INVESTIGATION OF THE IONIC COMPOSITION OF IONIZED LAYERS OF THE ATMOSPHERE

B. A. Martynov, V. G. Istomin

1. INTRODUCTION

One of the most important parameters of the ionosphere should be considered the chemical composition of its ions. Investigation of the spectrum of ions in the ionosphere (i.e., of its composition) is extremely important for solving a number of geophysical and astrophysical problems. Studies of this kind may, for example, help in solving such a fundamental geophysical and, at the same time, astrophysical problem as the Sun–Earth problem.

It is very probable that the spectra of ions at low latitudes of the Earth and in its polar regions (especially at times of auroras) will differ from one another. This difference must be due to the different mechanisms of ionization of the atmosphere acting in the regions named—ultraviolet irradiation at low latitudes and corpuscular radiation in the polar regions (especially during the polar night).

Investigation of the ionic composition of the ionosphere may also help in solving the problem of the origin and existence of ionized layers at different altitudes. This will become possible by studying the change in ionic composition over the course of a day (from day to night), and also by studying this composition under conditions of deep polar night, during a prolonged absence of such a powerful ionizing agent as the ultraviolet radiation of the Sun.

Finally, if we return to the conditions of propagation of radio waves, knowledge of the ion spectrum is of definite interest from the point of view of their effective cross sections for processes of collision with electrons. To construct a complete theory of radio-wave propagation, it is apparently impossible to neglect such collisions, and the effective cross sections of ions in this process are, of course, quite different from the effective cross sections of neutral atoms or molecules.

At present we possess only general, more or less reliable qualitative data on the composition of ionized layers, obtained by indirect methods. However, these methods do not make it possible to judge the most important matter—the variations in the composition of the ionosphere not from the qualitative, but from the quantitative side of the phenomenon. To detect these variations, mass-spectrometric investigations are needed, which can be carried out only with the aid of special measuring apparatus sent into the atmospheric layers under investigation. The difficulty of sending instruments to altitudes of the order of 200–400 km and higher has until now made such investigations impossible, and only very recently has the conduct of high-altitude experiments obtained a reliable foundation: special instruments have been created that make it possible to carry out the indicated measurements, and launching means have been created that make it possible to raise these instruments to great heights.

2. ARTIFICIAL SATELLITE AND THE STUDY OF THE ION SPECTRUM IN THE IONOSPHERE

At present, two ways of investigating the ionic composition of the ionosphere by direct methods are conceivable and possible: lifting the corresponding apparatus on rockets and using an artificial Earth satellite. The first method, as is known, is already being used to obtain information on ionic composition; however, unfortunately, it can provide only very limited information, since the duration of a rocket’s flight in the ionized regions of interest to us does not exceed several minutes. In addition, with the aid of rockets it is difficult, if not hopeless, at the present time to carry out simultaneous studies with sufficient density of sounding points, including even hard-to-reach regions of the Earth. The second method (still inaccessible to researchers for the time being) is the measurement of the ion spectrum by an instrument placed on an artificial Earth satellite. It is precisely this method that appears most promising in investigations of the ionic composition of the ionosphere.

One of the decisive advantages of experiments conducted on a satellite is the length of time the measuring apparatus remains in the layers under study. In combination with the enormous speed of motion of the satellite (\(\sim 8\) km/sec), this feature makes it possible to carry out multiple and almost simultaneous observations at points separated from one another by tens of thousands of kilometers. The difference in time between studies in the equatorial zone and in the polar region may amount to no more than 20–30 minutes. The repeated appearance of the satellite in the same zones (orbital period about 90 minutes) makes it possible to trace changes in ionic composition over time—changes associated with variations of solar activity in the ultraviolet or in corpuscular fluxes.

The problem of ionic composition in the night-time and daytime should likewise be well solved on a satellite. During the 90 minutes of its flight the satellite will be located both on the sunlit and on the shadow side of the Earth, measuring, respectively, the daytime and nighttime ion spectra.

Another possibility that presents itself when carrying out investigations on a satellite is that, owing to the elongation of the orbit, measurements of ionic composition can be made at different heights above the Earth. As a result, the investigations can encompass both of the most important layers of the ionosphere—the \(E\) and \(F\) layers. A satellite has yet another substantial advantage. A rocket, entering the rarefied layers of the atmosphere, begins to release a large quantity of “parasitic” gases, which surround the rocket with a peculiar gas cloud. This cloud consists of air that fills the rocket before launch, products of fuel combustion, and also gases formed during intensive evaporation of unused propellant. Because of the brevity of the rocket experiment, and also the enormous reserve of gases contaminating the atmosphere, the rocket does not have time to leave the contaminated region, which, naturally, adversely affects the measurements carried out at the surface of the rocket.

The situation is entirely different with a satellite: its carefully sealed small volume makes it possible to reduce the release of gases from within to a minimum, while the length of time the satellite remains in the rarefied layers creates the possibility of good degassing of the satellite surface and, what is especially important, degassing of the interiors of the measuring instruments themselves. This advantage of the satellite has an indirect, but sufficiently substantial, influence on the investigation of the composition of ions.

3. SOME GENERAL QUESTIONS OF CONDUCTING THE EXPERIMENT

In those regions of the atmosphere where an artificial satellite will be located, the medium is so rarefied that the free paths of molecules reach tens and hundreds of meters. The motion of the satellite itself takes place at a speed an order of magnitude greater than the gas-kinetic speeds of molecules. These conditions place the experimenter in a very difficult position, not only from the standpoint of choosing suitable measuring instruments, but also from the standpoint of the possibility of measuring the undisturbed parameters of the medium in which the satellite is moving.

When studying the ionic composition of the ionosphere it is extremely important to ascertain whether the instrument placed on the satellite will measure true or fictitious ionization. It is well known, for example, that the flights of meteors through the earth’s atmosphere are accompanied by intense ionization of the atmosphere. True, meteors fly at speeds considerably exceeding 8 km/sec, and the ionization is observed in lower layers than those in which the satellite will move; however, in view of the importance of this question for the results of all experiments in the ionosphere, it seems necessary to us to dwell on it in detail. In order to ionize the molecules of the surrounding medium, a body must possess an energy of not less than approximately 15 eV. The satellite itself does not possess such energy and therefore cannot directly ionize the gas molecules it encounters. But molecules rebounding from the surface of the satellite with a velocity equal to twice the satellite velocity (an absolutely elastic collision), i.e., with a velocity of 16 km/sec, already possess quite sufficient energy for ionization by impact. This is where the danger lies. To estimate the probability of this danger, it is necessary to determine whether gas molecules undergo an elastic or an inelastic collision with a rapidly moving surface.

At present it is difficult to decide this question definitively, since there are no reliable experimental data on the interaction of matter with gas molecules at such high velocities. But from a number of indirect data one may suppose that the overwhelming majority of molecules will be reflected from the satellite surface according to the laws of inelastic collision. Thus, for example, all modern theories of meteors (Sparrow, Epik, Gerloffson) proceed precisely from inelastic collisions of a meteor with gas molecules. Experiments also confirm the inelastic character of the interaction between a solid surface and gas molecules. The surface strongly absorbs the molecules striking it, which only after a certain time (\(10^{-4}\)—\(10^{-5}\) sec) “evaporate” from the surface with the velocity of thermal motion (Langmuir and Knudsen). However, taking into account the specificity of the experiment and the difficulty of an unambiguous solution of the question of the character of the collision, we shall consider the worst case, when all molecules colliding with the satellite are reflected elastically.

Let \(N_0\) be the number of molecules in a unit volume in the medium surrounding the satellite, \(s_0\) the surface of the satellite subjected to bombardment by molecules, \(v_0\) the velocity of motion of the satellite. During a time \(\Delta T\) the satellite will set into intense motion the following number of molecules:

\[ \Delta N = N_0 s_0 v_0 \Delta T . \tag{1} \]

Owing to the chaotic motion of the molecules and the unevenness of the reflecting surface, the reflected molecules will be directed toward the front half of a sphere whose center is the point of the initial impact*). They will move unhindered over a distance of the free path, character-

*) This is confirmed by the width of the “primary” meteor trails, which always agrees well with the magnitude of the free paths of molecules at the given heights.

only for the given altitudes, and for the first time will collide with molecules of the surrounding medium on the surface of a hemisphere of radius \(\lambda\), where \(\lambda\) is the mean free path.

Over the time interval \(\Delta T\) the satellite will be displaced by a distance \(\Delta \lambda\), and, consequently, the collision of the “fast” molecules with the “slow” molecules of the surrounding medium will occur in a layer bounded by radii \(\lambda\) and \(\lambda+\Delta\lambda\). The volume of half of a spherical layer with the indicated radii will be

\[ V=2\pi\left(\lambda^2\cdot\Delta\lambda+\lambda\cdot\Delta\lambda^2+\frac{\Delta\lambda^3}{3}\right). \]

Putting here \(\Delta\lambda=1\) and taking into account that \(\lambda\) is a quantity of the order of \(10^4\), we may, without great error, neglect in the brackets all terms except the first. Then

\[ V=2\pi\lambda^2 . \tag{2} \]

Fig. 1. Diagram of the interaction of a satellite with air molecules in the upper layers of the atmosphere.

The number of collisions \(\eta\) per unit volume during the time \(\Delta T\), taking (1) and (2) into account, will be

\[ \eta=\frac{\Delta N}{V}=\frac{N_0 s_0 v_0 \Delta T}{2\pi\lambda^2}. \tag{3} \]

But not all collisions lead to ionization. Therefore, in order to obtain the number of elementary ionization events \(\eta^*\) arising in a unit of the investigated volume, it is necessary to introduce into (3) the ionization coefficient \(\alpha\):

\[ \eta^*=\alpha\eta=\alpha\frac{N_0 s_0 v_0 \Delta T}{2\pi\lambda^2}. \tag{4} \]

In order to calculate how many of these newly formed ions will meet the satellite, let us turn to Fig. 1. In this figure the center of the sphere \(O\) is the place of the “initial” impact. From here the molecules with velocity \(2v_0\) scatter over the front hemisphere \(\overline{BB}\) with radius \(R=\lambda\). During the same time the satellite, having traversed a path \(\lambda/2\) (since its speed is half that of the fast molecules), will be at point \(A\). Taking into account that the ions formed possess thermal velocities, while the speed of the satellite exceeds these velocities by an order of magnitude, it is easy to see (Fig. 1) that, for sufficiently large \(\lambda\), only an insignificant fraction of the ions produced by it can meet the satellite. Indeed, owing to the indicated difference of velocities, the first collision of the satellite with these ions will occur at a distance \(\lambda/22\) from point \(D\).

With sufficient approximation it can be shown that in this case ions can reach the satellite only from that part of the surface \(\overline{BB}\) which is bounded by a circle of radius \(\lambda/20\). On this part of the surface there are \(N^*\) newly formed ions, with

\[ N^*=\eta^*\cdot s_1, \]

where

\[ s_1=\pi\left(\frac{\lambda}{20}\right)^2=\frac{\pi\lambda^2}{4\cdot 10^2}. \]

Substituting the values for \(\eta^*\) and \(s_1\), we obtain

\[ N^*=\frac{\alpha N_0 s_0 v_0 \Delta T}{8\cdot 10^2}. \tag{5} \]

The ions \(N^*\), like all the others, will move in any direc-

...and only a small part of them will reach the satellite surface. For simplicity of reasoning, one may assume that all \(N^*\) ions are concentrated at the center of a small sphere with center at \(D\) and from there, spreading in all directions, “irradiate” the satellite, which is located from \(D\) at the distance of the “first” collision, i.e., at a distance \(\lambda/22\). At this distance from \(D\) the density of the formed ions \(\eta_1^*\) will be equal to

\[ \eta_1^*=\frac{N^*}{s_2}, \tag{6} \]

where

\[ s_2=4\pi\left(\frac{\lambda}{22}\right)^2 \simeq \frac{4\pi\lambda^2}{5\cdot 10^2}. \]

Substituting the values of \(N^*\) and \(s_2\) into (6), we obtain:

\[ \eta_1^*=\frac{5\alpha N_0 s_0 v_0 \Delta T}{32\pi\lambda^2}. \tag{7} \]

It remains to substitute numerical values into (7). According to Herlofson, for meteors the coefficient \(\alpha\) is taken equal to \(10^{-4}\), for an altitude of \(250\ \text{km}\). \(N_0=10^{10}\); considering the satellite to be spherical with a radius, say, of \(25\ \text{cm}\), we have \(s_0=\frac{s}{2}=4\cdot 10^4\ \text{cm}^2\), the satellite velocity \(v_0=8\cdot 10^5\ \text{cm/sec}\), and \(\lambda=10^4\ \text{cm}\). We choose the time interval \(\Delta T\) so that during this time the satellite is displaced by \(1\ \text{cm}\), i.e. \(\Delta T=1/8\cdot 10^{-5}\ \text{sec}\). Then

\[ \eta_1^*= \frac{5\cdot 10^{-4}\cdot 10^{10}\cdot 4\cdot 10^4\cdot 8\cdot 10^5\cdot \frac{1}{8}\cdot 10^{-5}} {32\cdot 3.14\cdot 10^8} \simeq 20\ \text{ions}/\text{cm}^3. \]

Thus, the artificial ionization arising as a result of the motion of the satellite near its surface and equal to \(20\ \text{ions}/\text{cm}^3\) is negligible in comparison with the natural ionization, amounting to \(10^5\)—\(10^6\ \text{ions}/\text{cm}^3\). Such a ratio is due to the fact that the satellite passes as if through a tunnel whose ion walls are formed by the satellite itself. Owing to the enormous speed of the satellite and the large mean free paths of the molecules, the newly formed ions, in their overwhelming majority, do not have time to return to the satellite, and from this point of view the satellite moves as if in an undisturbed medium.

On the basis of the foregoing, it should be considered that in all work on the satellite the artificial ionization of the surrounding space created by it may be neglected. This same conclusion should also be extended to all other processes connected with the high energies of the molecules flowing around the satellite (for example, thermochemical reactions, etc.).

The second essential phenomenon associated with the high speed of motion of an artificial satellite is the presence of a deep vacuum behind it. This vacuum is formed because the speed of the satellite exceeds the gas-kinetic speeds of the molecules by an order of magnitude. In its flight the satellite punches a tunnel in the atmosphere, which the surrounding molecules do not have time to fill. As a result, behind the satellite there is formed a “wake cone,” into which only electrons and a small number of very fast molecules, always present in the gas, can penetrate. Measurement of the ionic composition of the ionosphere in the region of this cone is doomed to complete failure, since instruments that fall into the “wake cone” will generally cease to operate because of the insufficient density of ions. If the satellite is not oriented in space, then one must reckon with the fact that from time to time the apparatus will nevertheless enter...

into this “dead” space and its readings will be equal to zero. However, it by no means follows from this that the undisturbed ionosphere is characterized by an absence of ionization. With an oriented Sputnik this danger disappears, and the indicated measurements can be carried out with considerably greater reliability.

4. INSTRUMENTS FOR THE DIRECT STUDY OF THE IONIC COMPOSITION OF THE UPPER ATMOSPHERE

If analysis of the gaseous composition of the atmosphere can be conducted by several different methods, then for the direct study of the ionic composition of the upper atmosphere the only possible method, apparently, is the mass-spectrometric one. By installing a mass spectrometer on a rocket or a satellite flying in the ionosphere and transmitting the data by radio to the Earth, one may hope to obtain information on the mass composition of the ionized layers.

Like all instruments installed on rockets or satellites, the mass spectrometer must operate automatically and be, as far as possible, low in energy consumption. In addition, the design of such an instrument must meet a whole series of specific requirements, namely: mechanical strength, resistance to large overloads, vibration resistance, the ability to withstand brief significant rises in temperature, the ability to operate under conditions of high vacuum, etc.

The widely known “magnetic” mass spectrometers, i.e., instruments in which both electric and magnetic fields are used to separate the sample under investigation according to mass, are of little suitability for this purpose. With all the merits of modern magnetic mass spectrometers (high resolving power, great sensitivity, and relatively high accuracy of analysis), they possess a number of shortcomings that make their use in experiments of this kind extremely difficult. The analyzer of a magnetic mass spectrometer is an ion-optical system whose necessary elements are, first, a magnet and, second, a series of slits and diaphragms that form and restrict the ion beam in the instrument. These circumstances lead to the fact that the dimensions and weight of such instruments turn out to be rather considerable, and the instruments themselves require careful adjustment and tuning. An additional difficulty is the extremely small ion currents obtained from the collector of a magnetic mass spectrometer. The amplification of small direct currents is associated with considerable experimental difficulties, one of which is the large time constant of the whole device; this, in turn, imposes limitations on the speed of scanning the mass spectrum. The reasons enumerated apparently explain why up to now there has not been a single successful launching on a rocket of a magnetic mass spectrometer.

Alongside magnetic instruments there exist quite a few types of mass spectrometers not connected with the use of a magnetic field in the ion analyzer. One such instrument is the radio-frequency mass spectrometer, in whose analyzer crossed electric fields are used¹, and one of its varieties is the radio-frequency mass spectrometer of the Bennett type². This instrument, while possessing sufficient resolving power, can be made tens of times lighter than a magnetic mass spectrometer, operates faster, and gives currents at the collector that are 2–4 orders of magnitude larger³. An important feature of the latter instrument is also that, in the process of adjustment and operation, it requires no mechanical regulation. A shortcoming of the radio-frequency instrument

is the low resolving power (in comparison with magnetic mass spectrometers). However, for solving a number of geophysical problems mentioned above, this resolving power is quite sufficient. In this connection it makes sense to discuss in greater detail the principal characteristics of the radio-frequency mass spectrometer from the standpoint of the possibility of using it on an artificial Earth satellite.

5. RADIO-FREQUENCY MASS SPECTROMETER OF THE BENNETT TYPE

In the Bennett-type radio-frequency mass spectrometer, the principle used is that of separating ions according to their velocities. The main element of the instrument is the mass-spectrometric tube, which is a specially designed high-vacuum lamp with a large number of plane-parallel grids.

Let us consider a simplified diagram of the mass-spectrometric tube (Fig. 2). To a system consisting of three parallel, equally spaced grids there is applied a negative sawtooth accelerating potential \(V\). In addition, an alternating high-frequency voltage \(U = U_0 \sin(\omega t + \theta)\) is applied to the middle grid, its amplitude being small in comparison with the accelerating voltage \((U_0 \ll V)\).

An ion passing through this system of grids with a certain velocity, depending on its mass and on the magnitude of the accelerating voltage at the given moment, gains or loses a certain amount of energy, while its velocity remains unchanged to a first approximation. It can be shown that the maximum energy from the high-frequency field of such a three-grid system is received only by an ion that enters it at a certain definite phase of the high-frequency voltage and passes through it with a definite velocity \(v_0\). This optimal velocity is imparted, when the accelerating voltage is varied according to a sawtooth law, successively to ions of all masses within a certain range of mass numbers.

Fig. 2. Simplified diagram of the tube of a radio-frequency mass spectrometer.

If a fourth grid is then placed in the path of the ions, and a suitable positive retarding potential is applied to it, all ions can be stopped except those that have received the maximum energy from the high-frequency field, i.e., precisely those that have passed through the system with the optimal velocity.

Since the velocity of an ion depends on its mass and on the accelerating voltage, then, knowing the value of the optimal velocity for the given grid system and the magnitude of the accelerating voltage, one can determine the mass of the ions that have overcome the potential barrier of the fourth grid and reached the collector. The ion current of the collector can be amplified and recorded by some recording device. The record will consist of a series of ion-current peaks, each such peak corresponding to ions of a definite mass.

The mass of an ion is related to the accelerating sawtooth voltage by the following relation:

\[ M = \frac{0{,}266}{s^2 f^2}\, V, \tag{8} \]

where \(M\) is the mass number of the ion, \(V\) is the magnitude of the sawtooth accelerating voltage in volts, \(s\) is the distance between grids in centimeters, and \(f\) is the frequency in megahertz.

In practice, the system described operates unsatisfactorily; therefore, in Bennett’s radio-frequency mass spectrometer the ion analyzer is a system of three three-grid sections separated by drift spaces. These drift spaces are chosen to be of such a size that the time of flight through them by an ion traveling with the optimal (synchronous) velocity \(v_0\) is a multiple of the period of the high-frequency voltage. This is necessary so that an ion arriving at the first section with the optimal input phase of the high-frequency voltage preserves it for the other two sections as well. The magnitude of the drift space is commonly expressed by the number of periods of the high-frequency voltage (the number of cycles) occurring during the ion’s flight through it.

Fig. 3. Circuit diagram for connecting a 7—5-cycle tube of a radio-frequency mass spectrometer.

Different versions of mass-spectrometer analyzers of this type may contain different numbers of cycles. For example, there are 9—7-cycle, 5—9-cycle, and 7—5-cycle versions. On the basis of published data one may conclude that one of the best in resolving power is the 7—5-cycle tube version.

The complete connection diagram of the 7—5-cycle mass-spectrometric tube is shown in Fig. 3. Electrons emitted by the heated cathode are accelerated by grid \(1\) and, passing into the space between grids \(1\) and \(2\), ionize the gas contained in the tube. The ions formed are drawn out of the ionization space and accelerated by the constant negative potential of grids \(3\) and \(4\) and by the sawtooth negative potential of the analyzer grids \(5—13\). A bias potential is applied to the drift spaces \(A\) and \(B\) of the analyzer; it is selected so as to compensate for a certain increase in the ion velocity occurring in the second stage of the analyzer. Beyond the analyzer there is a group of grids \(14\), \(15\), \(16\), to which a positive retarding potential is applied. The last grid, \(17\), with a high negative potential, is needed to suppress secondary electrons that may be knocked out of the tube grids or from the collector.

Parameters of the instrument and its field of application

A radio-frequency mass spectrometer can be used for the analysis of both neutral and ionized gases forming part of the Earth’s atmosphere. When the instrument is used to analyze the ionic composition of the atmosphere, an ion source is not needed, since its functions are performed by the ionosphere itself.

A radio-frequency mass spectrometer with a 7–5-cycle tube has a mass resolution of about 20–25. Recall that the resolving power (resolution) of a mass spectrometer can be expressed by the ratio \(R=\frac{M}{\Delta M}\), where \(M\) is the mass number to which a definite peak of the ion current corresponds, and \(\Delta M\) is its width in units of mass numbers, measured at a specified level (at half the peak height, at the base, etc.). In the present case the resolving power of the instrument is defined at the base of the peak, and a resolving-power value of 25 means that the instrument is capable of completely resolving, for example, masses of 24 and 25 atomic units*).

The mass interval in which the instrument operates, as follows from relation (8), is determined by the geometrical dimensions of the analyzer (the distance between the grids of the three-grid section), the operating frequency, and the range of variation of the accelerating voltage. For example, the instrument described by Townsend\(^{3}\) covered the range from 5 to 48 atomic mass units.

An important characteristic of the instrument is the time within which a mass spectrum can be obtained. For a radio-frequency instrument this time is approximately 1 second and, if necessary, can readily be reduced. This property of the instrument is extremely important when its use is contemplated on rapidly moving objects—rockets or an artificial satellite.

The altitude interval in which the instrument operates is determined, on the one hand, by the geometrical dimensions of the tube (it is necessary that the mean free path be greater than or equal to the length of the radio-frequency analyzer), and, on the other, by the density of the ions entering the analyzer. It follows from what has been said that the region of altitudes above the Earth’s surface where an ion mass spectrometer can operate is bounded below by an altitude of the order of 100 km (here the mean free paths of molecules reach about 10 cm), and above by that altitude at which the density of natural ions is still sufficient to ensure normal operation of the instrument.

A simple calculation shows that, at a concentration of positive ions of the order of \(10^{5}\ \text{cm}^{-3}\), ion currents from the collector may reach \(10^{-8}\) ampere**). Apparently, the instrument can also operate at a lower concentration, of the order of \(10^{3}\ \text{cm}^{-3}\). As is known, the concentration of charged particles in the \(E\) and \(F\) layers is estimated to be of the order of \(10^{4}\)–\(10^{6}\ \text{cm}^{-3}\). Thus both these layers, whose heights above the Earth’s surface correspond to 200 and \(\approx 400\) km, are quite accessible objects for the investigations mentioned. At present it is not possible to specify precisely the greatest altitude up to which an ion mass spectrometer can be used, since it is not yet known at what altitude the ion concentration becomes less than \(10^{3}\ \text{cm}^{-3}\).

In 1954–1955 the first rocket flights of a radio-frequency mass spectrometer were carried out for the purpose of analyzing the ionic composition of the upper atmosphere\(^{4,5,6}\). In these works, some data were obtained on the mass spectrum of positive and negative ions at altitudes up to 219 km. Without entering here into a detailed analysis of the results obtained, we shall note only that they are, to a considerable extent, still fragmentary—

*) It should be noted that this value is attained when the instrument is used only for the analysis of neutral gases. The resolving power of an ion mass spectrometer, for certain reasons that will be discussed below, may differ significantly from this value.

**) Currents of this order were also detected during the rocket flight of the mass spectrometer\(^{4}\).

...unclear, and even contradictory. It would not be an exaggeration to say that, so far, all launches of the ion radio-frequency mass spectrometer have had the character of working out the experimental technique.

6. SOME PARTICULAR QUESTIONS OF CARRYING OUT THE EXPERIMENT

As far as can be judged from the published materials, one serious experimental difficulty has already emerged which complicates the use of the Bennett-type radio-frequency mass spectrometer in experiments to determine the ionic composition of the upper atmosphere. This concerns the influence of the rocket’s own charge acquired in the ionosphere. If the rocket (or satellite) on which the radio-frequency mass spectrometer is installed becomes negatively charged, this leads to a change in the operating regime of the instrument. First, this negative potential is added, wholly or in part, to the negative cycloidal accelerating potential \(V\) (equation (8)), as a result of which the mass scale of the instrument is shifted toward lighter masses. For each individual ion of a definite mass entering the analyzer, this additional potential will enter wholly or partially, depending on the configuration of the field around the satellite and on the distance over which the ion is accelerated, or experiences the latter influence. This leads to an increase in the spread in the velocities of the ions entering the analyzer and, consequently, to a deterioration of the resolving power of the instrument (broadening of the peaks in the mass spectrum). Secondly, this negative potential is subtracted (wholly or partially for the same reasons) from the retarding potential of the mass-spectrometric tube, thereby lowering the most effective retarding potential. This may lead to the appearance on the mass spectrum of false (so-called harmonic) peaks, which will make its interpretation difficult or impossible. In view of the fact that the causes of the appearance of negative charge are at present still unclear, and there is also almost no experimental material in this area, allowance for the influence of this factor is difficult*).

Taking into account the positive results that have been achieved with the use of the radio-frequency mass spectrometer on rockets, it is easy to see how expedient it appears to install such an instrument on an artificial Earth satellite. Indeed, as was already indicated above, a mass spectrometer installed on a satellite will make it possible to obtain detailed information on the composition of the ionosphere at different altitudes (within the satellite’s altitude interval), at different times of day, and at the most varied points above the Earth’s surface. In particular, with a proper choice of orbit it will be possible to obtain information from the difficult-to-access polar regions of the globe. From the standpoint of the quantity of information obtained, one successful satellite launch should be equivalent to hundreds of rocket experiments.

It should, however, be clearly understood that the experimental difficulties associated with the use of a radio-frequency mass spectrometer on rockets will increase manyfold when the instrument is installed on a satellite. For example, disturbances of the normal operation of the instrument associated with the satellite’s own charge will make the acquisition and processing of experimental data much more difficult, since

*) A radical solution would be to install, together with the mass spectrometer, an instrument measuring the satellite’s own charge; with the charge thus measured, the indications of the mass spectrometer would accordingly be corrected. However, the first difficulty—the deterioration of the resolving power of the instrument—is not thereby removed.

the charge will vary depending on the altitude of the satellite’s flight, its geographical coordinates, and the time of day. In particular, for an instrument installed on a satellite it will apparently be absolutely necessary to provide automatic adjustment of the retarding potential as a function of the charge acquired by the satellite.

A specific difficulty that complicates the performance of some experiments on a satellite is the high speed of its motion in orbit. Some of the difficulties arising in this connection were discussed above. Let us now consider the direct influence of the satellite’s speed on the operation of the radio-frequency mass spectrometer described. We shall examine three principal cases of orientation of the entrance aperture of the mass-spectrometer tube with respect to the satellite velocity vector:

a) The entrance aperture is directed backward. As was shown above, a cone of “molecular shadow” is formed behind the satellite, and only a very small number of fast ions can enter the analyzer—insufficient for normal operation of the instrument.

b) The entrance aperture is directed forward. In this case, the velocity \(v = 8 \cdot 10^{5}\ \mathrm{cm/sec}\) will be superposed on the components of the thermal velocities of the ions directed along the axis of the tube. This velocity will not produce any additional spread of the thermal velocities of the ions and will not impair the mass resolution. It will merely lead to a certain displacement of the ion-current peaks of the instrument on the mass scale toward smaller mass numbers. This will occur because the radio-frequency mass analyzer operates essentially as a velocity filter, passing to the collector ions that have passed through the analyzer with a certain definite (synchronous) velocity. The sawtooth sweep voltage imparts this synchronous velocity in turn to all ions whose masses fall within the range of the instrument. If, before entering the analyzer, the ions already have some ordered velocity directed along the axis of the tube, then the synchronous velocity will be reached for all masses at a lower sweep voltage and, consequently, the corresponding ion-current peaks will be shifted toward the lighter masses.

The shift will be different for ions of different masses. Let us try to estimate it. The added velocity \(v = 8 \cdot 10^{5}\ \mathrm{cm/sec}\) is equivalent to a change in the accelerating sawtooth voltage of the radio-frequency analyzer by the amount

\[ \Delta V_{\mathrm{eqv}} = \frac{m}{2q}\, v^{2}, \]

where \(m\) is the ion mass; \(q = 4.8 \cdot 10^{-10}\) CGSE is the ion charge; \(v = 8 \cdot 10^{5}\ \mathrm{cm/sec}\) is the added velocity. For singly charged ions of molecular hydrogen \((M = 2)\) we obtain \(\Delta V_{\mathrm{eqv}} = 0.68\ \mathrm{V}\). For argon ions \((M = 40)\), \(\Delta V_{\mathrm{eqv}} = 13.6\ \mathrm{V}\).

The radio-frequency mass spectrometer described by Townsend\(^3\) had a constant approximately equal to \(5\ \mathrm{V/atomic\ mass\ unit}\). Consequently, in this case the ion-current peaks would be shifted for hydrogen by \(0.136\) atomic mass units and for argon by \(2.72\) atomic mass units. (The relative decrease in mass number is \(6.8\%\).) Thus, when the entrance aperture of the mass spectrometer is located in the direction of the satellite’s flight, the instrument constant changes by a quite appreciable amount, which must be taken into account when interpreting the spectra.

c) The entrance aperture of the radio-frequency mass spectrometer is located perpendicular to the direction of the satellite’s flight. In this case neither the resolution

instrument nor its mass scale should change. There may occur only an apparent decrease in the relative content of heavy ions, due to unequal diaphragmation of the ion beam by the rings of the input (retarding) grids in the section of the tube before the radio-frequency analyzer. This effect can be estimated only by considering the specific design of the mass-spectrometric tube and by specifying definite potentials of the input grids and the body of the satellite. In the analyzer itself, the component of the ion velocity directed perpendicular to the axis of the tube will no longer produce such an effect, since in the analyzer all ions that reach the collector move with a velocity independent of their masses.

It follows from what has been said that when installing a mass spectrometer on a satellite oriented in space, method c) is the most acceptable; method b) is also acceptable, while method a) is completely inadmissible. In the case of a nonoriented satellite, it is necessary to know, for each instant of time, the orientation of the entrance aperture of the mass spectrometer and to take it into account in processing the spectra obtained.

Technical Difficulties

In addition to the difficulties of a fundamental nature considered above, the installation of a mass spectrometer on a satellite will inevitably involve a number of difficulties of design and technical character. First of all, it should be pointed out that the dimensions and weight of a radio-frequency mass spectrometer are still too large. Thus, for example, Townsend’s instrument3, with power supply for several minutes of operation (vertical launch), weighed 20 kg and occupied a volume of about 13 liters.

Another serious limitation on the use of a mass spectrometer on a satellite will apparently be imposed by the capacity of the memory device of the radio-telemetry system. To transmit a mass spectrum, which may include several tens of peaks of ion current, radio-telemetry channels with high resolution are required. Thus, for example, in one of the experiments with a radio-frequency mass spectrometer6, three channels with a polling rate of about 300 per second and one high-speed channel with a polling rate of 1200 per second were used to transmit the mass spectrum. For direct transmission from a rocket, existing radio-telemetry systems meet the necessary requirements with respect to the volume of transmitted information. The task, however, of creating so capacious a memory device for the time required for the satellite to pass between two receiving stations, and of transmitting all the accumulated information in a short interval of time, apparently presents considerable difficulties.

References Cited

1. P. A. Redhead, Canadian J. of Phys. 30, No. 2 (1952).
2. W. H. Bennett, J. of App. Phys. 21, No. 2 (1950).
3. J. W. Townsend, Rev. Sci. Inst. 23, No. 10 (1952).
4. C. Y. Johnson, E. B. Meadows, J. of Geophys. Res. 60, No. 2 (1955).
5. C. Y. Johnson, J. P. Heppner, J. of Geophys. Res. 60, No. 4 (1955).
6. E. B. Meadows, J. W. Townsend, J. of Geophys. Res. 61, No. 3 (1956).