COORDISATION
COMPOUL-DS OF BORON TRICHLORIDE
1253
(10) JOSHI: Current Sei. 14, 67 (1945). (11) JOSIII: I'roe. Indian .lend. Sei. A22, 225 (1945). (12) J O > HPIr:o c . 1ndi:in Sei. Congr., Phys. S e c t . -4listracts 26 (1916). (13) ,JOSHI: Current Sei. 16, 19-31 (1917). (14) JOSHI:P r o c . Indian Sci. Congr., Phys. Sect. Abstracts 25 (1947). (15j JOSIEI ASD DESIIJILXH: S a t u r e 147, SO6 (1941). (16) JOSHI . i s 1 1 L ~ DProc. : Indian Sci. Congr., Phys. Sect. Abstracts (1945). (17) SARAYAS S J ~ ~ Proc. I I : Indian Sci. Congr., Phys. Sect. Abstracts 28-9 (1946) (18) TIITARI: I'roc. Iridian Sei. Congr., Phj-s. Sect. -Abstracts 34 (1916).
CO~RL)ISA~TIO CO S M P O P J X OF H O R O S TRICHLORIDE. 1-1
THE S T ~ T E MPHOSGESC-BOROS S TRICHLORIDE ASD PHOSGEXE-ROROX TRIFLCORIDE DOSALD RAT LIARTIS
.4m
J O I I S PHILIP FAI'ST
-1-oges Choii i c u 1 Labai ator 11, T-nivei sit ij i ~ Ill? j no?s , I'rbana. 111%nois
IZcceived .I7ovember 10, 1948
O x er forty year> ago 13ttud (I) reported the existence of molecular compounds betn een phosgene arid aluminum chloride. Germann and Jersey (8) later reported that boron trifluoride and phosgene formed some compounds Iiut gave neither empirical formulas nor names for these conipounds. Thev pobtulated that boron trichloride also might be soluble in phosgene and form conipounds v i t h it. OTring t o the similar electronic structures of aluminum chloride and the boron halides, it is logical t o expect the latter t o form molecular compounds similar t o those formed by aluminum chloride. The boron halides are known t o bc good acceptor molecules ant1 t o form many coordination compounds ( 2 , 10, 13). With phosgene, it may be possible for either the oxygen or the chlorine atom t o act as a donor. Inasmuch as other molecules containing each of these atoms have been found t o be donors t o boron halides, it seemed likely that phosgene 11 ould form such compounds with both boron trichloride and boron trifluoride. I t was of interest t o investigate the phosgene-boron trifluoride system also hecause Rroir-n, Schlesinger, and Burg ( 3 ) reported that phosgene and boron trifluoride did not coordinate at temperatures as low as -120°C. The apparatus and procedure employed in these investigations h a w been descrihecl in earlier publications (3, 4,7 , 11, 14) except for t n o change>. Owing t o the large amount of current required t o operate the necessary relays for the controls of the automatic fractionating column, the u-ire contact in the control manometer becomes fouled and the mercury smuts the walls of the manometer. w t h the result that the contact sometimes fails. Therefore, a vacuum-tube circuit with thyratron tulles' is used noTv t o operate the ielays which in turn 1
General Electric T h v r a t r o n Tube S o G-57,
1256
D O S A L D R A T K L R T I S A S D J O H S PHILIP FAUST.
YI
actuate ( a ) the motor t o pump refrigerant into the condenser of the fractionating column and ( 6 ) the automatic stopcock. The valves required t o operate the automatic stopcock, which were constructed from bicycle valves and telegraph iounderi as previously reported (1l ) , have been replaced with solenoid valves such as are used in electric refrigerators.? These permit iz larger flow of air than those made from bicycle valves. thus insuring a more rapid opening and closing of the stopcock, \vliich in turn niaintains a more constant pressure in the distillation column. The phosgene iiqed in these investigations vas obtainrtl from the llathieson Chcmical Company of Eagt Ruthcrford. Sen- ,Jersey, and ivas piirifietl 1))- frac-
1 OiK) 0 945 ( e ) * ( 1 SO4 ( 3 )
0
so7
(cs
0 801 ( a ) 0 743 (esi 0 . G O l (1)) 0 654 (esr 0 599 (1)) 0 548 (fsl 0 503 ( f S l 0 495 (hi 0 498 (ds) (1 448 (15 .---
tionul (listilla1ion. 'h( ii*eczirig , 1)olrii \ oi q m ~ i ~ c (1151 l y illrtl samplc's I\ touid l o I)c> - 132.5"C. =t0.3'. 1lie 1)oron tiichluritk u d iri t h y invc4igatiori \\:I< oGtaincd &A) from thtl I\I:ithlebon Clirmicul c'ompnny and n as similarly purified hy frtwtional (list illation. The fiwzing point\ ot q m a t C 1 y ctistillcrl wiiplr.+ I\ ere found t o be -107.3"C. i 0.3". l'hc 1)oron trifliioi ide I\ u- obtained from the Harshan- Chcmical Company of Clcvcland, Ohio, and \vas puntied by tract ional distillation. Separately tliitilled samples of this g a ~Tvere found t o freeze at -127.6"C. f 0.3". O\ving t o the tact that boron trichloride and phosgene each boil above ice temperature, the calibrated flaslcs used t o establish the mole fractions were cxrcb
, 7
+ulciioitl
~ . , 1 1 \ c ,t!
pe SV-11, lion1 t h e € I ~ I I IT7nl\.c V Comp:lr:\ , C'h~c,igo,Illinois
COORDISATION COMPOUNDS O F BOROS TRICHLORIDE
125’7
never filled irith either of these gases t o a pressure as great as their vapor pressure at 0°C. The pressure of phosgene i m s never permitted t o esceed 495 mm., inasmuch as its vapor pressure at 0°C. is 504 mm. (16). Similarly, the pressure of boron trichloride was niaintained below 4.50 nim., since 470 mm. is its vapor pressure at 0°C. (18). -0C.
'lie pressure of the boron trifluoride in the calibrated fla& was kept just helow i atm. t o ai-oid errors due t o compressibility. THE SYSTEM PHOSGI2SE-BOROS
THIFLL-ORIDE
Thc data for this system are listed in tahlc 1 and d ~ p i c t c din the phase rule diagram in figure 1. These indicate the existence of t w o compoiin(lq whew (nom-
1258
DOSALD R h T 11.kRTIS A S D J O H Z PHILIP FAUST.
VI
positions may he expressed hest as COCl?.13F3and (COCL)?.RF,, having freezing-point values of -1343°C. & 0.5" and -13i.O"C. =k 0.5", respectively. Both compounds condense as colorless liquids and freeze as ivhite solids. A minimum point was observed on each side of each maximum in the diagram. These minima \\-ere found at 23.2 =k 0.2 mole per cent boron trifluoride and -143.0"C. =k 0.5", 38.4 i 0.2 mole per cent h r o n trifluoritle and -138.0"C. i 0.5", and 60.4 i 0.2 mole per cent boron trifluoride and - 142.3"C. & 0.5". Although this invesBigation aft'orcls no evidence as to n-hether the chlorine or the oxygen is the donor atom in these compounds, it is of interest t o revienthe literature on this subject. The following structures may he postulated for these compounds:
A
R
Structures (' and F are supported 1-, the \\-ark of Brown, Schlesinger, and Burg (51, who studied the relat iJ-e st:it)ilities of molecular compounds formed by boron trifluoritle \\-ith acetone, trimethT.lacetaldeli~-de,chloral, and awtyl chloride. They concluded that it. is the carbonyl oxygen in acetyl chloride which is the donor atom rather than the chlorine atom. The remaining structures find support in the work of See1 (11). €€e p r e p a i d the colorless solid conipound CHL4COF.RF3 by allowing acetyl fluoride t o condense n i t h t m o n trifluoride, with or without an inert solvent such as carbon tetrachloride, chloroform, or liquid sulfur dioxide. The solution of @H3COF.BF3 in liquid siilf'ur dioside x i s found t o be an electrolyte, thus giving evidence for the existence of CH3CO+and BF4-. There is little doubt t,hat the fluorine atom is t,he donor atom in the ahove (~oiiipo~uid. inasmuch as See1 demonstrated 1,calculations involving lattice energim, heats of vaporization, and ionization energies that the boron trifluoride coultl not be att,ached t o the carbonyl osj-gen atom. If esperimentally possible, it xould be of interest t o test these compounds
COORDISATIOS
COUPOUSDS
OF BOROS
TRICHLORIDE
1259
of phosgene and boron trifluoride t o see if they possess sufficient ionic character t o be conductors. However, their lox melting points indicate that they are probably predominantly of c ovalent character. The structures involving coordination through the chlorine (-4, E, D, E) are supported also by the analogous compound formed by phosgene and aluminum chloride, COCI~~AlC13, which ionizes t o yield the ions (6):
c1
Cl-C=O+
I and Cl-A1-Cl-
I
c1 If one of the chlorine atoms of phosgene can donate t o the boron atom of lioron trifluoride, the question naturally arises as t o whether or not the >econd chlorine atom can behave likewise. I n this study, there wa5 no evidence for a compound between 2 moles of boron trifluoride and 1 mole of phosgene. However, it must be borne in mind that the second chlorine n-ould lie lesc likely t o donate after the first chlorine atom became a donor, oning t o the resultant displacement of electrons within the phosgene molecule. The fact that the carbon-to-oxygen linkage is different in the two molecules of phosgene in structure F arises from the shift in electrons about the carbon atom of the phosgene molecule which is the donor t o the boron trifluoride. Luder and Zuffanti (9) give a mechanism u-hich has been postulated for the displacement of a pair of electrons of a carbonyl oxygen, leaving an electron deficit on the carbon :
e1
e1 1
C : : 6 - + + C : O : +BF,
..
c1
..
e1
C1
..-
F I
I
+
+C:O:&-F
..
C1
I
F
This allo\\-s a second molecule of phosgene t o add t o the carbon having the electron deficit. During the course of the investigation of this system, glass formation \vas encountered in the range bet\\-een 30 and GO mole per cent boron trifluoride. Fur t8n-opoints in this range the usual technique for obtaining freezing points \\-as used (49.5 and 33.8 mole per cent boron trifluoride). However, for the other points a method \vas used in which the freezing-point cell \\-as cooled 11y imniei,sion in liquid air until a slush \\-as obtained in the freezing-point cell. The liquid air \\-as removed and the contents of the freezing-point cell allon-cd t o \\-arm until the solenoid stirrer just began t o operate. The contents of the cell then were inimersed in liquid air until the stirrer stopped, care being exercised not to overcool the slush. -4 short break in the cooling curve \\-as thus obtained on the recording potentiometer. The accuracy of these points is not considered t u be better than &0.5'C.
TABLE 2 Dnta fo: flit s:istcvi phosgene-boron trichloride UOLE I'R.ACTIOS O F
I
I
BCIi
FREEZIXG P m s r
X O L E FR.ACTIOS O F
__
__.
BC13
FREEZISC P O I S T
___ -__
10.002
:'-,
1,000
~
0.S46 ::I' 0.500 k1'1 0.753 'ai 0.704 h l 0.662 (bi 0.606 !b) 0.550 i b ' 0.501 f i ) ) -~
10.3 107.3 1os.3 110.0 111.cj 113.6 115.0 118.0
I
119.6
121 .o 123.0 126.5
~~
&0.3
*0.003
0.458 f c ) 0.424 (b) 0.354 ( e ) 0.312 (c) 0.264 (c) 0.216 ( c ) 0.164 ( c ) 0.122 ( c ) 0.070 ( c ) 0.000
- "C.
128.0 129.6 135.3 139 .0 142.3 141 .0 138.0 135.0 133.0 132.5
1
1
1 1 ~
I
i
~
"
:I, b,
"C
.
and c are different analyses.
I05
I13
I21
123
137
I4 5
I
I
C O 6 R D I S A T I O S COMPOUSDY O F BOKOS TRICHLOHIDE T H E SYSTEM PHOSGESE-BOROS
1261
TRICHLORIDE
Inasmuch as compounds are obtained by allon-ing boron :rifluoritle t o react with phosgene, it was of interest t o see ii boron trichloride, although a iworer acceptor molecule, would behave in an analogous manner. With this system, just as with the one with boron trifluoride, thcw is thc possibility of the phosgene coordinating t o the boron trichloride through the carbonyl oxygen or through the chlorine. Examples of ketones (15) as ne11 a s nlkyl chlorides (12) coordinating with boron trichloride are known, and therefore there is a possibility of coordination taking place n-ith phosgene. =Iccordingly, the system phosgene-boron trichloride was studied by thermal analysib. The data in table 2, depicted in the phase rule diagram in hgure 2 , indicate t h a t phosgene and boron trichloride do not form any coordination compounds. A eutectic point \vas found a t 25.6 i 0.2 mole per cent boron trichloride and -142.3'C. i 0.3". However, since the tivo compounds are formed between boron trifluoride and phosgene only a t low temperatures, and since lioron trichloride is a poorer acceptor molecule than boron trifluoride, it is not surprising that no compounds exist between phosgene and boron trichloride. h s predicted by Germann (6), boron trichloride and phosgene are niiwible in all proportions. SUJIK4RT
The system phosgene-boron trifluoride \vab studied by thermal aiialyQs. Evidence was obtained for the eiistence of t u o compounds, COCl,.BF, and (COCl?)?.BF3, having freezing points of - 134.3"C. f 0.5' and -137.0"C. i O.jO, respectively. 11inima n-ere found at 23.2 f 0.2 niolc per cent boron trifluoride and -143.0"C. i 0.5', 35.4 =t0.2 mole per cent boron trifluoride and -138.O"C. f O.jO,and 60.4 f 0.2 mole per cent boron trifluoridc and - 142.3"C. i0.5". A thermal analysis of the system phosgene--horon trichloride revealed that the components are soluble in all proportion
