li""'KY ('(HU'Us. Mg"'l Vul 26. }ICI. 2. pp. PrIl>1ed in G"",l Rnla,,,

249- ~52.

OI%-II!1()4/1I6 S.1 00 + 0 00

198(,

""f&"n",n

Jilumaill lId

THE ORlGIN OF THE ELECTRIC CURRENT IN FU'S EXPERIMENT- AN ANAL YSIS OF FU'S WORK ON REALlZING MAXWELL'S "DEMON" YU-ZHJ WANG Ch~ngdu

loSlllUte of Radio Fngincermg. Chengdu. Sichuan.

Peo pl~'s

Repuhlic of Chi na

(Recl·i!'e" ( May 19H5) Abstraet- In 1982. Fu presented his wOfk o n lh~ theore!Îcal and ~xperi mcnta ! re~llizali on or Maxwell's "demon". [1] Fu dcclared Ihal a perpetuai cu Trcnt was ohwÎ uoo in hi~ de\"il:e, This means Ihat a new maehme of the second kind of perpclUal motIon has been lrI\"CnIN. This JliIpeT prcscnt$ a detal lai analysis of Fu's ....·orl.:. and the followmg conclusions have heen oblamed. (1) ln Fu's cltpenmenl. the objcc t eon lrolled br tne magnelic field was clcctrons in muss. nOlllldlvidual clcctrons. The magnetie field did nOI r.eparale electrons individually into a hlgh-spœd group and a low-spced group. For Ihi~ rea~on, il was not a "demon". The pre
250

WANG:

FUS WORK ON MAXWELL'S DEMON

eleclric cncrgy mille from heat, and was proof of the possibility of transfonning heal from a single thermal source iuta work withoul any deleterious effcct, and thus violulcd the second law. This mcans Ihal the proccss bctween the IWO emiltcrs was rcally li process for the cntropy ta decrease. If SO, one can imagine Ihat, if the whole devicc (including the exlernal circuit and the galvanornctcr which was ~upposed to he heal rcsistant) \l'cre brol:ght 10 a high temperature star, a perpetuaI clcctric currcnt would be obtained wi/hout supplying any heater power. This would really he a perpetuaI motion machine. In the following paragraphs, an analysis of Fu's work is presented and the Tcal origin of the current determined.

ANALYSIS OF FU'S WORK

Firstly, we have ta distinguish the essential difference between Fu's meehanism and Maxwell's meehanism. Apart from eontrolling e1eetrons instead ofmolecules, there was another important difference, i.e. Fu's mechanism did not separate electrons individuaBy, did not divide and gnher electrons into a high-specd group and a low-speed group. What was controlled by the magnetic field were e\cctrons in mas:>, not individually. Thercfore, the magnetil: field did not possess the function of the wntrolling service as the "demon"' does. In Fu's device, Ihe flowing of e1ectrons was eontinuous, nol a cumulative process whieh is necessary in Maxwell's deviœ. [t is obvious that the process in Fu's devicc was one including the motion of a huge number of small particles, in spite of the fact that the current was a mere 0.35 pA. This just enters the field ruled by thermodynamies. The above consideration leads the author to believe that the eunent flowing in Fu's device was not the "Maxwell eurrent". Essentially. the prob\cm whieh has 10 he c1arified is: What was the origin of the eurrent? To determine the origin of the eurrent, t he firsl thing of importance is that the cathode accepting electrons (8 in Fig. 5 of Fu's paper [1 )) need not possess the ability 10 emit electrons. Therefore, R can he rcplaced by a common metal plate. Seeondly, we can displaee B, which is a metal plate, now to the top-side of the cathode, and a parallel-plate vacuum diode is achieved. Now, when A is heated, a small eurrcnt should build up in the circuit, provided the distance between the two plates is short enough . This is aetually, as is well-known. the initial vclocity current of a vacuum diode, ils value being approximately several hundred lI A. This fact is wdl-known to all vacuum tube enginecrs. The initial current cornes from the tail part of the velocity distribution curve for the free elcetroll5 in metals, i.e. the part having maximum velocilY . Actually, in the modern vacuum tube indu~try, a type of heat-·eleetrie eonverter, called the thermo-

ekctron energy converter, which ulilizes thermionie emission has been developed (0 transform heat into c1ectrical energy. h is reported that Ihere are tens of energy converters of this type used in satellites. In comparison with solar cells, its advantage is the avoidance of using a direction regulator. Now the whole problem bceomes the following: If a vaeUllln diode or modern hcat-·electric converter, ineluding its external circuit, brought to a high lemperature star, does the thermionic initial velocity eleetron eurrent appear?

This prohlem hears close relation to the thcrmoelectric phenomenon in mctals or semicondw.:tors. [t can bc clarified, stcp by step, as follows . Thermocleetrie phenomena arc of Iwo different kinds, namely, the Peltier clreet and the Thomson effel:!. The Peltier clTeet has to have two mdal wires of di ~ kinds, connected together, the two joining points being at different temperatures, resulting in an eleetrie eurrent in the circuit. As the eurrcnt cireulates, the high-tempcrature point ahsorbs heat while the low-tempcraturc one desorbs, so as to keep their temperatures constant. Sinee the desorbing amount is smalkr than the absorbing amount, the net dilTerence changes to electrical encrgy whidl i, able to do work or transform into joule-heat. Il was apparent that the Peltier etfeet strietly obeys the second law: it needs both a high-temperature heal tank and a low-tempcrature one. If not, the lowtemperature junetion should raise its tcmpcmture, while the high-tempcrature one should deerease. At las!, the tempcratures of both j unetions tend to lx: equal, and the eurrent vanishes. Aeeording 10 the cleetron theory of llletals, the cause of a thermocouple e.m.f. is due to the differem;e in the number of frce clectrons exehanged, whil:h depends on the 10l:al tcmperature, in the two wires whjl:h have different cleetron l:oneentrations. Now, we centre our attention on the case of one eonduetor only If there is a tempcrature differenee at its two points, the free elcl:tron wnœnlrations should be ditferent, and so a thermo e.m.f.