II17AT O F .IDSOIIPTIOS O F C'ARBOS MOSOSIDE O S COPPER C'-ITALTST
-_-_
BY RALPH A L O S Z O B E E B E
I n an earlier communication, the n-riter in collaboration with H. S, Taylor1 described a method for the direct nieasureiiient of heats of aclsorption. Both the Integral and differential d i m of Qqthe hezt of adsorption per niol of ndsorbed gas, n-ere given for hydrogen on n nickel adsorhent. The therinnl
TIC. I
cc nd.s?rl,cd effect was found to vary in a surprising nianner according t o the past history of the cataj-st. h recent papcr liy B. Foresti' made this point a particularly important one. The follo1ving measurements of Q for a totally different gasmetal combination shollld, therefore. he presented at this time. Preparation of the Catalyst .-The zdsorhent sample vias prepared 11y reduction of copper oxide granules in a streem of pure, dry hydrogen at I;o'180' C, reoxidation in a stream of oxygen at the same temperature and subsequent reduction by hydrogen a t a temperature that vas kept a t all times 1
Beebe and Taylor: J. -Am. Chcm. Soc., 46,43 (1924). 185 (1925).
* B. Forrsti: G R Z Z .45, ,
HEAT O F . i D S 3 R P T I O S O F C . I R B O S M O N O S I D E O S C O P P E h
I539
below 135' c'. The 5;g sample thus prepared ndsorlied 11.5 cc of carlion inonoxide and 6.0 cc. of hydrogen at ooc-'and 760 min. pressure. -After iiieasiireinents of Q hatl been completed for the catalyst in this state. n-hich n-ill he called the "active" sample, due t o some mischance, the catalyst became poisoned so that its adsorptive capacity for carlion nionoxitle n-as cut don-n t o approximately 7.0 cc. nt o°C and 760 iiiin. The catalyst in this reduce(; adsorptive capacity will lie designated as the 'Lpoisonetl" samp!e. l-nfortunntely. the writer is unable t o state the cause of the poisoning. -1compnrison of the isotherms I antl I1 in Fig. I for adsorption of C'O liy tlie catalyst in tlie **active"antl "poisonetl" statcs n-ith the siinilnr pair of c i i r w s olit:iinec! by Pc,o.;el for a cop1;er catalyst liefore antl after poimninp liy mercury. n-oultl 1e:itl one to attriliiite the poisoning in the present instance t o a siinilar cauye. The obeerration,s recorded in this paper were coin!:letecl in ,Jiine 102;. The niitlior hoped at that time to repcat the mensiirements for a czt::lyst poisonet1 under controlled conditions. 1-iiforeseen interrupticns h:;w nintle this inipossilile rliiring the past !-ear. The striking ant1 a1;pzrently acoi11:iloii~ nr.ture of the data olitnined, lion-ever, make it atlrisnlile to present them n-ithOUT irirther { I c k y . DiscusZion of Es!:eriiiient::l Method.-The method tlescrilietl in ihe ewlier pa1;er as employed in this research. The teml:erntiire of eraruntion wis 140' t o 1 j 0 " C ' . -1s ~ r i pointed s out therein for tlie care of hytlropen an(l col:l;er, ;liere IT-SS 1;r~cticallyno tlrift in tlie cc!sorprion of cnrlion monoxitle I)>- the copper. This iiicc!e for very relialile ixlues. since all the heat ~ v a q lil?crateil :Iltiiost instantaneously. In the memurenient of tlie differentid \-'dues it TKIS fou~itlpossible to ohtnin tlie thei~innl(jntn for much siii~Jle1~ increments of s,the ~ o l i ~ m 2tleor13etl. c t lian had prei-iously lieen coiisitkrec! prncticnhle. s.iate(l in the discussion of the method it was necessary t c 1lialie a correction for tlie thermal effect accruing from (a) compression of the gas in entering the cxlorimeter. (11'1 cooling of tlie gas froin rooiii teiiiperatiue t o 0°C'. I n deterinin-tion of integral values. TT-lien the final pressure of gas over the catalyst is 760 111111: this correction is an important one. T o cite an a c t u d case. in Determination I , Tali!e I. the true temi?erntureriFe~~--aso" .j.oC', the correction for effects (a) and (11) n-as 0.~07;. hence tlie rise due t o the aclsorption was 0 . ~ 4 4 5 . There v a s , therefore, a I ; correction. S o t e that in this case effects (a) and (11) for tlie 4 2 . 5 3 cc CO occupying the free space elioi-e the catalyst hatl to lie considered 2.s n-ell as for the 11.04cc ntlsorhecl. The correction Iiecaiiie of minor importance, however. in ohtaining tlie differential d u e s j since as is evident from Curve I, Fig. I , the pressure of CO in the gas phase in equilibriuni with as much as 7 . 0 . of the total 11.0 cc which is cdsorlietl at 760 nim. is only 6.0 iiiin, i. e. there is practically no correction for gas to fill the free space of the chamher. In Det. 4 C the true temp. rise was o."14j, the correction vias O".OOT, antl hence theteinp. risedue to adsnrption was 0 . ~ 1 3 8 .Here thz c3rrection isonlya 3 y o n e . I t hasheenfound possible to ohtain con tent results for increments as siiiall as 1.0cc. for the Pease: J. =Im. Chem. SOC.,45.
2300
(19231,
1540
RALPH ALONZO BEEBE
first j . 0 cc of the total 11.0 cc adsorbed, i. e. while equilibrium pressure is still lo^. As the isotherm shows, the equilibrium pressure increases rapidly for amounts adsorbed beyond j . 0 cc, and the correction again becomes important. The differential values of Q are, therefore, less reliable for these last portions.
Results.
TABLE I Integral Values for Heat of -Adsorption of CO on Copper Sample
Determination
“Active” Copper “Poisoned” Copper
Q
1 (T-T,! Cc (c-760) True rise in adsorhed by sample Temp.
Heat, of adsorption in calories per mol
I
11.04
0.445
12,200
2
0,405
11,200
3
92 I 1 .o; 7.10
0.423 0.358
I I , 700
I
IO.
I5doo
TABLE I1 (Q at Partial Pressures)
Differential T-alues for CO on Copper Sample
Determination I
A
In
,1 2B 3 X 2
“Active” Copper
0
74.9 2 . 2
2.2
j8.I
0
3 =
3. I
4 -4 4B 4c
0.8 1.8
A 2 A 2B 2c
6
0
0.j
I
Final P in m m Hg
6
3B 3C
4D 5 -1 6 -4
“Poisoned” Copper
Initial P in mm Hg
1.9
_ -
0.
j
1.9 5.7 jj.6
0 8 1 8 8.8 8.8 1 4 0 . 8 0
> .O
< I . O
> o
