone pressures around 8 mbar and nitrogen pressures between 20 and 40 mbar use relative line widths of pure rotation transitions were detected. In the MIR, ozone pressures of 4 to 8 mbar and Foreign gas pressures from 750 to 950 mbar change the ozone hold out after adding nitrogen, oxygen or air relative absorption coefficient determined. At the foreign gas It was observed that after reaching a pressure min mums the pressure when the foreign gas expands into the cuvette approx. 10 mbar (at a total pressure of 900 mbar). At the same time, the temperature of the thermally insulated kitchen rose vette by 0.1 K. The temperature change comes from adiabati compression and heating of the gas in the cuvette and subsequent heating of the cuvette. The gas in the Va kum system and in the reservoir cools accordingly, whereby the change in temperature due to the much larger one Quantity in the larger volume of the vacuum system and the reservoir is less. Due to the higher temperature gradient of the Kü vette and its surroundings, the heat flow is corresponding there greater. As a result, the average temperature and thus the Pressure of the gas initially. The subsequent slower tem temperature compensation with the environment leads to the subsequent Pressure increase. Due to the rapid temperature adjustment in the The cell closes and the pressure is measured in the cell Pressure minimum for no systematic error of the total pressure.

Validation of the procedures for determining the ozone content

To validate the procedures in connection with a Bestim The ozone content became line widths of pure rotation far infrared transitions at ozone pressures of 8 mbar and Nitrogen pressures measured between 20 and 40 mbar. Theoreti Line thicknesses can be based on quantum mechanics calculations and the experimentally determined, highly precise permanent electrical dipole moment of ozone is calculated who the. (K.M. Mack and J.S. Muenter, J.Chem.Phys. 66, p. 5278  (1977)). The experimental line weights were derived from spectra determined with a high-resolution Fourier transform Spectrometers were measured. The photometric accuracy of the spectrometer and the accuracy of the evaluation process were measured by measuring the line thickness of the stable molecule Carbon monoxide and by comparison with theoretical values si made. The agreement of theoretical and expe rimental line widths within 1% are validated on average probably the method for determining the degree of decomposition in Ge entire system and the cuvette after filling in ozone as also the spectroscopic method for determining the temporal Chen course of the ozone content during and after the addition of Foreign gas.

Pressure range of the ozone and the foreign gas

The process has so far been in the range from 4 to approximately 14 mbar applied; however, it can also be easily applied at pressures use between 1 and 40 mbar. At very small pressures below 1 mbar it must be noted that outgassing and leak rates then represent a larger systematic source of error and furthermore, the degree of decomposition may be greater under certain circumstances. A larger amount of ozone should therefore be assumed here and should be expanded from the cuvette several times with simultaneous spectroscopic measurement of the initial pressure can be set. Ozone levels in the bar range can be also with the help of a secondary standard (i.e. line weights, Absorption coefficient) using a spectrometer with a correspondingly high, photometric accuracy determine spectroscopically. Leave ozone pressures above 40 mbar also reach themselves, although the risk of explosion at Production of larger amounts of pure liquid ozone increases.

The process has so far been used for extraneous gas pressures in the range of Tested 20 to 40 mbar. The lower limit for the extraneous gas pressure is principally given by the ozone pressure; to higher Pressures down to 100 mbar can be easily achieved. An out  However, the pressure range can expand to 1000 mbar the clear enlargement of the pressure broadening lines wide be problematic. In the FIR- and MIR range unresolved spectra, which the determ relative line widths of individual transitions ren, so that there is possibly only relative Absorptionskoef efficient determination.

If the temperature of the cuvette differs significantly from the temperature of the vacuum system 1 , equation (7) must be modified to determine the degree of decomposition in the cuvette to take into account the change in particle number density with temperature in accordance with the general gas equation.

Claims (7)

1. Device for generating a defined mixture of ozone and foreign gas, characterized in that the main components of the device are a vacuum system ( 1 ) and a closed container ( 2 ) with a temperature sensor ( 22 ) assigned to it.
wherein in the vacuum system ( 1 )
a vacuum pump ( 10 ) is connected via a vacuum valve ( 12 ₁) to a first cold trap ( 13 ) which is connected to a main pipeline ( 19 ) via a vacuum valve ( 122 ),
between the pump (10) and the vacuum valve (12 ₁) a first bypass line, branching off (19 ₁₀) the valve (12 ₃) is connected via a further vacuum back to the main pipeline (19) is Schlos sen,
Pressure transducers ( 11 ₁, 11 ₂), an outlet ( 18 ) via a vacuum valve ( 12 ₄) and a foreign gas inlet ( 17 ) via a vacuum valve ( 12 ₈) are connected to the main pipe ( 19 ), an ozone inlet ( 16 ) a vacuum valve ( 12 ₅) is connected to the inlet of a second cold trap ( 14 ) which is filled with a means for storing ozone, and the outlet of which is connected to the main pipeline ( 19 ) via a vacuum valve ( 12 ₇),
a second detour line ( 19 ₁₄) connects the ozone inlet ( 16 ) via a vacuum valve ( 12 ₆) directly to the main pipeline ( 19 ),
the inlet of a third cold trap ( 15 ) for condensing and cleaning ozone is connected via a vacuum valve ( 12 an ) to the main pipe ( 19 ), and its outlet via a vacuum valve ( 12 ₁₁) with a vacuum valve ( 21 ) of the container ( 2 ) is connected, whereby
a third detour line ( 19 ₁₅) connects the main pipe ( 19 ) via a vacuum valve ( 12 ₁₀) directly to the container ( 2 ) assigned vacuum valve ( 21 ). 2. Apparatus according to claim 1, characterized in that largely materials with a minimal reactivity with ozone are used, and that both all ver used connections in the form of polyethylene hoses and a connecting pipeline from the cold trap ( 15 ) to the container ( 2nd ) are as short as possible. 3. Apparatus according to claim 1 or 2, characterized in that the container ( 2 ) in terms of a favorable Volu men / surface ratio is carried out in a simple manner in cuboid, Ku gel or cylinder shape and only one connection for the connection with the vacuum system ( 1 ), the vacuum valve ( 21 ) for separating the container ( 2 ) bezüg Lich the vacuum system ( 1 ) as close as possible to the container ( 2 ) is attached. 4. A method for providing a defined ozone-foreign gas mixture in a closed container, characterized in that a device according to one of claims 1 to 3 is used, wherein

• a) is based on purified, condensed ozone in the cold trap;
• b) the vacuum system is conditioned and characterized, and
• c) the ozone content is calculated from a measured pressure (p _a ) and a degree of decomposition (z _b ) in the container ( 2 ) and the temperature (T) measured with the temperature sensor ( 22 ).

5. The method according to claim 4, characterized in that for conditioning and characterization in step b)

• b₁) is evaporated (15) of pure ozone in the cold trap, fall into the cooling (15) warmed to ambient temperature and the total pressure (P _a) is measured;
• b₂) then the ozone in the cold trap ( 15 ) is completely condensed and the oxygen pressure (p₀₂) is measured;
• b₃ pumped) of oxygen initially in the closed cold trap (15) and then with an open cold trap (15) of the pressure up to the vapor pressure of ozone at 77 K is lowered,
• b₄) Cycles from steps b₁) to b₃) are carried out briefly one after the other until calculated degrees of decomposition within 10% are the same and remain the same for successive cycles.

6. The method according to claim 4 or 5, characterized in that the container ( 2 ) is filled by evaporation of condensed in the cold trap ( 15 ), pure ozone, the cold trap ( 15 ) warmed up and the total pressure (p _a ) is measured ;
the container ( 2 ) is separated from the vacuum system ( 1 ); the vacuum system ( 1 ) is evacuated;
after a certain dwell time, the container volume is expanded into the vacuum system ( 1 ) with simultaneous condensation of ozone in the cold trap ( 15 ), and
the oxygen pressure (p₀₂) taking into account the volume ratio (q _v ) from container ( 2 ) to the overall system ( 1 , 2 ) is determined and the degree of decomposition (z _b ) in the container ( 2 ) is calculated. 7. The method according to any one of claims 4 to 6, characterized in that a gas with a defined ozone content of foreign gas is added and the ozone content is determined during and after the addition of the foreign gas in that

• k) an absorption spectrum of the ozone in the container ( 2 ) formed as a cuvette is measured with a spectrometer;
• l) a reservoir volume connected to the vacuum system ( 1 ) is filled with the foreign gas mixture provided;
• m) the foreign gas is first expanded into the vacuum system ( 1 ) and then into the cuvette, the foreign gas pressure having to be greater than the ozone pressure;
• n) after the pressure measurement, the cuvette is separated from the vacuum system ( 1 );
• o) absorption spectra of the ozone in the cuvette ( 2 ) are measured at corresponding time intervals;
• p) relative line strengths or relative absorption coefficients are determined from the absorption spectra obtained in steps k) and o), and the change in the ozone content is determined therefrom.