For Cs on tungsten (f _o = 4.5 eV), H » 2.8 eV is obtained from (7.2), which agrees well with the experimental value: H » 3 eV [8]. Equation (7.2) easily explains the well known experimental fact that Cs is bound more strongly to those monocrystal faces that have the greater vacuum work function (for example, the (100) face of a monocrystal of tungsten with (f _o = 5.20 eV).

where N_max is the greatest number of particles which can be arrayed on 1 cm of surface, with dense packing. The value f = 1 corresponds to a monolayer.

As the extent of coverage increases, adsorbed Cs ions begin to be affected more by mutual interaction leading to a decrease in adsorption energy. Thus, for Cs on W, the heat of adsorption for a single added Cs atom decreases from about 3 eV at the beginning of a monolayer to about 2 eV at the completion of the monolayer. As the density of the adsorbed ions increases, the positive potential of the adsorbed layer increases, and it becomes energetically more advantageous for the electrons to partially transfer to the adsorbed ions. This leads to a decrease of the dipole moment per adatom, and also to a decrease of the ionic and an increase in the covalent components for binding the surface layer to the substrate. Experimentally, it is usually found that the work function is a minimum with coverage close to or somewhat less than a monolayer.

It should be emphasized that the above model for ion binding is rather crude. It does not take into account covalent bonding and the polarization of electron shells. Even the concept of the surface charge density is unsatisfactory if we are talking about distances of the order of atomic distances. From this ion model, it is difficult to understand, for example, why Ba and Th are bound more firmly to the tungsten surface than is Cs. Unlike Cs on W, barium on tungsten does not show an increase of heat of adsorption as the substrate vacuum work function increases. This can be seen from Table 2.1 [17].

With the substrate in equilibrium with the adsorbate vapor (for example, Cs or Ba), a fractional coverage is established at which the rate of desorption is equal to the rate of incidence of adsorbate atoms from the gaseous phase (1/4 N_{an a}). In this case, according to (7.1), the coverage is determined by the expression

The work function is mainly a function of coverage: f = f (0).

It follows from (7.4) that the pressure should increase exponentially with temperature to maintain constant coverage. If pressure is kept constant as electrode temperature increases, then a monolayer coverage is at first retained (even to elevated temperature), but then the coverage decreases with increasing temperature until the surface is completely free of the adsorbate. Since the heat of adsorption increases as the coating decreases, the transition from q = 1 to q = 0 usually spans a rather wide range of temperatures. At P = const and with variation of temperature, the emission current varies along the typical S-shaped curves shown in Fig. 2.13. The slope of segment I corresponds approximately to the work function of the adsorbate. The slope of section III corresponds to the work function of the pure substrate metal. In the intermediate zone II, the work function increases as temperature increases. This leads initially to a slowing down of the increase of emission current, but when the effect of changes in 0 begin to dominate the increase of T (in the exponent of (1.7)), the emission current decreases with temperature (the descending segment of the S-curve). The emission current again begins to increase only at a coverage of about 0.1, when the increase of the work function is slowed down as it approaches the work function of the substrate. Segment II is more elongated and smoother (i.e., with a smaller amplitude of current variation) for a polycrystalline surface, as compared to a monocrystal surface, because different patches have different heats of adsorption and different work functions.

8. Surface Ionization [20]

where g_i is the statistical weight of the ion and g_a is the statistical weight of the unexcited atom. It is obvious from

*As the ordered structure of the adsorbed layer varies, the work function may also vary, even with a constant degree of coverage. For example, Cs adatoms on a (111) face of a Ge crystal readily combine into Cs₂ molecules, which form different regular structures on the surface [18]. In this case, the work function increases considerably (up to 1 eV) above that of an amorphous Cs layer. With the formation of Cs₂ molecules on the surface, it is apparently more convenient for the Cs adatoms to take part of their charge from the substrate. However, during adsorption of Cs on the (110) face, the order-disorder transition hardly changes the work function [19].

where E_ea is the energy liberated during formation of the negative ion (electron affinity). Also, for desorption of an atom in an excited state with energy E_k, we have

where g_k is the statistical weight of the state with excitation energy E_k.

To determine the total desorption rate, it is necessary to add the desorption rates for all possible states.

As already noted, the most suitable electrodes for TIC cathodes are monocrystalline surfaces. However, the electrodes of real converters are usually cylindrical; therefore, oriented vapor deposited coatings are now used extensively for cathodes. Tungsten, precipitated out from tungsten chloride or fluoride vapor with a preferred (110) orientation is used most frequently. Such cathodes are rather stable and have a high vacuum work function (4.9-5.1 eV). Because of their grainy, faceted structure, they sometimes exceed smooth monocrystalline specimens in output performance [24].

The main feature of the adsorption system for the TIC is the dependence of the work function on the electrode temperature and the cesium vapor pressure. Therefore, the family of S-curves for a number of cesium pressures (or atom arrival rates) yields the essential information about the system. These families are presented in Fig. 2.14 and 2.15 for polycrystalline W and Mo.

As already noted, in metals with a higher vacuum work function f _o, the energy of desorption is also greater; therefore, at p_Cs = const, the same degree of coverage occurs at a higher substrate temperature. This significantly increases the emission currents and shifts the maximum

Because of the comparatively low heats of desorption for cesium on refractory metals (H » 3.5 - 2.5 eV as q ® 0), the "working" pressures which provide the required cathode emission at the temperatures of practical interest are rather high, in the range of 10^-1 to 10 torr and above.

The descending segment of the S-curves is used for cathode operation. The region of the ascending segment of the S-curves, where the work function is close to minimum, is used for anode operation (T_A » 800 - l200° K).

The curves for the dependence of the work function and heat of desorption on the degree of coverage yield more complete information about the adsorption system. The S—curves, which were discussed with formula (7.4), can be easily constructed from these data. However, direct measurement of the degree of coverage q (or of surface density N) can be made simply only at low substrate temperatures and with non-equilibrium films, i.e., with those films for which the rate of desorption is negligible, and therefore, which may exist for a long time without the incidence of adsorbate atoms.

and the method that measures Auger transitions in the adsorbed atoms have been used successfully to measure coverage. The first method yields good results when the accommodation coefficient equals 1 and when there is negligible diffusion of the adsorbed atoms to the unexposed sections of the specimen (to the other side of the specimen, to the holder, etc.). The degree of coverage is then strictly proportional to the spraying time. The beam is usually calibrated by ion current from a hot filament placed in the beam. The hot filament ionizes the beam by surface ionization.

Determination of the degree of coverage for films in equilibrium with the vapor is possible only at low cesium pressures (l0^-5 to 10^-6 torr) and is a difficult problem.* But the state of the adsorbed film may appreciably affect its properties (and therefore the degree of coverage). Therefore, the values of H(q ) and f (q ) obtained at these low temperatures and pressures are to be extrapolated to the range of temperatures and pressures corresponding to normal TIC operating conditions, This extrapolation should be made very carefully. @

*The most careful measurements of the degree of coverage under these conditions were carried out by Langmuir and Taylor [8].

@In some cases, for example during adsorption of Ba on the (100) face of tungsten [25], variation of the surface properties has been observed upon heating.

The data presented in Fig. 2.17 for the heat of desorption of cesium from the various surface orientations of monocrystalline tungsten were obtained for a layer in equilibrium with a beam of cesium atoms and at a rather high substrate temperature. The dependence of work function of these same orientations on the surface concentration (taking data from field emission measurements) is presented in Fig. 2.18. The S-curves determine the emission current and work function as a function of two parameters: the Cs vapor pressure and substrate surface temperature. In practice, it is very useful to be able to extrapolate the data on the work function obtained at low Cs pressures (that is, when there is no electron and ion scattering in the interelectrode gap) into the range of higher (more practical) pressures and higher temperatures. For such extrapolations, it is desirable, at least approximately, to reduce the two parameters to one.

where P⁰ is some constant (with the dimensions of pressure) determined experimentally for the surface. With ordinary degrees of coverage, 0.2 < 0 < 1, the heat of adsorption satisfies H(q ) » kT/lnq , and it is possible to assume with good accuracy that the degree of coverage q

and the surface work function f ( q ) are dependent on a single parameter kT ln(P_Cs/P^o).

The dependence of f on kT ln(P_Cs/P^o) for polycrystalline tungsten with preferred surface orientation of (110) is shown in Fig. 2.19. The experimental points, obtained from measuring the electron and ion emission at different pressures and temperatures, can be plotted on a common curve with a spread of 0.02—0.03 eV [26]. If we assume that

In practice, the reservoir temperature T_R usually varies within a comparatively narrow range (500 ~ 700° K). Because of this T_R in the brackets of (9.3) can be replaced by some average value, and one may assume that the degree of coverage and the work function of a given surface are universal functions of a single parameter T/T_R The relations f (T/T_R) are called T/T_R plots or Rasor curves [27]. Because of the simplicity of plotting, they have found extensive application in the processing of experimental data, although they should yield a somewhat greater spread of points for different values of P_Cs than the dependence of f on kT ln(P_Cs/P^o) considered above. As an example, the experimental data for the work function of the main faces of a tungsten

monocrystal in Cs vapor are presented in Fig. 2.20a as a T/T_R plot.

where H_int(q ) and f _int(q ) are some functions weakly dependent on the surface properties and determined by the degree of coverage q .

As can be seen from formulas (9.4) and (9.5) (to the approximation given) the surface properties are determined by a single easily measured parameter--the vacuum work function f _o. Different metal surfaces with identical f _o should behave identically. The family of Rasor curves for different work functions, for which H_int(q ) and f _{int( q )} were calculated from experimental data [8], is presented in Fig. 2.20b. These curves agree satisfactorily (within 0.1-0.3 eV) with experimental data for polycrystalline surfaces of refractory metals in the range of TIC operating temperatures and pressures [27]. However, a greater spread, which indicates the necessity of a variation in the selection of H_int(q ) and f _{int( q )} , is observed for monocrystalline surfaces.

Barium has two valence electrons capable of binding with the substrate. The first barium ionization potential is equal to 5.15 eV, the second is equal to 15 eV. Therefore, one may expect that the ion adsorption model for Ba is not appropriate. Actually, unlike Cs, the vacuum work function of the substrate plays a secondary role in the adsorption of barium, and the role of the crystalline structure of the substrate becomes dominant. It is obvious, from Table 2.1 for the heats of desorption of Ba from different tungsten faces, that the (110) face (with the maximum work function) has the worst adsorption capacity (low desorption). The heat of desorption of Ba on the (100) face which has a small work function is more than double that of the (110) face.

Calculated per atom, the adsorbed barium changes the surface work function slightly less than cesium changes it (Fig. 2.21); the maximum decrease of the work function for barium on tungsten is approximately 0.5 eV less than for cesium.

of the function f ( q ) (curve 5 in Fig. 2.21a) was observed at rather high substrate temperatures (» 1000° K) on polycrystalline substrates [29] and also on the (100) face of a tungsten monocrystal [30].*

Unfortunately, there is poor agreement among the results from various authors [17, 25, 29-32] for the function H(q ) for the Ba-W system. Data on the heat of desorption presented for a tungsten monocrystal in Fig. 2.21b were taken from the same investigation [25] as were the corresponding curves of f ( q ) in Fig. 2.21a.

A beam of Ba atoms was used to produce the coverage in most of the barium measurements. A condensed barium phase can be maintained in equilibrium with the vapor only at high temperatures (» 1000° K) because of the relatively low vapor pressure of barium [33]. For an equilibrium measurement, the whole experiment must be maintained at this high temperature. Such measurements have been carried out only recently. Data obtained in [34, 35] for tungsten are presented in Fig. 2.22,

* It is interesting to note that the presence of even small amounts of oxygen during barium adsorption (residual pressure of P_O2 ³ 10^-9 torr) leads to the occurrence of a minimum on the curve f ( q ) and to an additional decrease of the work function.

and the S-curves obtained for molybdenum in barium vapor are presented in Fig. 2.23.

It is obvious from Fig. 2.23 that the smaller change in surface work function compared to cesium (with the same degree of coverage) is more than compensated for by the high barium heat of desorption. Even at P_Ba » 0.01 - 0.1 torr, the maximum emission reaches values of the order of tens of amp/cm², and the maximum of the S-curves occurs in the most favorable range (from the TIC viewpoint) of cathode temperatures (» 2000° K).

In conclusion, consider the problem of the simultaneous adsorption of barium and cesium. The presence of Cs in the gap of a TIC should hardly affect the electrode work function, because (with approximately identical cesium and barium pressures) the surface should be coated mostly by single atoms of barium, owing to the large difference in heats of desorption of cesium and barium. This conclusion has been confirmed by experimental results [34,36]. Even at a barium pressure P_Ba = 4 x l0^-4 torr, the work function remains practically constant over a wide range of P_Cs and begins to decrease significantly only for P_Cs ³ 1 torr.

cesium vapor is to introduce electronegative elements or compounds into the converter: oxygen, halogens, or their compounds with cesium. The joint adsorption of electronegative atoms and cesium increases the cesium heat of desorption and leads to a much greater decrease in the work function, not only of the cathode, but more important, of the anode. The cesium pressure required to obtain the optimum cathode work function decreases accordingly, because the general output characteristics of the TIC are significantly improved [37].

Oxygen adsorption. Oxygen transforms to a chemisorbed state, beginning at very low temperatures. In this case, the first adsorbed layer is bound to the surface of the metal very strongly (Table 2.2 [40]), while the succeeding layers are much more weakly bound. Oxygen adsorption (as well as the adsorption of other electronegative atoms) usually leads to an increase in the work function of the metal surface.

*A typical example of this reaction is the "water cycle" with Mo or W [39]. When water vapor impinges on the surface of a Mo or W cathode, the H₂0 molecule dissociates and volatile oxides of these metals form, which are then desorbed together with the hydrogen. The oxides condense on some colder surface where they are reduced by hydrogen. The atoms of the metal remain on this surface. The resulting water is evaporated and may again impinge on the cathode.

even reduce the work function of the metal [41,42]. Oxides begin to form on the surface (even on oxygen—resistant metals) at temperatures around 700 - 1000° K, and further heating results in evaporation not only of oxygen but of the oxidation products as well. Also, etching of loosely packed planes into densely packed facets can occur for some metals - for example, the (100) and (111) planes of W are etched to (112) and (110) facets [42,43].

A free oxygen layer is completely removed from a metal surface only at very high temperatures - for example, at about 1400° K for Mo and about 2000° K for W. At higher temperatures, the adsorbed layer may exist only with a continuous supply of oxygen to the surface from outside or by diffusion from inside the metal.* These equilibrium oxygen layers on refractory metals increase the substrate work function. Some dependence of the work function of W on temperature at different oxygen pressures is presented in Fig. 2.25.

the metal as temperature increases. These oxygen-saturated metals maintain an adsorbed film for a long time in the absence of an external oxygen source. @

*Complete removal of oxygen from the metal requires heating to about 1700° K for Mo and about 2500° K for w [44].

@Removal of oxygen from the volume requires careful heating at T » 2600° K for Ta [41]. Oxygen is removed from nickel only upon heating in a hydrogen atmosphere or during ion bombardment at T » 1000° K [46].

Among the other electronegative gases, fluorine is most like oxygen in its adsorption properties. The remaining halogens interact more weakly with the metal substrate, have a lower heat of desorption, and alter the work function to a lesser extent [47].

The properties of complex adsorption layers. The role of electronegative atoms in complex coatings is that they increase the effective dipole moment of the electric double layer created by the Cs atoms on the metal surface and increase the binding energy of the cesium to the substrate. It is often assumed that these increases are the result of an increase of the substrate work function due to oxygen adsorption, but this point of view is too simple and does not explain the many features of this complex adsorption system. Most likely, the electronegative atoms and cesium atoms on the substrate surface enter into a complex interaction simultaneously with the atoms of the substrate and among themselves [48].

The change in the work functions of pure and oxidized tungsten is presented as a function of the degree of cesium coverage in Fig. 2.26. It is obvious that even with very small amounts of adsorbed oxygen, which essentially do not affect the substrate work function (curve 2), the contribution of each Cs atom to the reduction of the work function increases significantly. The cesium coverage at optimum (lowest work function) decreases by a third. With large amounts of adsorbed oxygen, which greatly increases the work function over that of pure tungsten, there is only a small further decrease in the optimum coverage, but a marked decrease in the minimum work function itself (curve 3). An increase in the heat of desorption with the presence of electronegative atoms is manifested most clearly in the shift of maximum emission in the S-curve to a higher temperature, an increase by as much as 200 - 300° .

We note that complex compounds consisting of cesium oxide, metal oxides of the substrate, and free Cs atoms may form on the substrate surface at low temperatures. These coatings (typical of photocathodes, for example) have very low work functions (» 1 eV). However, they are rapidly transformed at elevated temperatures; and. their work function increases so that, in the presence of different electronegative additives, it does not decrease below 1.3 - 1.4 eV at the lowest anode operating temperatures (» 700 - 800° K). (See, for example, [49].)

The S-curves are determined by the flow rate of electronegative atoms to the surface, if this flow is constant. However, the most stable and reproducible results are observed if the electronegative gas compounds with cesium are used rather than the gases themselves. These compounds dissociate upon reaching the hot electrode surface so that the final emission result is the same as that for the incidence of an equivalent rate of the gases themselves.

Data for adsorption of cesium oxide on the most promising cathode materials, Re and W [50], are presented in Fig. 2.27. It is obvious from the figure that an increase of the cesium oxide flow to the substrate (determined by the temperature T_Cs2O of the oxide solid phase) is accompanied by a decrease in f only to a specific limit (with constant substrate temperature T). A further increase in the cesium oxide flow leads to an increase in the work function. The maximum decrease in the work function is somewhat greater for Re and reaches 0.3 - 0.5 eV for the conditions investigated, which corresponds to an increase of emission by approximately an order of magnitude.

An even greater increase of emission is observed with the introduction of fluorine [51]. There is also evidence that the emission from metals in cesium vapor increases with the addition of such gases as CO and H₂.

Metals which easily absorb oxygen (Ta, Ni and Cu), once subjected to its effects, yield a significant (and more stable) increase of emission in cesium vapor without an external oxygen source. In this case, the presence of low oxygen pressures has no effect on the emission, because the main input of oxygen atoms comes from inside the substrate. Changes in the S-curves, which in this case may have complicated appearance [52], is observed only with comparatively large additive concentrations. This effect is weaker when the substrate is W, Re, or Mo.

When cesium is adsorbed on a surface with a free oxygen layer,* relatively stable coatings may be formed only with a preliminary heating of the oxygen layer, to remove excess oxygen. This surface yields stable and reproducible S-curves, if the surface temperature is not raised, and if the surface is not subjected to prolonged ion or electron bombardment. The emission for cesium adsorbed on these coatings may exceed by orders of magnitude the emission for cesium adsorption on the pure metal. @

The presence of electronegative atoms is especially significant because of their effect on the anode work function. In cesium vapor, even a comparatively small amount of electronegative gases (primarily of oxygen) which is inevitably present in a real converter, reduces the anode work function to values of 1.5 - 1.6 eV. The anode work function may decrease to 1.3 - 1.4 eV with special introduction of electronegative components. The role of electronegative additives and of residual gases in converters filled with Ba is even more important, since in the absence of these additives, the anode has the work function of pure barium:

f _Ba = 2.4 eV,

@For example, heating of the oxygen layer on the tungsten to 1600° K leads to an increase of emission in the Cs vapor by more than a factor of 2 [53].

TIC reduces to the solution of the Langmuir problem for the passage of current through a diode [54,55] and is outlined in detail in [56-68].

Consider also a TIC with plane-parallel electrodes. This assumption is appropriate even for coaxial TICs since the radii of the electrode curvatures are always very large compared to the interelectrode gap. Assume also that the anode temperature T_A is rather low, so that the electron emission from the anode can be ignored.

When current passes through this type of TIC (if the potential difference between the anode and cathode is small), the space charge of the electrons creates a potential barrier in the interelectrode space, which retards electrons emitted by the cathode (Fig. 2.28). Only those electrons which are capable of overcoming the potential barrier* pass to the anode, and the current j through the converter will be related to the emission current j_S° by an expression similar to (1.7): Huh?

Here V_m is the potential height of the barrier (referenced to the cathode).

When a potential difference is applied across the vacuum space, the potential barrier V_m will decrease gradually and the current will increase. When the applied potential increases so much that the resulting field at the cathode reaches zero [E(0) = 0], the barrier disappears, the current through the converter reaches the cathode saturation current

Application of an increasing electron-retarding field also eventually leads to the disappearance of the potential extreme in the interelectrode space. When this occurs the current is directly related to the value of the retarding voltage ô V_dô :

To plot the characteristic in the intermediate region (which naturally is of primary interest, because it contains what corresponds to the usual operating point of the TIC) it is necessary to establish the relationship between the value of the barrier I’m? and the voltage across the gap V_d. This relationship will be a function of the TIC parameters T_C, j_S° , and the size of the gap, d. We begin by considering the potential distribution of the interelectrode space.

*For convenience, instead of the ordinary potential y , consider the potential for a single negative charge V = -y ("the potential for an electron")--see Fig. 2.28.

for the range 0 £ x £ x_m,

for the range x_m £ x £ d

where f_m(n ) is the Maxwellian distribution function using the cathode temperature.

When solving this problem it is convenient to introduce the dimensionless potential (Huh?) y , referenced to the point of the maximum potential:

The solution is then expressed in the form of a special function (Huh?) y (x ), which is presented in Fig. 2.29. For given values of T_C, j_s^o, and d, this function makes it possible to find the potential distribution between the electrodes for any value of conduction current j. To do this it is necessary, having been given the value of current, to calculate in sequence V_m (from formula (12.1)), (Huh?) y ₀ = eV_m/ kT, x ₀(y ₀) (from the graph in Fig. 2.29), and x_m = -x ₀/c. By knowing V_m and x_m, and again by using the graph of y (x, the value of the potential V(x) for any point of the interelectrode gap 0 £ x £ d can be found.

By calculating the value of V(d) = V_d for a number of values of

converter current in this manner, it is possible to plot the desired function j = f(V_d) over the range where a potential maximum exists, and therefore, to plot the full current-voltage characteristic of the TIC for given parameters T_C, j_s^o, d and f _C.

It can be shown that the function f(V_d), expressed in relative values of j/ j_s^o and eV_d/ kT_C, is clearly specified by the dimensionless parameter

The family of these curves, calculated for different values of R, is presented in Fig. 2.30. In this case the current—voltage characteristic for the given TIC parameters is the curve for the corresponding value of R, shifted toward the region of negative (operating) voltages by an amount equal to the contact potential different D V = (f _C - f _A)/e *.

The limiting case of R ® 0 corresponds to the absence of space charge, and therefore, for this case the characteristic is an exponential up to the value of V_d = 0, where the current becomes constant at the saturation value. For R > 0, the total emission current is removed upon application of a potential difference V_d (>0), which increases as R increases, and a deviation of the characteristic from the above exponential begins at ever lower retarding potentials V_d.

It is obvious from Fig. 2.30 that no appreciable decrease of current by space charge occurs for R £ 10. It is easy to estimate the maximum interelectrode gap permitted for efficient operation of TICs in the vacuum mode, using this critical value. By assuming T_C » l500° K and j_s^{o »} 10 amp/cm² (these figures correspond to the TIC output of practical interest, of the order of several W/cm²), at R = 10 we find d » 2.6 x l0^-4 cm, a distance of the order of the Debye length L_D.

Values of the interelectrode gap for TIC operation in the vacuum mode are shown in Fig. 2.31, where the calculated values of the nominal output are presented as a function of [58]. The effect of the contact potential difference of the cathode and anode on the TIC output is also clearly evident from Fig. 2.31. Since the cathode work function cannot be too high (to allow sufficient emission current), an anode with an even lower work function is required, to allow a significant voltage output. But a lower limit is set to the anode work function (according to the temperature), so that the electron emission from the anode remains low compared to that from the cathode.

The occurrence of excessive anode emission leads to a decrease in the output power of the TIC both as a result of reverse electron current (from the anode to the cathode), and as a result of an increase of the space charge, which reduces the electron current from the cathode [59].

*It follows from Fig. 2.28 that V_o = V_d - D V.

b, L. van Someren, Thermionic Conversion Specialists Conference, Carmel, p. 18 , 1969,.Thermionic emission in an accelerating field. As the electric field increases and all the patches become exposed , Fig. 2.8c,, b approaches 1. This occurs when the applied field becomes approximately equal in intensity to the patch field, E = D f /ea , where D f is the work function difference of adjacent patches and a is their linear dimension,. Thus, the value of the external field at which transition to the normal Schottky effect occurs , b = 1, is determined by the specific surface properties of the emitter and may vary over a wide range-from tens to 10⁴ V/cm Language, English , eng,, Call Number, TK2955 .T4412 1969, LCCN, 88197004, Dewey Decimal, 621.31/243, ISBN/ISSN, -

7. T.V. Krachino and T.L. Matskevich, Radio Eng. Electron. Phys., 14, 1660 , 1969,.Thermionic emission in an accelerating field. As the electric field increases and all the patches become exposed , Fig. 2.8c,, b approaches 1. This occurs when the applied field becomes approximately equal in intensity to the patch field, E = D f /ea , where D f is the work function difference of adjacent patches and a is their linear dimension,. Thus, the value of the external field at which transition to the normal Schottky effect occurs , b = 1, is determined by the specific surface properties of the emitter and may vary over a wide range-from tens to 10⁴ V/cm, Language, , Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

13. G.M. Rosenblatt and Pang-Kai Lee, J. Chem. Phys., 49, 2995 , 1968,.Evaporation. Even smaller values of a , » 10^-4, have been measured for the , 111, face of a monocrystal of arsenic, where evaporation occurs in the form of As₄ molecules, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

14. W. Winterbottom, J. Chem. Phys., 51, 5610 , 1969,.Evaporation. Complex formation is more probable in a deep opening or in an effusion chamber than during vacuum evaporation from a smooth surface. Thus, during evaporation of NaCl, the concentration of the dimer Na₂Cl₂ increases as the depth behind the aperture increases, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

15. Yu. A. Logatchov and B. Ia. Moyzhes, Surface Science, 17, 504 , 1969,.Surface energy. The dipole moment of the double layer near the surface of solid solutions of ionic crystals was calculated in, but the corresponding calculations have not yet been carried out for metal alloys, Language, English , eng,, Call Number, QD506 .S8, LCCN, , Dewey Decimal, , ISBN/ISSN, 0039-6028

16. Ya. I. Frenkel’, Statisticheskaya fizika, izd-vo AN USSR , 1948,.Adsorption and its effect on the work function. Frenkel formula, Language, , Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

17. U.V. Azizov and G.N. Shuppe, Sov. Phys. - Solid State, 7, 1591 , 1966,.Adsorption and its effect on the work function. Electropositive additives in TICs., Language, , Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

18. R.E. Weber and A.L. Johnson, Journal of Applied Physics, 40, 314 , 1969,.Adsorption and its effect on the work function, Language, English , eng,, Call Number, QC1.J83, LCCN, , Dewey Decimal, 530.5, ISBN/ISSN, 0021-8979

19. A.G. Fedorus and A.G. Naumovets, Surface Science, 21, 426 , 1970,.Adsorption and its effect on the work function, Language, English , eng,, Call Number, QD506 .S8, LCCN, , Dewey Decimal, , ISBN/ISSN, 0039-6028

20. E. Yu. Zandberg, N.I. Ionov, Poverkhnostnaya ionizatsiya, "Nauka" , 1969,.Surface ionization, Language, , Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

21. J.M. Houston and H.F. Webster, Adv. in Electronics and Electron Physics, L. Martin, ed. vol. 17, Academic Press, New York, p. 125 , 1962,. Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

22. a, R.L. Aamodt, I.J. Brown, and V.D. Nichols, Journal of Applied Physics, 33, 2080 , 1962,. Language, English , eng,, Call Number, QC1.J83, LCCN, , Dewey Decimal, 530.5, ISBN/ISSN, 0021-8979

b, W.A. Ranken, et. al., Advanced Energy Conversion, 3, 235 , 1963,. Language, English , eng,, Call Number, TK2896 .A5, LCCN, , Dewey Decimal, , ISBN/ISSN, 0365-1789

23. L.W. Swanson and R.W. Strayer, J. Chem. Phys., 48, 2421 , 1968,. Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

24. a, P. Batzies, P. Schroder-Bobo, and G. Wahi, 3rd Int. Conference on Thermionic Electrical Power Generation, Jülich, p. 845 , 1972,. Adsorption of cesium on refractory metals,Language, English , eng,, Call Number, TK2955 .I58 1972, LCCN, 75313167 //r893, Dewey Decimal, 621.312/139, ISBN/ISSN, -

b, J.C. Danko and G.W. Titus, Thermionic Conversion Specialist Conference, Miami Beach, p. 82 , 1970,. Adsorption of cesium on refractory metals, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

c, J. Dunlay and R. Meyers, ibid, p. 86 Adsorption of cesium on refractory metals,

d, R.J. McCandless and D. R. Wilkins, ibid, p. 227 Adsorption of cesium on refractory metals,

53

25. Yu. S. Vedula, et.al, [24a], p. 135.Adsorption of cesium on refractory metals, Electropositive additives in TICs,

26. G.N. Hatsopoulos and F. Rufeh, [6b], P. 1.Thermionic emission in an accelerating field,

27. N.S. Rasor and C. Warner, Journal of Applied Physics, 35, 2589 , 1964,.Adsorption of cesium on refractory metals, Language, English , eng,, Call Number, QC1.J83, LCCN, , Dewey Decimal, 530.5, ISBN/ISSN, 0021-8979

28. R.J. McCandless and D.R. Wilkins, [24b], p. 210.Adsorption of cesium on refractory metals,

29. G.E. Moor and H.W. Allison, J. Chem. Phys., 23, 1609 , 1955,.Electropositive additives in TICs, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

30. a, Ya. P. Zingerman, V.A. Ischuk, and V.A. Morozovskii, Sov. Phys. - Solid State, 2, 2030 , 1961,. Electropositive additives in TICs, Language, , Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

b, Yu. V. Zubenko and T.L. Sokol’skaya, Sov. Phys. - Solid State, 3, 1133 , 1961,.Electropositive additives in TICs, Language, , Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

31. B.V. Bondarenko, E.D. Konovalov, and E.A. Tishin, Sov. Phys. - Solid State, 14, 3078 , 1973,.Electropositive additives in TICs, Language, , Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

32. H. Utsugi and R. Gomer, J. Chem. Phys., 37, 1706 , 1962,.Electropositive additives in TICs, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

33. R.E. Honig and D.A. Kramer, RCA Review, 30, 285 , 1969,.Electropositive additives in TICs, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

34. J. Psarouthakis and W.J. Levedahl, Proc. 1st Int. Conference on Thermionic Electrical Power Generation, London , 1965,.Electropositive additives in TICs, Language, English , eng,, Call Number, TK2950 .I57 1965, LCCN, 68103750 //r862, Dewey Decimal, 621.312/139, ISBN/ISSN, -

35. I.G. Gverdtsiteli, et.al., [9], p. 1097.Electropositive additives in TICs, Thermionic emission in an accelerating field,

36. V.D. Bondarenko, A.I. Loshkarev, and B. Sh. Ul’masbaev, High Temp., 8, 201 , 1970,.Electropositive additives in TICs, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

37. D. Lieb, F. Rufeh, [24b], p. 471.Adsorption of cesium on refractory metals, Electronegative additives in TICs,

38. R.L. Wagner, [6b], p. 411.Thermionic emission in an accelerating field,

39. a, R.L. McKisson, Advanced Energy Conversion, 3, 137 , 1963,.Electronegative additives in TICs, Adsorption of cesium on refractory metals, Language, English , eng,, Call Number, TK2896 .A5, LCCN, , Dewey Decimal, , ISBN/ISSN, 0365-1789

b, G. Wahl, J. Demny, and P. Batzies, [24a], p. 896.Electronegative additives in TICs, Adsorption of cesium on refractory metals,

40. B. McCarroll, J. Chem. Phys., 46, 863 , 1967,.Electronegative additives in TICs, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

41. a, O.L. Fehrs and R.E. Stickney, Surface Science, 8, 267 , 1967,. Electronegative additives in TICs, Language, English , eng,, Call Number, QD506 .S8, LCCN, , Dewey Decimal, , ISBN/ISSN, 0039-6028

b, B.J. Hopkins, C.B. Williams, and P.C. Wilmer, Surface Science, 25, 633 , 1971,.Electronegative additives in TICs, Language, English , eng,, Call Number, QD506 .S8, LCCN, , Dewey Decimal, , ISBN/ISSN, 0039-6028

42. A.E. Bell, L. W. Swanson, and L.C. Crouser, Surface Science, 10, 254 , 1968,.Electronegative additives in TICs, Language, English , eng,, Call Number, QD506 .S8, LCCN, , Dewey Decimal, , ISBN/ISSN, 0039-6028

43. J.C. Tracy and J.M. Blakey, Surface Science, 13, 313 , 1969,.Electronegative additives in TICs, Language, English , eng,, Call Number, QD506 .S8, LCCN, , Dewey Decimal, , ISBN/ISSN, 0039-6028

44. S. Dushman and J.M. Lafferty, Scientific Foundations of Vacuum Technique, 2nd ed., John Wiley and Sons, New York , 1962,.Electronegative additives in TICs, Language, English , eng,, Call Number, QC166 .D85 1962, LCCN, 61017361 /L/r962, Dewey Decimal, 533.5, ISBN/ISSN, -

45. P. Batzies, [9], p. 1357.Electronegative additives in TICs, Thermionic emission in an accelerating field,

46. J.W. May and L.H. Germer, Surface Science, 11, 443 , 1968,.Electronegative additives in TICs, Language, English , eng,, Call Number, QD506 .S8, LCCN, , Dewey Decimal, , ISBN/ISSN, 0039-6028

47. C.W. Jowett and B.J. Hopkins, Surface Science, 22, 392 , 1970,.Electronegative additives in TICs, Language, English , eng,, Call Number, QD506 .S8, LCCN, , Dewey Decimal, , ISBN/ISSN, 0039-6028

48. J.M. Chen, Journal of Applied Physics, 41, 5008 , 1970,.Electronegative additives in TICs, Language, English , eng,, Call Number, QC1.J83, LCCN, , Dewey Decimal, 530.5, ISBN/ISSN, 0021-8979

49. T. Alleau and M. Bacal, [24b], p. 434.Electronegative additives in TICs, Adsorption of cesium on refractory metals,

50. R. Langpape and A. Minor, [9], p. 1367.Electronegative additives in TICs, Thermionic emission in an accelerating field,

51. R. Langpape and A. Minor, [34].Electronegative additives in TICs, Electropositive additives in TICs,

52. R.G. Wilson, Surface Science, 7, 157 , 1967,.Electronegative additives in TICs, Language, English , eng,, Call Number, QD506 .S8, LCCN, , Dewey Decimal, , ISBN/ISSN, 0039-6028

Industrial Engineering Chemistry

53. I. Langmuir, Industr. Eng. Chem., 22, 390 , 1930,.The vacuum mode of TIC operation, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

54

54. V.R. Bursian, ZhRFKhO, 51, 289 , 1921,.The vacuum mode of TIC operation, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

55. L.N. Dobretsov and M.V. Gomoyunova, Emission Electronics, Israel Program for Scientific Translations, Jerusalem , 1971,.The vacuum mode of TIC operation, Language, English , eng,, Call Number, QC721 .D5913, LCCN, 71181530 //r872, Dewey Decimal, 539.7/2112, ISBN/ISSN, -

56. A.I. Ansel’m, Termoelektronnyy termoelement, izd-vo AN USSR , 1951,.The vacuum mode of TIC operation, Language, , Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

Journal Electronics

57. H. Moss, J. Electronics, 2, 305 , 1957,.The vacuum mode of TIC operation, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

58. H.F. Webster, Journal of Applied Physics, 30, 488 , 1959,.The vacuum mode of TIC operation, Language, English , eng,, Call Number, QC1.J83, LCCN, , Dewey Decimal, 530.5, ISBN/ISSN, 0021-8979

Journal Electronics and Control

59. P.A. Lindsay and F.W. Parker, Brit. J. Electronics and Control, 7, 289 , 1959,.The vacuum mode of TIC operation, Language, English , eng,, Call Number, , LCCN, , Dewey Decimal, , ISBN/ISSN,

Hosted by www.Geocities.ws