Патент USA US2412236код для вставки
ec, w, 1946. H. w. wAsHBURN 2,412,236 MASS SPECTROMETHY Filed Deo. 9, v1943» ' 2 Sheets-Sheet l MY @6% mmRkSÄ mkDS Ä w \%, A TTOÍFNEYS Deco i0, 1946. H, W, WASHBURN ‘ 2,412,236 f MASS SPECTROMETRY INVENTOR. H/wœ n /4./ M45/@amv Patented Dec. l0, 1946 - 2,412,236 UNITED STATES PATENT OFFICE 2,412,236 \ Mass sPEcTnoMETRY Harold W. Washburn, Pasadena, Calii'., assignor to Consolidated Engineering Corporation, Pasadena, Calif., a corporation of California. Application December 9, 1943, Serial No. 513,526 20 Claims. This invention relates to gas analysis and par ticularly to quantitative analysis of gaseous mix tures by mass spectrometry. This application is a continuation in part of my co-pending application Serial No. 320,802, (Cl. 73-18) 2 simple quantitative relationship between the composition of the mixture to be analyzed and the various ions formed therefrom in the mass spectrometer. Thus I have found that, if the pressure in the sample chamber from which the filed February 26, 1940. gas mixture is admitted into the ionization cham A mass spectrometer is an apparatus employed ber is suiliciently low the various components of for producing and sorting ions. One known a gaseous mixture to be analyzed will flow into form of mass spectrometer comprises a sample the ionization chamber at mutually independent chamber. an ionization chamber, an analyzer, 10 rates, i. e. the rate of ñow of each component and a collector. A 'gas mixture to be analyzed will be in accordance with the partial pressure is introduced from the sample chamber through of the component in the mixture and independ an orifice into the ionization chamber and is ent of the partial pressures of the other compo there bombarded by electrons emitted by a fila nents. Hence the portion of the mixture which ment, so that molecules in the mixture become enters the ionization chamber will be quantita positive ions. As a result of their charge, the tively as well as qualitatively representative of positive ions are accelerated toward an exit slit the mixture. in the ionization chamber. After passing through The pressure in the sample chamber should this slit the ions are accelerated further toward be reduced so far that the mean free path of a second slit, which is kept at a large negative 20 the molecules in the region of the conduit> potential with respect to the iirst slit. Hence the through which the mixture is admitted into the positive ions pass through the second slit at high ionization chamber is large compared to the least velocity and enter the analyzer, where they arel cross sectional dimension of this'conduit. It ap subjected to the action of a magnetic ñeld that pears that optimum results are obtained if the causes them to pursue a curved path. The ra 25 pressure in the sample chamber is so low that dius of curvature of this path for a -given ac the mean free path of each type of molecule celerating voltage depends upon the ratio of the present is at least twice the least cross section charge on the ion to its mass or atomic weight, dimension of the conduit. However, irrespective which ratio is hereinafter sometimes referred to of the particular ratio between the mean free as “speciiic mass.” In consequence, the ions of 30 path of the molecules and the cross section of low mass follow a path of short radius in the the conduit through which they enter the ioniza analyzer, while those of larger mass follow a tion chamber, the fact remains that in- any in path of greater radius. stance the pressure in the sample chamber can At the exit end of the analyzer there is an be reduced to a point below which the molecules exit slit, and by proper adjustment of the mag will ñow from sample chamber to ionization netic ñeld or the accelerating voltage or both, chamber at mutually independent rates, and the radius of path for ions of a given mass can that when this condition prevails it becomes be adjusted so that these ions are directed at much simpler to determine the quantitative com the split, pass through it and strike a collector, position of the gas mixture from its spectrum. where their quantity is measured, for example 40 In short. conditions of “molecular ñow” into the by a galvanometer connected to the collecter ionization chamber can be obtained by reducing through a suitable vacuum tube amplifier. the pressure in the sample chamber and when By varying the magnetic field or the acceler molecular ilow is established quantitative an ating voltage, the diverging ion beams of dif alysis with a mass spectrometer becomes rela ferent speciñc mass formed from the gas mole 45 tively simple. In-order to obtain a uniform cules in the mixture can be brought successively mixture within the sample chamber, the diifu~ through the exit slit and discharged to „produce sion within the latter must be rapid, which is a series of ion currents which represent the mass the case when the pressure is low and the shape spectrum. If a quantitative relationship can be of the chamber is appropriate, i. e., relatively found between the molecular components of the- 50 wide in proportion to its length. mixture and the ions formed therefrom, the mass In summary, my invention contemplates :dow ing the mixture (preferably by pressure) from analysis of the mixture. the sample chamber into the ionization chamber. As the result of my investigation, I have found while maintaining the pressures in both cham-. that it is possible to establish and maintain a 55 bers. Si? 112W thateach component flows into the spectrum becomes a means for the quantitative 2,412,236 3 quantitatively and qualitatively the same as that in any other portion of the chamber. How ever, the rates of diiïusion of the components of gas sample at those temperatures and pressures ` ionization chamber at a rate dependent on the partial pressure of that component in the mix ` ture and independent of the partial pressure of any othîí component. Moreover, the net with drawal of each component from the ionization ' which I prefer to maintain in the chamber usually are such that an adequate degree of homogeneity is obtained without agitation. Another factor which aids in establishment of region by the pumping system (taking into ac count that there is some diiîusion of the com ponents in the opposite direction) should be in l dependent of the partial pressures of other com the required linear relationship of the mass spec trum of the gas mixture to the mass spectra of its pure components obtained under similar con ponents present. ».This can be accomplished (a) by having a pumping speed of the system em ployed to exhaust the ionization chamber high (as compared with the pumping speed of the ex ditions is the maintenance of a pressure in theionization chamber such that each component is haust port of the ionization chamber) and (b) by ‘ introducing a “bottle neck” between the pump and the chamber, so that a high pumping speed in the pump (as compared with the bottle neck) is obtained. In such case, each componentiiows through the ionization chamber at the rate it would have ii' it alone were present. The pressure in the ionization chamber should ionized to the same extent in the same manner that it would be ionized if it alone were present. In` other words, the- pressure in the ionization ‘ region (i, e. in the space in which the ions are - formed and through which they travel until they mlñll another requirement, in that it should be l such that the number of ions derived from an 1 individual component in the ionization step varies ‘ in accordance with the partial pressure of that enter the analyzer) should be so lowthatv the ions formed do not collide substantially with each other or with uncharged particles. vBy avoiding such 'collisions' in this region, inter change of charge between particles, secondary ionization of particles by the original ions, and ` combination of ions with other particles, are pre vented.' In other words, .the _possibility of> col component and independently of the partial pres ` sures of other components, intermolecular col l lisions being minimized by the low pressure. Again, pressure should be so low that ionization of each component proceeds independently of " the presence of other components. Molecular ñow through the conduit probably lision in the ionization chamber may be mini mized by making the mean free path> of each and every type of ion present greater than the dis tance to. be travelled by these ions from their point of formation to the point at which they enter the analyzer. ' ' ' Thirdly, in order-to obtain -the desired linear relationship between the spectrum of the'mixture and the several spectra of its individual com l arises under these conditions because the mole , cules in passing through the conduit into the ionization chamber strike the walls of the con duit to a much greater extent than they strike each other. The molecules, upon striking the walls of the conduit, rebound with a velocity which ponents, collision between ions andluncharged molecules should be avoided in the analyzer and~ this requires that the pressure prevailing in- the analyzer be so low that the mean free path of ' is controlled primarily by the character and tem- ’ ' perature of the conduit surface and is independ 40 the ions at the prevailing pressure in the analyzer is greater (and preferably much greater) than ent of the original velocity of the molecule strik ing the wall. Moreover, since the molecules do the distance travelled by the ions from the point of ionization to the point of collection. At the same time, the space charge and interior surface effects’ in the ionization chamber should be kept not collide with each other to a substantial ex ` tent, if at all, in the conduit, fast-moving mole cules have little or no opportunity to impart ecules. The result of the collision with the con- l low, and electron emission should be kept uni- ' form. dependent only‘upon the temperature and nature junction with the accompanying drawings in their velocity characteristics to slow-moving mol These and other features of my invention will duit wall and the lack of collision between the be more thoroughly understood in the light of molecules themselves is that each kind of- mole cule flows through the conduit at a rate which is 50 the following detailed description taken in con which: of the conduit surface, and upon the molecular cule. - ing a mass spectrometer, which may be operatedv In any event, and whatever be the explanation, in accordance with my invention; _ the fact remains that under the conditions spec Fig. 2 illustrates a modiñed form of the inlet iñed above. each kind of molecule (i e. each component) flows into the ionization chamber at system of the mass spectrometer of Fig. l; Fig. 3 represents graphically the time decay curve of ion currents produced by the ionization an independent rate. This phenomenon coupled with other conditions discussed hereinafter can be employed to establish a linear relation be tween the mass spectrum of -the gas mixture and the mass spectra of any of the individual com ponents obtained under similar conditions. And this linear relation greatly simpliñes quantitative of a sample vof pure gas in the mass spectrom eter of Fig. 1; and . Figs. 4, 5 and 6 represent graphically the in tensities of certain ion currents measured under A Y.'w standard conditions for CO2, iso-butane, and analysis with a mass spectrometer, especially of l gas mixtures containing components which crack under the conditions prevailing in the ionization chamber. The establishment of the required linear rela ` tionship is further aided if, during the analysis ` period, the contents of the sample chamber is agitated with a view to maintaining it homoge , neous throughout and so that the portion of the ' sample immediately adjacent the conduit is j Fig. 1 is a. schematicl diagram, partly inr cross» section, showing a gas analysis apparatus includ ` weight and partial pressure of that kind of mole normal butane, respectively. Referring to Fig. 1, an unknown gas mixture held in a sample chamber `I is admitted to‘an ionization chamber 2 through an inlet capillary tube 2' and a jet 1, and withdrawn from the ion ization chamber by evacuation through an out let port 4. Thus, the 'bore of the capillary tube acts as an inlet port or oriñce 3. . ` Gas within the ionization chamber 2 -is bom barded by electrons drawn from a helical ñla 1 2,419,930 ' 5 _ , 6 . ment type cathode l into the space within a‘grid type anode l which is maintained at a positive _' in a co-pending patent application of Harold W. Serial No. 513,527, filed. December 9, potential with respect to the cathode. Positive y Washburn, 1943,- and entitled Mass ions are formed from molecules .thus bombarded and these ions are accelerated toward a grounded collimator tube I 0 by virtue of- a high- positive potential maintained at the cathode 5 `and the spectrometry. Gas to be analyzed is gathered ’in a detach able container 30 and _the latter is attached> to the sample chamber through a conduit I4. ' Prior to introduction of a gas mixture from _the detachable container I0 to the sample chamber. Sonie oi' the accelerated ions pass through a - stop cocks 3|, 32 collimator slit. 9 and proceed through asecond 10 Il is kept_ closed, are opened while a stop cock in order to- evacuate sample collimator slit II, thereby forming a limited het chamber I andthe connecting conduit 34. When erogeneous positive ion beam which, when a key the pressure within the sample >.chamber has been K is closed, is deñected downwardly by anv elec reduced to a suitable value, say one micron, the trostatic field maintained between a pair of plates ' valve y3I"is closed and some of the unknown gas Il by' a battery I2. The stream oi' ions pro mixture is admitted into the sample chamber by ceeds through a gap 20 in a chamber Il where ' opening the valve 33 fora time. pressure said stream is bent upward bya magnetic ileld oi' the unknown gas mixture within the sample provided by an electromagnet I5. _chamber I is measured by means of a pressure Due to thefcombined eilects of the electric gauge 35, as a McLeod gauge. In case-t o and magnetic ilelds and the geometry of the 20 much gas such is admitted to chamber I a portion of.' mass spectrometer, positive ions vof a predeter said gas may be withdrawn by opening the cockmined mass-to-charge ratio vare caused to pass II for a short time interval. anode 6 by a high voltage battery 8. _ _ through a narrow. exit slit I6 and -fall upon a col- « tcßlector I'I connected in-conventional manner toa - B’rid oi' an electrometer tube I8. The intensity of the ion current fallingupon the collector Il is i) A -stop cock 40 is opened to cause gas to ilow into the ionization chamber. The rate of iiow of a pure gas through the capillary tube 2’ is given by the equation measured by a galvanometer G in the output oi' a D.-C. ampliñer‘A connected in conventional manner to the electrometer tube. - - The particulaPmass-to-charge ratio of theions which fall upon the collector I'I may be changed- by varying the current through a coil I9 which provides the magneto-motive force for =rate of now in c. c./sec., referred to a -unit establishing the magnetic ñux in the .gap_2l)pressure of one dyne/cm?, ' _ through which the ions are caused to ilow. By 35 R=radius of tube. :length of t?be. changing the magnetic neld, ions of different charge-to-mass ratios are caused to fall succes d1=density of said'puregas at a pressure yoi! one ' where: sively upon the collector. This produces a series of ion currents which can be measured and em dyne/cm?, » , > a Z=mean free path of molecules within cham- l ployed for determining the constituents of- an 40 ber I, _ unknown gas mixture admitted into the ioniza p1=pressure in chamber I, tion chamber from the sample chamber._ p2=pressure in ionization chamber 2. The space from the collimator tube II) to the l'n most practical cases to be considered here, electrometer tube I8 is maintained ata very low _ the pressure in the sample chamber I will be pressure by means of vacuum pumps` connected . large compared to the pressure in the ionization at exhaust ports 22, 23, so that the mean free chamber 2 so that, if the radius of the inlet port path of molecules in said space exceeds the 3 is small compared to the mean free path of the length of the paths traversed by the ions-in their molecules in chamber I, Equation 1, reduces to travel from the anode 9 to the collector I1. In the foregoing gas analysis procedure, the ~ pressure within the ionization chamber 2 is Mp! maintained large enough to provide ion currents of suitable intensity. Preferably the. pressure is - When the radius of the inlet port 3 is smalll low enough for the mean free path to be large compared to the mean free path ol.' the molecules, compared to the dimensions of the ionization 55\jew molecular collisions occur at that »point and chamber. A pressure suitable for this purpose hence the rate of iiow through the tube becomes lies within a range of about 10 to 40 mp.- Hg. y independent oi' the internal viscosity of the gas. The pressure within the 4ionization chamber Thus, when a gas mixture is being admitted to may be measured by means of a Knudsen gauge 24 and controlled by adjustment of a poppet 60 the ionization chamber through the inlet port 3, the flow of molecules of one type will be substanf valve .25 at the mouth of an exhaust tube 2B en tially unaffected bythe ñow of molecules of any closing the outlet port 4. A suitable Knudsen other type present. The rate of` ilow of -each gauge is described in articles by J. W. N. Du component is- governed by Equation 2 where d1 mond, and W. M. Pickles, Jr., in- the Review of and p1 are respectively the densities and partial _ Scientiñc Instruments, volume VI, page 362 pressures corresponding to the individual com (1936). ' ponents. l At the end of a poppet valve shaft .21 _oppo site the poppet valve is a soft iron armature 28 by means of winch the position of the poppet valve may be adjusted by the action of an exter . nally operated magnet (not shown). Although it is shown in a vertical plane, the valve shai't 21 preferably is horizontal. A detent 4I determines ' the position of maximum closure. The outlet system is described in more detail and claimed _ ’ In another form -of my irîvention illustrated in Fig. 2 an inlet port 3' consists of a small oriñce in a plate 4’. In` this case also each com' ponent oìfoagas mixture will flow through the inlet“ port 3’ at an independent rate if the mean free path of the molecules is large compared with the radius of the oriiice. ‘ The rate of now of pure gas through either inlet port 3 or inlet port 3' varies inversely as 2,412,286 the squareI root of the molecular >_weight of said to maintain the mean free path o! molecules within the chamber approximately equal to the' length of the chamber. I have found, however, While the equations of flow (l) and (2) given that the pressures required to maintain the mean hereinbefore are strictly applicable only tc pure gases, I have found that in general, if I main- 5~ free path suillciently large for this purpose, lare unnecessarily low and that we canmaintain mix tain the mean free path of the molecules at the tures sumciently uniform at still higher pressures. _inlet port 3 or‘l 3’ large compared to the radius The time constant which measures the period R, collisions between molecules of different during which a‘given degree of mixing occurs in ïkinds of gas near or within the inlet portare ' made so infrequent that molecules of different 10 a binary mixture is given by sas. kinds ñow through said orifice substantially un impeded by the presence of other molecules. ‘ ' It is clear that the effective radius of the `funnel-shaped flanged end of capillary tube 2' X2 ~' . f “n ' (3) where D=diiïusion coefficient; I :length of 3 is greater than the radius of the bore of the tube 15 sample chamber. Fora mixture ‘of hydrogen and oxygen (having litself’. For this reason the funnel-like end of an interdiffusion constant of 0.7 at S. T. P.) in the tube 2’ is preferably mounted, as shown,~on a sample -chamber 10 cm. in diameter at a pres the low pressure side of the orifice where the _sure of 0.10 mm. Hg, the mixing period is mean free path is largest. At a, suitable working pressure the mean free 20 102 0.10 , `lpath in ionization chamber 2 will be very large i _T“’fîîßfmo , compared with the radial thickness of the annu- . « A lar space between valve 25 and cone-shaped valve seat 29. At 10 mu Hg and 0° C., for in ' 37.24 v ` stance, the mean free path of nitrogen molecules 25 I have found that I can provide a substantial is 650 cm. At such pressures each component of a. mixture will ñow out of the exhaust port 4 1 ly uniform mixture in the sample chamber if the volume of gas admitted to the ionization cham at an independent rate inversely proportional to sa... .ber during the mixing period is sufllciently small the molecular weight of said component. Under the conditions prescribed above,v the ion4 30 compared with the volume of the sample cham ber. Thus, for example, the quantities of hydro , currents detected at the collector I1 will repre sent the sums of the currents which would be observed for the individual components if these Í were present alone, andthe measurements of the gen and oxygen ñowing through a simple orifice such as the inlet port 3’ having a diameter of l mm. during the above calculated mixing period ‘ several’ioncurrent‘s may be used to determine 35 T are 0.67 cc. and 0.16 cc. respectively. -the constitution of the original gas mixture. From the foregoing description it is clear that Since each of these quantities of gas- is very small com pared to the volume of the sample chamber, it is clear that the mixture in the sample chamber _is substantially homogeneous atv any instant dur component gas through the- ionization chamber 2 substantially independent of the presence of"40 ing the transfer of gas to the ionization chamber. Thus the portion of gas near the oriñce is sub other components. However, when extreme ac stantially typical of the gas remaining in the curacy is required, it is also desirable to provide ‘ I am able to maintain the rate of flow of each l ‘ some method for,- maintaining the gas right at sample chamber. „ ` By. so maintaining the gas in the sample cham the entrance end of the inlet'port (3 or 3') sub ’ stantially typical of the entire mixture within 45 ber substantially homogeneous, complex-correc tions that might otherwise be required due to the sample chamber. Otherwise, the mixture variations in sample concentration with time are flowing into the ionization chamber 2 will be ‘ seriously affected by the rates ofinterdiiïusion Y avoided. However, the degree of homogeneity required and hence the sample chamber pressure of the components within the sample chamber I and the anlysis of observations made correspond- 50 permissible depends on the degree of accuracy' required. ‘ ingly diiiicult. The process fof obtainingv uni' "form distributions ofthe various components is i retarded by the collisions which occur between unlike molecules. I prefer to resort to stirring the mixture me-~ chanically to maintain the mixture homogeneous when the ¿gas in sample chamber I isat too high . I maintain the mixture within sample chamber 55 >a pressure for interdiilusion to occur rapidly Isubstantially homogeneousv in either of two enough for my purpose. s . By' controlling the operating conditions of a mass spectrometer in accordance with (the prin rates within the sample chamber or (2) by stir-_-` ring the mixture mechanically. - I prefer to main-J’ " ciples hereinbefore explained each component of tain the mixture substantially uniform Ithrough- 60 a gas mixture is` caused -to ñow through the ion- , ' ways; (l) by maintaining rapid interdiffusion out thel sample chamber by maintaining the rates of interdiiîusion within said sample chamberY rapid compared to the rate at which gas is admit ' _ ted to the ionization chamber. /I achieve >this re sult by employing a sample chamber of proper G5 shape and by maintaining the pressure within the sample chamber low enough for the mole cules to distribute themselves throughout that chamber so rapidly that the mixture is main tained substantially uniform and the mixture ad- 70 jacent the mouth of the orifice is always substan 'f tially typical of the mixture present in the chamber. . _ ` One way to maintain the mixture substantial ization chamber 2 independently of the presence of other components; ions are derived from- each component within theionization chamber 2_sub stantially in direct proportion to the partial pres sure of each component; and as a result the mass spectrum for a mixture is a linear superposition of the mass spectra .of the individual components of said mixture. ' l Consider the conditions which exist during the analysis of a known pure gas such as CO2 con tained in sample chamber i. Prior to admitting the CO2 into the ionization chamber 2, the indica tion of the galvanometer G is zero. When the inlet system of the ionization chamber 2 is opened ly uniform throughout the sample chamber I, is 75 by turning the stop cock t0, the partial pressure 2,412,2se . 9 of CO2 Within the ionization chamber 2 begins to rise. Ions produced by electronic bombardment It is clear that if a mixture of any of the afore mentioned gases is admitted to the mass spec of CO2 are formed in proportion to the partial trometer under the operating conditions which pressure of CO2. After a short time interval, of the order of one or two minutes, dynamic pres- 5 I have prescribed hereinbefore, each component of the mixture will act independently of each of sure equilibrium is established between the sam the other components. Accordingly, the spec ple chamber l, the ionization chamber 2, and the trum observed for the mixture will be a super exhaustpumps~ There'after the sample cham position- of the separate spectra of the gas com- ' ber pressure decreases substantially exponen ponentsv combined in proportion to the amounts` tially andthe ion density in the ionization cham-y l0 of the respective components present in the mix ber 2 decreases in a corresponding manner. Part of the ions formed traverse the collimators Ill-Il and ions of a predetermined mass-to-charge ra tio are caused to fall on the collector I1. ture. For a mixture the intensity of a mass spec trum line formed by ions having ' a mass-to charge ratio of R is y In Fig. 3, I have illustrated graphically the 15 (4) variation of ion current with time, measured after opening the stop cock 40. The curve represents where KR; is the sensitivity of the mass spectrum the collected ion` current fora given ion such as for ions of mass-to-charge ratio R and derived CO+ having a mass-to-charge ratio of 28 formed from a unit amount of gas component i, and X1 by' bombardment of CO2. Abscissae represent 20 is the quantity of component :i present in the time, and ordinates represent the logarithm of mixture. ' _ the reading of the galvanometer G. After the Now assume that a mixture 4of ethane, propane, stop cock 40 is opened, the ion current increases and normal butane is being analyzed, and that rapidly, shortly reaching a maximum and there the partial spectrum for this mixture consists of after decreasing substantially exponentially as. indicated by the straight line portion L of the » curve. _ The time constant of the decaying ion current depends on many factors, including the volume of the sample chamber l, the dimensions of the inlet ports (3 or 3’), and the molecular Weight of the gas being analyzed! For the analysis of some mixtures containing CO2, only the CO2 ions having a mass-to-charge ratio of 28 (C12O16+), 29 (C13O16+ and 012017), 30 (C13O17+), and 44 (C12O216+) are of interest. The corresponding galvanometer deñections may be measured at convenient predetermined stand ard times of 2, 4, 6, and 12 minutes to obtain a standard mass spectrum. A spectrum for CO2 standard time galvanometer deflections Cao-:9.9, C44=14.8, 058:41, corresponding, respectively, to ions having mass-to-charge ratios of 30, 44, and 58. From Equation 4 and the table itis clear that for this case ` , (5) (6) (7) . where X1, X2 and X3 are the quantities ói.' normal butane, propane and ethane, respectively, in the sample. Solving Equations 5, 6 and 7 simulta neously, it is found that the contents of the sam ple are, respectively: obtained in this manner is shown in Fig. 4. In this graphe abscissae represent mass-to-charge ratios and ordinates represent galvanometer de iiections per microlitre at standard temperature The example just given shows that where thel and pressure of CO2 originally present in the 45 number and nature of the components of a gas sample chamber. mixture is known, the composition of the gas may Figs. 5 and 6, respectively, represent similar be determined by reading the galvanometer de standard spectra for iso-butane and normal bu iiections corresponding to a limited number of tane for mass-to-charge ratios of 28, 29, 30, 43, different ions produced by electronic bombard 44, 57 and 58. ment of the mixture. In general, the number of 50 diil’erent ion currents measured should be at least The intensities of the ion currents measured equal in number to the number of components at standard times are given more exactly in the contributing to the production of said ions. Ob table fol~ CO2, isobutane, normal butane, pro viously, if the number of observationsl exceeds the pane, and ethane. The tabulated values repre sent galvanometer deflections per /rl standard 55 number of components present the extra obser vations may be used to check the results. temperature pressure of the respective gases for In case it is not known in advance of the analy mass-to-charge ratios of 28, 29, 30, 43, 44, 57 and sis what components are ` present, the nature 58, obtained at the standard times given in co1 umn 1. of the components may be determined by a study of the complete mass. spectrum of the mixture 'rattle~ Standard time Ethanc Propane Normal butane 'e 'e ‘ge n 60 or by supplementary methods. butane C02 está? ‘ses In any case, standard spectra are determined for gas components contributing to the presence of particular ion currents measured for a mix ture, and the composition of the mixture deter , mined by comparing the mass spectrum for the mixture with the mass spectra of the compo nents. The calculations of the composition of the mixture are simplified by the method be cause of the control maintained on the rates of 70 110W. While I prefer to obtain the standard spectra An examination of the partial spectra repre for pure gas from samples of the pure gas, it is sented in the table and Figs. 4, 5 and 6, shows that clear that standard spectra for n pure gases may the spectra diifer widely and may be utilized in be obtained if desired from the spectra for n dif identifying the respective gases. 75 ferent mixtures of these pure gases. Other 2,412,236 11 modifications of the method may be made where pure gases are unavailable. The numerical example given above illustrates how my method of mass spectrometry may be utilized to determine the composition of a gas mixture. It has particular advantages when the two or more gas components present in the gas sample produce ions of the same mass-to-charge 12 ’ path within a suitable range in accordance with the principles herein set forth. _ The relative amplitudes of ther standard mass spectral lines of .any gas `as illustrated in Figs. 4, 5 and 6 are dependent on the decay rates oi the ion as well as upon the conditions of ionization. ~ For any given set of conditions, how ever, the composition of the mixture may bev de terminedby obtaining standard spectra of the ‘ My method of mass spectrometry is particu 10 components of a gas mixture together with a standard spectrum for the mixture. larly useful in the analysis of 'a gas mixture the In the actual analysis of a gas mixture certain components of which yield some ions of the same steps in addition to those already described above ' mass-to-charge ratio. And the method of analy are desirable. ,j sis is essential to mass'spectrometry when one By means of the rheostat R the current in the or more of the components yield only ions which 15 coil I9 is adjusted to a value which produces a are also produced by ionization of other compo ' magnetic ñeld which causes ions of a predeter nents possibly present. mined mass-to-charge ratio to fall on the col Not only can the method be used in the analy lector l1. f ` Y sis of a mixture of -several hydrocarbon gases To obtain accurate readings it is desirable to having different molecular weights. It may also 20 - measure the background spectrum due to residual be used to measure the concentrations of hydro gases in the ionization chamber prior to opening carbon mixtures made up of a plurality of struc the inlet system.` To do this I measure the back turally different hydrocarbons having the same ground ion currents corresponding to those ions molecular weight. For example, to measure the which I also measure from the sample. This concentrations of iso-butane and normal butane 25' background spectrum is preferably measured just in a mixture known’to contain only these two before or just after a gas sample is run. The gases, it is only necessary to measurek ion cur background spectrum- is subtracted from the rents corresponding to two of the common ions 4spectrum observed for the mixture, prior to com formed. The pair of ions having mass-to-charge puting the composition of the mixture according ratios of 57 and 58 may be used for this purpose. 30 to Equation 4. It is to be understood, of course, An examination of the table and Figs. 5 and 6 that the measurement or the background is not ywill show that other pairs of ions are also suit necessary where the background is of negligible able. magnitude. From the foregoing illustrations itis clear that When analyzing small samples of gas, such as ' the method of the invention may be utilized to 35 soil gases containing hydrocarbons or other pe obtain rapid and accurateanalysis of gas mix troleum indicators, observations of the intensities tures where conventional gas analysis methods of the ion beams Amay be made successively at are slow, tedious and inaccurate. 'different times and corrections applied' to the ob The procedure described above is also -particu servations to compensate for the loss of gas from » larly useful where the gas sample to be analyzed 40 the sample during the observation times. is very small. For this reason the method is ap In case ionization currents are measured for a > plicable to soil gas analysis for petroleum pros mixture at times other than standard times, the pecting purposes and leads to an accurate knowl readings may be corrected to standard times by edge of the minute contents of soil gases where applying to the mixture readings, correction 45 l other methods fail to separately identify the factors corresponding to the decay rates of the various gases present. gas components contributing to said ionization In the usual method» of soil gas analysis, currents. In the apparatus used such corrections groups of hydrocarbons are only roughly identi are of the order of 1 to 5% per minute. While ñed and measured. Individual hydrocarbon con this correction procedure neglects differences in ` stituents of soil gases cannot be completely sep 50 decay rates for gases of diiîerent molecular Y Y arated and identified by conventional gas analy weights (which may contribute to a given ion - sis procedures. By analyzing soil samples in ac current), nevertheless`such corrections are sufii cordance with my method, however, it is possible ciently accurate for many commercial purposes. Yto identify individual hydrocarbons present in Where extreme accuracy is desired, the spectra 55 said samples. ' . of the separate components are corrected to the When soil gases are extracted from soil sam times corresponding to the times at which the ples collected in the vicinity of a petroleum de ionization currents are determined for the mix posit, minute quantities of hydrocarbons such as ture. ethanenpropane and butane are normally found. When the gas sample to be analyzed is small, ’Such hydrocarbons, or other substances, which 60 its composition may be determined by the method may be indicators of petroleum deposits, may be outlined above. If the sample is large, certain identified by my method even when non-indi simplifications may be made in the computation cators such` as methane CHi and ethylene C21-I4 procedure. A large sample chamber may be used are present. ' " In adapting my method to soil gasr analysis I 65 to hold a large sample, the inlet port 3 may be made smaller in diameter, and the analysis car prefer to concentrate significant hydrocarbons ried out without exhausting the sample during by any conventional method, such as temperature the run. With large samples contained in large separation, prior to introducing the sample into sample 'chambers and admitted slowly to the the mass spectrometer sample chamber l. While ionization chamber the composition of the sample 70 it is not possible to completely separate minute does not change appreciably during the course of quantities of hydrocarbons from each other, yet the readings and the time decay of the various by concentrating them the introduction of rela ion currents is not appreciable. Under these con tively large amounts of petroleum indicatorsinto ditions the standard times at which the readings the sample chamber is facilitated while still maintaining the total pressure and the mean free 75 areV made need not be determined accurately, if ratio. Y 2,412,236 13 at all. Under some conditlonsit is clear that the decay of ion currents will not be appreciable in the time interval during which readings are made and that for all practical purposes the readings may be considered as having been made simul taneously. I claim: ' sure therein at level such that the number of ionsv derived from a component varies in accord ance with the partial pressure of that component and independently of the partial pressuresfof .the other components. _ > " 5. In a method of analyzing a gaseous mixturev » involving admitting the mixture from a sample -1. In a method of analyzing a gas mixture with chamber into an ionization zone through a pas: a, mass spectrometer having an ionization cham ber and a sample chamber connected thereto, the 10 sage, ionizing'components of the mixture ‘in the zone, withdrawing the resulting ions from the improvement which comprises ñowing the mix z'one, and determining the` ture from the sample chamber into the ionization chamber while maintaining the sample chamber amounts of with- . drawn ions of a selected mass-to-charge ratio, the improvement which comprises flowing the ccm pressure and the ionization chamber pressure so low that each component of the mixture flows 15Y ponents of the mixture simultaneously into the ionization zone but at mutually independent rates from the sample chamber into the ionization by maintaining in the sample chamber a pressure chamber at a, rate dependent on the partial pres ‘ that is higher than in the sure of that component in the mixture and in ionization zone but so low that_the mean free path of molecules of the dependent of the partial pressure of any other component of the mixture. _ components of the V'mixture in the sample lcham - 2. In a, method of analyzing a gas mixture with 20 ber is at least as long as about half of the least cross-sectional dimension of the passage, and a mass spectrometer having an ionization cham maintaining the pressure in the ionization zone ber and a sample chamber connected thereto, such that the number of ions derived from each the improvement which comprises flowing the component varies in accordance with the partial mixture from the sample' chamber into the 25 pressures of that component and independently ionization chamber rwhile maintaining the sample of the partial 'pressures of the other components. chamber pressure and the ionization chamber 6. In the analysis of a, gas mixture containing’ pressure so low that each component of the mix a plurality of components with a mass spec ture iiows from the sample chamber into the _trometer having an ionization chamber and a ionization chamber at a rate >dependent on the partial pressure of that component in the 'mixture 30 sample chamber connected thereto through an and independent of the partial pressure of any oriñce, the improvement other component of the mixture, and ionizing components of the mixture in the ionization sure iiowing the mixture through the orifice from the sample chamber into the ionization chamber while maintaining the pressures in the chamber while maintaining the pressure therein such that the number of ions derived from an in dividual component varies in accordance with the partial pressure of that component and in dependently of ‘ the partial pressures of other components. 3. In a method of analyzing a gas mixture with a mass spectrometer having an ionization cham ber and a sample chamber connected thereto, the which comprises pres- 35 chambers at such values as to flow each com ponent at the same rate with which it would flow if it alone were present, simultaneously ionizing leach component inthe ionization chamber while 40 improvement which comprises iiowing the mix maintaining the pressure inthe ionization cham ber at a value such that each component is ionized to the same extent that it would be ionized if it alone were present, and measuring the rate of formation of resulting ions of a selected mass-to charge ratio. ture from the sample chamber into the ionization ’ chamber and simultaneously ionizing components _ ‘ ' ' 7. In analyzing a gas mixture with a mass spectrometer involving passing the mixture from of the mixture in the ionization chamber while a simple region through an oriñce into an ioniza maintaining the sample chamber pressure and tion region, the improvement which comprises the ionization chamber pressure so low that each component of the mixture flows from the sample 50 ñowing each component from _the sample region into the ionization region at a rate which varies chamber into the- ionization chamber at a rate directly with the partial pressure of said each dependent on the partial pressure of that com component and inversely as the square root of the ponent in the mixture and independent of the partial pressure of any ’other component of the molecular weight thereof by maintaining the pres mixture, the pressure in the ionization chamber sure in the sample region greater than the pres sure in the ionization region and at a value at also being such that the number of ions derived which the mean free path of molecules is suiii ciently large for the molecules of each component to pass through said orifice without substantial collision with other molecules. from an individual component varies in accord ance with the partial pressure of that component and independently of the partial pressure of the other components. 4. In a method of analyzing a gaseous mixture involving admitting'the mixture into an ioniza tion zone from a. sample chamber, ionizing com ponents of the mixture in the zone, withdrawing resulting ions from the zone, and determining the amount of withdrawn ions of a selected mass 60 8. In a method of analyzing a gas mixture with a mass spectrometer, the improvement which comprises diiîusing each component of a gas mix ture into and out` of an lionization region at an ' independent rate, and maintaining the pressures of the gas in the mass spectrometer at a value such that the diiïusion rates of the respective components are inversely proportional to the flowing the components of the mixture into the square root of the molecular weights of the re ionization zone simultaneously and at rates de spective -components, and measuring the relative pendent upon the partial pressure of the respec 70 rates of formation of ions of diiîerent mass-to tive component and independent of the partial charge ratios produced in said region. pressures of the other components by maintain 9. In a method of analyzing alimited quantity ing the pressure in the sample chamber below a of a gas mixture originally contained in a limited speciñc level, and ionizing the components in the ionization zone while-maintaining the pres 75 sample region of a mass spectrometer in which only a single ion current corresponding to a single to-charge ratio, the improvement which comprises - - 2,412,236 16 15 ionization zone, the ionization of components of measured at any one time while the pressure of a component in said mixture is diminishing at a the mixture in said zone, the withdrawal of the resulting ions from the zone, and the determina tion of the amounts of withdrawn ions of difier -predetermined mass-to-charge ratio may be substantial rate, the improvement which com prises the steps of continuously admitting said mixture into the ionization chamber of said mass spectrometer While maintaining the sample region pressure and ionization region pressure at levels such that the respective components of the mix- . ture iiow from the sample region into the 'ioniza 10 tion region at mutually independent rates, suc cessively measuring ion currents corresponding to ions of different mass-to-charge ratios, said measurements being made at predetermined times after initiating the ñow of said mixture into said ent mass-to-charge ratios, the improvement which comprises admitting the components of the mixture simultaneously to the ionization zone but at mutually independent rates, and maintaining the pressure in the ionization zone at such a low value that the distance travelled in the zone by ions being withdrawn from the zone is relatively short as compared with the mean free path of molecules of the gaseous com ponents of said zone, whereby the amounts of ions of each mass-to-charge ratio formed inthe ionization of the mixture are equal respectively to the linear sum of the quantities of such ions which would be formed if each of the compo nents were present alone. ionization chamber, separately admitting into said ionization chamber known quantities of sub stances corresponding chemically to the respective components of said gas mixture, and measuring 13. In a method of analyzing a gas mixture, ion currents of said mass-to-charge ratios at times 20 the improvement which comprises the steps of corresponding to said predetermined times after pressure iiowing the gas mixture from a high initiating the ñow of each of »said substances into pressure sample region into a low pressure analysis region while maintaining the pres composition of said mixture by comparing the ion sures in said regions at values such that each 25 currents measured for said mixture at such times component in the mixture ñows at a. rate which with the ion currents measured for said sub is independent of the presence of other compo stances at corresponding times. nents present in the mixture, ionizing molecules 10. In a method of mass spectrometry involv of each component in the low pressure region ing the ñow of diñerent components of a mixture while maintaining the pressure in said region so from a limited sample region into an ionization 30 low that ions are produced from the components . said ionization chamber, and determining the region at such different rates that the composi tion of the mixture is changing during the analy- - sis, the improvement which comprises 'ñowing the `mixture into the ionization region, ionizing the ` in amounts corresponding to the partial pres ,Y sures of the respective components in said low pressure region, and measuring the respective rates of formation of ions of different mass-to mixture `in the ionization region, successively 35 measuring at predetermined times the rates of formation of ions of different mass-to-charge ratios formed while the amount ofthe mixture charge ratio so formed. ‘ 14. In a method of analyzing a, gas mixture with a mass spectrometer having an ionization region, and a sample region connected thereto in the sample region is decreasing so as to obtain a restricted orifice, the steps which com a mass spectrum of the mixture, similarly obtain 40 through prise pressure iiowing diiîerent components of ing mass spectra of substances corresponding , the gas mixture from the sample region into the chemically to individual components, and deter mining the composition of the mixture by com paring the measured rates of formation of the respective ions in the mass spectrum of the mix ture with those in the mass spectra ofthe sub ionization region through the orifice at difier ent rates whereby the-relative amounts of the components remaining in the sample chamber are changed during the iiow process, while main taining the pressure in the sample region at such stances and in this determination compensating a value that the molecules of the respective for changes in the composition of said mixture components enter the oriñce from a relatively during analysis by reducing the measurements small part of the sample region adjacent the of the ion formation rates of given mass-to 50 mouth of the orifice without .any substantial pro change ratio in the mixture and in the individual portion of collisions with other- molecules and substances to a common time basis by correct mixing the components remaining in the sample ing the measurements in accordance with the region rapidly enough in relation to the diiïer rates of iiow of the individual components. ences in said ñow rates and the dimensions of 11. In a method of analyzing a gaseous mix- .Ul Cil the sample region to maintain the mixtureof lture involving admission of the mixture into an the gas sample substantially homogeneous ionization zone, the ionization of components of throughout the entire sample region, thereby the mixture in said zone, the withdrawal ofthe maintaining a typical sample in the small part resulting ions from the zone, and the determina tion of the amounts of withdrawn ions of differ 60 lof the sample region adjacent the mouth of the oriñce representative of the entire sample in the ent mass-to-charge ratios, the improvement sample region. which comprises admitting the components of 15. In the analysis of a gas mixture contain the mixture simultaneously to the ionization ing a plurality of components with a mass spec zone but at mutually independent rates, and having an ionization chamber, a sample maintaining the pressurel in the ionization zone 65 trometer chamber connected thereto through a first aper at such a vlow value‘that ions being withdrawn from the zone do not collide substantially with molecules of the gaseous mixture, whereby the amounts of- ions of each mass-to-charge ratio ture, and a low pressure zone connected thereto through a second aperture, the improvement which comprises pressure flowing the mixture through the ñrst aperture while maintaining the formed in the ionization of the mixture are equal 70 pressures in the two chambers at such low val respectively to the linear sum of the quantities ues that each component flows into the ioniza of such ions which would be formed if each of tion chamber at the same rate with which it the components were present alone. would flow if it were present alone, ionizing each 12. In a method of analyzing a gaseous mix component in the ionization chamber while the ture involving admission of the mixture into an 75 17 2,419,230 pressure in the ionization chamber is of such a low value that each component is ionized in the same manner that it would be ionized it it were present alone, withdrawing the resulting ions of each component simultaneously from the ionization chamber at the same rate that the ions of each component would be withdrawn if that nected to the intermediate chamber through a second conduit, and a stopcock in said second conduit. 18. In a mass spectrometer. the combination which comprises an ionization chamber, a sam ple chamber, an intermediate chamber, a ñrst conduit having a restricted oriiice therein con component were present alone and were being nectingthe intermediate chamber to the sam ionized, pressure flowing all of the non-ionized ple chamber, a valve in said conduit between said molecules from the ionization chamber into the 10 oriñce and said intermediate chamber, means low pressure zone by maintaining the relationship for exhausting the intermediate chamber inde between the pressures in the ionization chamber pendently of'the sample chamber and connected and such zone at such values that the non to the intermediate chamber through a second ionized molecules of each component flow into conduit, and a valve in said second conduit. ` the zone at the same rate at which they would 15 19. In a mass spectrometer, the combination flow if that component were present alone, and which comprises an ionization chamber, a sam measuring the rate of withdrawal oi' the result ple chamber, an intermediate chamber of iixed ing ions oi’ a selected mass-to-charge ratio, volume, means for admitting a portion oi’ a sam whereby the contribution of each component to ple from the sample chamber into the inter the total number of such ions withdrawn is the 20 mediate chamber in gaseous form, a conduit con same as it would be if the component were pres ent alone. 16. In a mass spectrometer, the combination which comprises an ionization chamber, a sam necting said intermediate chamber and' said ionization chamber, a valve in said conduit, means for indicating the pressure of gas present . in said intermediate chamber, a second'conduit ple chamber, a conduit connectingthe sample 25 connected to the intermediate chamber, and chamber to the ionization chamber, and a tube . means connected to the intermediate chamber sealed within said conduit with at least part of through the second conduit withdrawing a por the tube spaced from the conduit wall and ex tion of any gas contained therein. tending i'rom the sealed portion thereof in the 20. In a mass spectrometer, the combination direction of the sample chamber and opening to 30 which comprises an ionizationchamber, a sam ward the sample chamber. ple chamber, an intermediate chamber, a first 17. In a mass spectrometer, the combination conduit having a restricted oriñce therein con which comprises an ionization chamber, a sam necting the intermediate chamber to the ioniza ple chamber, an intermediate chamber of fixed tion chamber, a valve in- said conduit between volume, a i‘lrst lconduit containing a restricted said orifice and said lvintermediate chamber, orii‘lce connecting the intermediate chamber to evacuating means, a second conduit connecting the sample chamber, a stopcock in said conduit. the intermediate chamber and the evacuating means i’or exhausting the intermediate chamber means,y and a valve in said second conduit. independently of the sample chamber and con HAROLD W. WASHBURN.