Electrodeposition of Metals from Organic Solutions III. Preparation and Electrolysis of Titanium and Zirconium Compounds in Nonaqueous Media W. E. :REID, JR., J. M. BISH, AND ABNER BRENNER National Bureau of Standards, Washington, D. C. ABSTRACT Numerous nonaqueous solutions of .Ti and Zr compounds were electrolyzed in an attempt to electrodeposit the metals. Ether solutions containing halides, hydrides, borohydrides, and organo-metallic compounds of Ti and Zr were the most promising. A mixed type of bath containing both hydrides and borohydrides yielded Ti-A1 alloys containing about 6% Ti. Similar baths containing Zr instead of Ti gave alloy deposits containing up to 450-/oZr. New methods of preparation of titanium and zirconium boro- hydrides were developed. Coatings of Ti would be of considerable value for pro- cting steel against corrosion. For this reason, methods for ?ctrodepositing Ti have been under intensive investiga- )n by many workers during the past few years. Attempts to electrodeposit Ti go back over 100 years hen Becquerel (1) claimed to have deposited a Ti-Fe loy from aqueous solution. Experiments on aqueous ectrolysis have continued since that date without success, ~spite claims to the contrary. Russell (2, 3), investigating ectrode potentials, reported that Ti amalgamated with g to a very slight extent. He electrolyzed an acid solution TIC13 using a water-cooled rocking Hg cathode at a high trrent density. He claimed a deposition rate of 0.2 g i/hr. A method was patented (4) for plating Ti on a ~se metal from strongly alkaline solutions at low current ~,nsity. About 20 mg Ti were deposited by 0.1 amp-hr. hin, black coatings of Ti on Pb, Zn, or Sn cathodes were ~posited from solutions containing Ti2(S04)3 and Na2S04 ; 7). Tartaric acid solutions were also suitable. Grat- anskii (8) obtained thin Ti films on a Cu cathode from dfanilic acid solutions. A patent was issued (9) covering m electrolysis of salts melted in their water of crystalliza- on; titanium sulfate trihydrate was one of the salts entioned. Fink (10) patented a bath consisting of Ti02, IF, HC1, and gelatine with traces of Cu to depolarize le deposition of Ti. Keys and Swann (11) reported their ~ability to plate Ti from aqueous solutions. Ti contaminated with oxides has been deposited from tucous sols containing colloidal Ti, but it could not be ectrodeposited from solutions of its salts (12). A very thin ~posit of Ti-Zn alloy was obtained from aqueous solutions retaining a peroxide and tartrate (13). Some alloy de- )sits containing Ti were formed from anhydrous and tueous systems (14). Three general references on Ti are tcluded in the bibliography (15-17). Experiments on the electrodeposition of Zr have also ,iled to yield any successful metallic deposits. Bradt (18) aimed to have established optimum conditions for plating right Zr from solutions of zirconyl sulfate on a Cu cathode. he deposits were very thin. Laubengayer (19) confirmed m findings of Bradt, but Plotnikov (20) obtained only 21 H at the cathode. An investigation of the electrodeposition of Zr from organic plating baths was unsuccessful (21). Holt (22) investigated the eleetrodeposition of Zr from aqueous and organic plating baths without obtaining any successful metallic Zr deposits. This shows that up to now no substantiated evidence has been presented for the successful electrodeposition of Ti or Zr from either aqueous or organic plating solutions. Although Ti metal can be deposited from high temperature fused salt baths (23, 24), the operational difficulties in- volved make the development of a plating bath capable of operation at room temperature highly desirable. The available evidence indicates that these metals are probably too reactive to be electrodeposited from aqueous solutions. This conclusion has focused attention on the use of organic solvents as a plating medium. Although organic plating baths have been known for many years (25), it was not until the development of the hydride aluminum plating bath by Couch and Brenner (26) that the consistent deposition of good plates from an organic plating bath was demonstrated as feasible. SOLVENTS In the exploratory phase of the work here (1951-52), Sherfey and Senderoff of this laboratory dissolved Ti halides in a number of solyents and electrolyzed the resulting solutions. No successful deposition of the metal resulted. However, the systems investigated are listed in Table I to indicate the diversity of solvents tried. The conclusion is that Ti halides were not the proper type of compound. Of the solvent types used, ethers were the most suitable. Atmospheric oxidation of ethers to peroxides and other compounds occasionally caused difficulties. For example, the first additions of lithium aluminum hydride to 1,2- dimethoxyethane or tetrahydrofuran reacted very vigor- ously. This was caused by impurities resulting from atmospheric oxidation. Acetaldehyde was identified in the dimethoxyethane. The latter is so sensitive to oxidation that even a very brief exposure to the atmosphere resulted in formation of peroxides. Ethers should be free of oxidized ) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66 Downloaded on 2014-06-17 to IP
Transcript
Electrodeposition of Metals from Organic Solutions
III. Preparation and Electrolysis of Titanium and Zirconium Compounds in Nonaqueous Media
W. E. :REID, JR., J. M. BISH, AND ABNER BRENNER
National Bureau of Standards, Washington, D. C.
ABSTRACT
Numerous nonaqueous solutions of .Ti and Zr compounds were electrolyzed in an attempt to electrodeposit the metals. Ether solutions containing halides, hydrides, borohydrides, and organo-metallic compounds of Ti and Zr were the most promising. A mixed type of bath containing both hydrides and borohydrides yielded Ti-A1 alloys containing about 6% Ti. Similar baths containing Zr instead of Ti gave alloy deposits containing up to 450-/o Zr. New methods of preparation of titanium and zirconium boro- hydrides were developed.
Coatings of Ti would be of considerable value for pro- cting steel against corrosion. For this reason, methods for ?ctrodepositing Ti have been under intensive investiga- )n by many workers during the past few years. Attempts to electrodeposit Ti go back over 100 years hen Becquerel (1) claimed to have deposited a Ti-Fe loy from aqueous solution. Experiments on aqueous ectrolysis have continued since that date without success, ~spite claims to the contrary. Russell (2, 3), investigating ectrode potentials, reported that Ti amalgamated with g to a very slight extent. He electrolyzed an acid solution TIC13 using a water-cooled rocking Hg cathode at a high
trrent density. He claimed a deposition rate of 0.2 g i/hr. A method was patented (4) for plating Ti on a ~se metal from strongly alkaline solutions at low current ~,nsity. About 20 mg Ti were deposited by 0.1 amp-hr. hin, black coatings of Ti on Pb, Zn, or Sn cathodes were ~posited from solutions containing Ti2(S04)3 and Na2S04 ; 7). Tartaric acid solutions were also suitable. Grat- anskii (8) obtained thin Ti films on a Cu cathode from dfanilic acid solutions. A patent was issued (9) covering m electrolysis of salts melted in their water of crystalliza- on; titanium sulfate trihydrate was one of the salts entioned. Fink (10) patented a bath consisting of Ti02, IF, HC1, and gelatine with traces of Cu to depolarize le deposition of Ti. Keys and Swann (11) reported their ~ability to plate Ti from aqueous solutions. Ti contaminated with oxides has been deposited from
tucous sols containing colloidal Ti, but it could not be ectrodeposited from solutions of its salts (12). A very thin ~posit of Ti-Zn alloy was obtained from aqueous solutions retaining a peroxide and tartrate (13). Some alloy de- )sits containing Ti were formed from anhydrous and tueous systems (14). Three general references on Ti are tcluded in the bibliography (15-17). Experiments on the electrodeposition of Zr have also
,iled to yield any successful metallic deposits. Bradt (18) aimed to have established optimum conditions for plating right Zr from solutions of zirconyl sulfate on a Cu cathode. he deposits were very thin. Laubengayer (19) confirmed m findings of Bradt, but Plotnikov (20) obtained only
21
H at the cathode. An investigation of the electrodeposition of Zr from organic plating baths was unsuccessful (21). Holt (22) investigated the eleetrodeposition of Zr from aqueous and organic plating baths without obtaining any successful metallic Zr deposits.
This shows that up to now no substantiated evidence has been presented for the successful electrodeposition of Ti or Zr from either aqueous or organic plating solutions. Although Ti metal can be deposited from high temperature fused salt baths (23, 24), the operational difficulties in- volved make the development of a plating bath capable of operation at room temperature highly desirable.
The available evidence indicates that these metals are probably too reactive to be electrodeposited from aqueous solutions. This conclusion has focused attention on the use of organic solvents as a plating medium. Although organic plating baths have been known for many years (25), it was not until the development of the hydride aluminum plating bath by Couch and Brenner (26) that the consistent deposition of good plates from an organic plating bath was demonstrated as feasible.
SOLVENTS
In the exploratory phase of the work here (1951-52), Sherfey and Senderoff of this laboratory dissolved Ti halides in a number of solyents and electrolyzed the resulting solutions. No successful deposition of the metal resulted. However, the systems investigated are listed in Table I to indicate the diversity of solvents tried. The conclusion is that Ti halides were not the proper type of compound. Of the solvent types used, ethers were the most suitable.
Atmospheric oxidation of ethers to peroxides and other compounds occasionally caused difficulties. For example, the first additions of lithium aluminum hydride to 1,2- dimethoxyethane or tetrahydrofuran reacted very vigor- ously. This was caused by impurities resulting from atmospheric oxidation. Acetaldehyde was identified in the dimethoxyethane. The latter is so sensitive to oxidation that even a very brief exposure to the atmosphere resulted in formation of peroxides. Ethers should be free of oxidized
) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66Downloaded on 2014-06-17 to IP
Dimethylformamide + Li Alcohols Dimethylformamide -t- Na Butyl alcohol N , N - dimethyltrichloro- Ter t -butyl alcohol acetamide Carbitol Ethanol (0~ and 25~ Nitriles Ethanol + boron trifluoride Ethanol -{- ethyl ether -i- Dial lylcyanamide
HC1 Dimethylcyanamide Ethanol -}- HC1 Acetonitri le Ethanol + HC1 (aqueous) Acrylonitri le Ethanol + Na Oxydipropionitrile Ethanol -{- te t ramethylam-
Benzothiazole But ly ether -t- A 1 C 1 3 Methylmagnesium bromide 2-Ethoxyethanol Methylmagnesium bromide 1,4-Dioxane -{- ether 1,4-Dioxane + Li N-methylpyr id inum methyl 1,4-Dioxane ~- Na sulfate Ethyl ether Methyl sulfate Ethyl ether -[- A1Br3 Methyltrichlorosilane Ethyl ether -~ A1Br3 q- 3-Nitro 4-chloro benzo-
LiAH4 trifluoride E thy l ether -b A 1 C 1 3 Ni t romethane E thy l ether -k A1Cla q- Isopropyl t i tanate
LiAH~ Ethyl ether -~ Li Inorganic Ethyl ether -}- LiBHt Ethyl ether -{- Na PC13 Ethy l ether -~- NaBHt S2C12 Tetrahydrofuran S02C12
SOCI2 Ketones TiI t Acetone Water
compounds when used as solvents for organometall ic, hydride, or borohydride type compounds, such as were used in this study.
A necessary prerequisite of a plating bath is tha t a high concentrat ion of complexed ions be mainta ined in solution. This made necessary a s tudy of the solubilities of T i and Zr compounds in various organic solvents. Approximate solubilities obtained at room tempera ture are given in Table I I .
T~ AND ZR COMPLnX ~-~YDRIDES
Ti (III) borohydride solutions.--The borohydrides con- s t i tute a new type of electrolyte, the electrolytic proper- ties of which have not been previously investigated. These compounds may be divided roughly into two classes: the ionic borohydrides and the covalent borohydrides. The ionic borohydrides, such as NaBH4, are quite stable even when exposed to the atmosphere. They are insoluble in the usual organic solvents. The covalent borohydrides, such as those of Be or A1 are unstable and are much more reactive. They react spontaneously with the atmosphere, often with explosive violence, and must be studied in an
TABLE II . Solubilities of Ti and Zr compounds in various organic solvents
) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66Downloaded on 2014-06-17 to IP
Vol. 104, No. 1 E L E C T R O D E P O S I T I O N O F M E T A L S 23
inert atmosphere. The covalent borohydrides are usually very soluble in organic solvents, and this property makes them of interest as electrolytes.
Hoekstra and Katz (27) prepared t i tanium(III ) boro- hydride, Ti(BH4)a, by passing TiCt4 over LiBH4. Titanium (III) monochloroborohydride, TiCI(BH4)2, was formed when TIC14 reacted with AI(BH4)a. Ti(BH0a was green and the monochloroborohydride was blue. Ti(BH4)a was re- ported to be unstable at room temperature, decomposing completely within several days (27).
Ti(BH4)a was prepared by this method, but it was un- satisfactory for preparing more than milligram quantities. Ti(BH4)a for plating experiments was prepared by the reaction of a Ti halide with Al(BH.t)a or LiBH4 in ethereal solution. All experiments were performed in a He atmos- phere, and no a t tempt was made to isolate the resulting Ti(BH4)a. As shown in Table I I I , some Ti halides reacted with AI(BH4)a to give green solutions, and some gave blue solutions. Assuming that the ether solutions had the same color as the compounds, this indicated in some cases the solutions probably contained a mixture of both types of t i tanium borohydrides.
Results of the electrolysis experiments with the Ti(BH4)a solutions are included in Table I I I . Pure metal was not obtained. Ethereal borohydride solutions prepared from Ti halides all yielded nonmetMlic, black, moisture-sensitive deposits on electrolysis. None of the solutions conducted very well. A tetrahydrofuran solution of t i tanium (III) borohydride tetrahydrofuranate did not conduct at all.
The black deposit obtained fi'om electrolysis of the green solution of TiC14-AI(BH4)a (Table I I I , No. 7) con- tained 19% Ti and may be considered typical in composi- tion and appearance for the nonmetallic Ti deposits described in Table I I I . I t also contained 19% A1, 33% C1, and undetermined B. As this solution was electrolyzed, the intensity of the green color diminished and a brownish- red material which formed at the A1 anode soon discolored the solution.
Ethyl ether solutions of Ti borohydride were unstable on standing for several days at room temperature, as indicated by a change in color from green to brown, a decrease in conductivity, and precipitation of a brown solid.
Addition of hydride aluminum plating bath (26) to TiC14-AI(BH4)a solution gave a brown solution (Table I I I , No. 9). Electrolysis of this solution gave a good metallic deposit containing Ti, A1, and B. Maximum Ti content of an alloy deposit obtained from a bath of this type was 6.5%. A deposit obtained from this same bath four days later contained 1 6% Ti, 94.7% A1, and 2 6% B.
Uncertainty regarding composition of the borohydride compound in the solutions led to at tempts to prepare pure Ti(BH0a by a new method satisfactory for producing large quantities. This involved the reaction of diborane with a Ti ester of the type Ti(OR)4. Titanium borohydride tetrahydrofuranate, Ti(BH4)a.C4HsO, was formed in tetrahydrofuran. Preparation of this compound is discussed in the Appendix.
An ether solution containing t i tanium borohydride tetrahydrofuranate and hydride aluminum plating bath was electrolyzed. The resulting alloy deposit contained
TABLE I I I . Electrolysis of ethyl ether solutions of Ti (BH4)a and Zr(BH4)4
Preparation of electrolyte
T i o r Zr Reagent salt present added
At(BH,) 1. K2TiF6
2. TiCla
3. NaTiC14
4. TiFa 5. TiF4
6. TiCla
7. TIC14
8. TiBra
9. TIC14
10. ZrCI~
11. ZrC14
LiBH4
LiBH4
LiBH4 LiBH4
A1 (BH4) 3
A1 (BH~)
LiBH4
AI(BH4)a + A1 plating bath
A1 (BH4) 3 + A1 plating bath
A1 (BH4) 3
Wt. % Nature of Ti or
deposit Zr;in deposit
N o n e
Black, moisture sensitive
None
Li deposit - - Black,
moisture sensitive
Black, 22.4 moisture sensitive
Black, 19.1 moisture sensitive
Black, moisture sensitive
Metallic 6.5
Metallic 22.9
Black, 34.2 moisture sensitive
* Zr(BH4)~ solutions are colorless.
6.3% Ti. Attempts to prepare a plating bath by direct addition of LiA1H4 solution alone instead of addition of the hydride aluminum bath to Ti(BH4)a solution at room temperature caused decomposition as indicated by a vigorous evolution of gas and the formation of a black solid which discolored the solution. This would indicate that LiA1H4 reacts or complexes with A1Cla preferentially so that its reactivity with Ti(BH4)a is decreased.
Titanium aluminum hydride.--In the reaction of LiA1H~ with Ti(BH4)a, t i tanium aluminum hydride, Ti(A1H4)4, may be formed first but then decomposes immediately. This seems likely because the compound has been re- ported in the l i terature (28) as being stable only at dry ice temperature or below and decomposing on warming to room temperature. The compound is insoluble in ethyl ether at its temperature of existence. Some of the com- pound was prepared in methyl ether a t -80~ and the solution was electrolyzed as it warmed up slowly to the boiling point of this solvent. The compound did not appear to be very soluble in this ether, and conductivity was very low. No metallic deposit was obtained.
ZR-AL ALLOY BATHS
Type A bath (Zr(BH4)4 and LiA1H4).--Zr(BH4)4 in ether did not conduct, and ether solutions of LiA1H4 were only slightly conducting. When these two solutions were mixed, however, the resulting solution conducted very well, and
) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66Downloaded on 2014-06-17 to IP
24 J O U R N A L O F T H E E L E C T R O C H E M I C A L S O C I E T Y January 1957
a sound metallic deposit containing about 8% Zr was ob- tained. The ratio of LiA1H4 to Zr(BH4)4 necessary to obtain a metallic deposit from these ether solutions was critical. This was shown by a comparison of baths contain- ing the following ratios: (a) 1.7; (b) 1.0; and (c) 0.5. Only in the case of (b) was a metallic deposit obtained on electrolysis. The other two solutions gave black moisture- sensitive deposits. Solution (a) was unstable and, after standing for several hours at room temperature, a metallic deposit was formed which coated the wall of the reaction vessel. Similar decomposition occurred with solution (b), but much more slowly and to a lesser degree. Solution (c) did not decompose, remaining unchanged even after standing several days.
Type B bath (Type A bath with A1C13).--The successive addition of A1C13 to Type A baths resulted first in a marked increase in the amount of Zr in the alloy deposit, and then a gradual decrease. However, even when the AICh content was relatively large, the Zr content of the deposit was still greater than in the alloys from Type A baths. The result suggests that A1C13 is an important component of these baths. Several baths of this type were made up with differ- ing ratios of Zr(BH4)4 to A1CI~ to determine the relation between bath composition and deposit composition. Re- sults of this study are listed in Table IV. These data show that Zr content of the deposits does not vary regularly with Zr content of the baths.
Type C bath.--In earlier work, AI(BH4)3 and ZrCla were used instead of pure Zr(BH4)4. The deposits obtained were similar in appearance and Zr content to the alloys from Type B baths. The deposit with the highest Zr con- tent obtained from this type of bath contained 44.5% Zr, 53.0% A1, and 2.5% B. Due to the greater number of variables in Type C baths, they were not studied as extensively as Type B baths.
Characteristics of the Zr Plating Bath
These baths are similar to the aluminum hydride bath (26) in that they are subject to decomposition by the atmosphere. As Zr and Ti dissolved anodically with poor current efficiency in these solutions, A1 anodes were used in all cases. Although this metal dissolved anodically with good efficiency, some decomposition still occurred at the
TABLE IV. Electrodeposited Zr-A1 Alloys from plating baths containing hydrides, borohydrides and chloride
anode. With Zr and Ti anodes, decomposition was much more rapid.
The alloy baths were unstable and gave no deposit after a few weeks use. They appeared to be stabilized by A1Cl~, which also improved the appearance of the deposits. Those baths which contained only a small amount of A1Cla relative to Zr(BH4), were so unstable that metal deposition usually ceased after the bath had stood for 24 hr. On prolonged electrolysis, the Zr content of the deposits usually decreased more than could be accounted for by depletion of the Zr content of the solution.
At a LiA1H4--Zr(BH4)4 mole ratio greater than unity, spontaneous decomposition of the bath occurred, and a metallic mirror deposited on the surface of the electrolysis cell. Analysis of one such deposit gave the following results: 12.4 Zr, 78.8 A1, 9.2 B (per cent by weight).
If the complexes in these baths are similar to those of the hydride aluminum plating bath, then Li t t should be as effective as LiA1H4, but this was not the case. L i i I did not react with an ether solution of Zr(BH04, and the re- sulting suspension did not conduct. When this experiment was repeated with equimolar quantities of A1Cln, LiH, and Zr(BH4)4, the resulting solution gave no metallic deposit on electrolysis.
Ethyl ether plays a more important role than tha t of a mere solvent in these baths, as LiA1H4 and Zr(BH4)4 do not react in its absence. In the presence of ether they reacted vigorously to form a brownish-purple solution. Since the resulting solution conducted well and gave an alloy deposit when electrolyzed, the ether apparently served both as a complexing and ionizing medium for the LiA1H4.
The Zr compound responsible for deposition of the alloys is probably a zirconium borohydride-aluminum hydride complex which is formed in solution. This view is supported by the observation that LiA1H4. conducts poorly in ethyl ether, and Zr(BH,I)4 does not conduct at all. When the two solutions were mixed, a white solid of empirical composition Zr(A1H4)4 formed immediately. With excess Zr(BH~)4, the white solid dissolved to give a solution of good conductivity. The solubility of Zr(A1H4)4 in an excess of Zr(BH4)4, and the good conductivity of the resulting solution indicate the presence of a complex.
Preparation of the compound Zr(A1H4)4 is included in the Appendix since it has not been reported previously in the literature.
ORGANOMETALLIC COMPOUNDS
Organotitanium compounds.--Only within the last few years have methods become available for preparation of compounds containing T i - - C bonds (29-31). Such com- pounds would have greater solubility and greater stabili ty in organic solvents than the hydride type previously used in plating solutions. Although preparation of such com- pounds is time-consuming, i t was thought that their s tudy would be of interest in view of the fact that other metals can be deposited from baths containing organo- metallic compounds (32).
Two organotitanium compounds were studied: bis- (cyclopentadienyl)titanium(IV) dibromide and phenyl- t i tanium triisopropylate.
) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66Downloaded on 2014-06-17 to IP
Vol. 10~, No. 1 E L E C T R O D E P O S I T I O N O F M E T A L S 25
A quanti ty of bis-(cyclopentadienyl) t i tanium(IV) dibromide (29) was prepared, and numerous solutions of this compound in various solvents were electrolyzed. In no case was Ti deposited. Solubility of this compound in organic solvents is very low, and its solutions are weakly conducting, indicating that i t does not form ionic com- plexes easily. Solubilities of this compound are listed in Table II .
Phenylt i tanium triisopropylate was prepared as de- scribed by Herman and Nelson (30). The compound was not isolated, but its resulting ether solution was electro- lyzed. Conductivity was very small and was not improved by addition of N a i l or phenyllithium. Electrolysis of this solution did not give a metallic deposit.
Organozirconium compounds.--Up to the present, the only organozirconium compound described in the literature is the bis-(cyclopentadienyl)zirconium (IV) dibromide (31). In view of the lack of success with the similar Ti compound, no a t tempt was made to synthesize this Zr compound.
Addition of Zr(BH4)4 to an ether solution of phenyl- lithium turned the solution black and caused evolution of heat. A similar reaction occurred with ZrCh and phenyl- lithium. In both cases the isolated reaction product gave a good Michler's ketone test for a carbon to metal bond. Although the exact composition of this compound was not determined, analyses indicated that it contained an appreciable amount of C. When the solvent was com- pletely removed from this compound, the product was a tan powder (mp 175~176 which reacted vigorously on exposure to the atmosphere. I t was soluble in phenetole, benzene, and tetrahydrofuran, and sparingly soluble in ethyl ether and dimethoxyethane. Ethyl ether solutions of this compound conducted slightly, but no metallic deposit was obtained on electrolysis. Although the compound was not very soluble in ether, it dissolved readily when phenyl- lithium was added to a suspension in ethyl ether. A solid separated out of the benzene solution after several weeks. The precipitate was insoluble in the solvents previously mentioned and did not react with phenyllithium. This com- pound, discussed further in the Appendix, is probably an impure phenylzirconium derivative.
PERFLUOROACID SALTS
The presence of the fluorinated hydrocarbon chain in perfluoroacid salts makes these salts highly soluble in many organic solvents and suggests their use in organic plating baths. The Ni, Ag, Ti, and Zr salts of pel~uoroace- tie acid were prepared as described in the Appendix.
None of these salts yielded a conducting solution in ethyl ether or benzene. The Ni salt of perfluoroacetic acid dissolved in dimethylformamide to give a conducting solution from which a Ni deposit was obtained. These deposits were stressed, poorly adherent, and contained organic matter.
A Ti derivative containing chloride was prepared by direct reaction of TIC14 with trifluoroacetie acid. An a t tempt to prepare the Ti salt in benzene solution was unsuccessful. TIC14 reacted in benzene solutions of silver trifluoroacetatc or silver perfluorobutyrate to form a white precipitate. I t probably consisted of the Ti compound,
which is relatively insoluble in benzene, mixed with AgC1. Conducting solutions were obtained with the t i tanium trifluoroacetate derivative dissolved in dimethoxyethane, anisole, or dimethylformamide. No deposit was obtained from any of these solutions on electrolysis.
FUSED ORGANIC SALT SYSTEMS
Several organic salt systems, which are molten at low temperatures, were studied. Two quaternary ammonium tetrachlorodibromotitanates which were the double salts of TIC]4 with tetraethylammonium bromide and ethyl- pyridinium bromide (33) were prepared and investigated. These salts melt near 200~ Electrolysis of each of these salts at 250~ gave no deposit. Addition of either of these salts to the ethylpyridinium bromide-A1C13 type of A1 plating bath (34) gave a black, tar ry deposit when electro- lyzed. These deposits contained Ti.
When this bath was electrolyzed with a Ti anode, metal- lic deposits of a Ti-A1 alloy were obtained, but with very low cathode efficiency. This bath yielded deposits con- taining up to 14% Ti. Alloys of higher Ti content could not be obtained, presumably because of electrolytic decomposi- tion of the bath on prolonged electrolysis.
Hurley-Wier baths containing the following compounds were also electrolyzed: tetraisopropylti tanate, bis (cyelo- pentadienyl)ti tanium(IV) dibromide, and sodium ti tanium chloride. The anode was A1. In no case was a metallic deposit obtained.
For use in fused salt electrolysis, a series of anhydrous tr ivalent Ti halogen derivatives of the general formula MTiC14 was prepared. The sodium derivative was prepared by reduction of TIC14 with Na in an inert solvent a t 100~ The green salt obtained reacts with many polar organic solvents, and behaves much as TIC13. I t dissolves in fused halide melts to form a solution from which Ti may be obtained on electrolysis. These compounds have been described (38) and will be the subieet of a future publica- tion.
LOWER VALENT TI COMPOUNDS IN ORGANIC SOLUTIONS
Lower valent Ti compounds were prepared in various ether solutions by the reduction of TIC14 with various reducing agents (see Appendix). Electrolysis of these solutions did not yield a metallic deposit.
PREPARATION OF TIC13
An essential part of this program was preparation of new compounds and simplification of present methods of preparation of compounds used in electrolysis studies. An improved method for preparation of TiCla falls under the lat ter category.
By reaction of TIC[, with hydrogen in a bomb at 500~ and 2000 lb/in. 2, 12% was reduced to TIC13 in 30 hr. If allowed to proceed for a longer time this reaction may give a rather high yield of TIC13. This method has certain ad- vantages over other procedures employing hydrogen in that a lower temperature is used, so that the possibility of contamination with TiCl~ is minimized, and the reaction requires no attention while in process. Presumably, the use of higher pressures would considerably speed up the reaction. The preparation of this compound is discussed further in the appendix.
) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66Downloaded on 2014-06-17 to IP
26 JOURNAL OF THE ELECTROCHEMICAL SOCIETY January 1957
Manuscript received May 21, 1956. This paper was pre- pared for delivery before the Boston Meeting, October 3 to 7, 1954. The work was sponsored by the Wright Air Development Center under U S A F Contract No. AF(33- 616)53-11.
Any discussion of this paper will appear in a Discussion Section to be published in the December 1957 JOURNAL.
R E F E R E N C E S
1. M. BECQUEREL, Ann. chim. phys., 48, 337 (1831). 2. A. S. RUSSELL, Nature, 127, 273 (1931). 3. A. S. RUSSELL, Nature, 142, 210 (1938). 4. E. POKORNY, (I. G. Farbenindustrie, A. G.), German
Pat. 605,551, Nov. 13, 1934. 5. ~i. HAISSINSKY AND H. EMMANUEL-ZAvIZZIANO, Compt.
rend., 204, 759 (1937). 6. ~i. HAISSINSKY AND H. EMMANUEL-ZAvIZZIANO, J .
(RD), Virginia Ins t i tu te for Scientific Research, 1953.
13. Progress Report No. 3, Contract No. AF 33(616-75), School of Mines and Metallurgy, Univ. of Missouri, 1952.
14. Final Report , AF 33(038)50-1085, U. S. Bureau of Mines, Materials Laboratory, 1953.
15. J. BARKSDALE, "T i t an ium," The Ronald Press Co., New York (1949).
16. C. A. BROPHY, B. F. ARCHER, AND R. W. GIBSON, "Ti tan ium Bibliography 1900-1951," Battel le Me- morial Inst i tute , Columbus (1952); see also 1952 Supplenmnt by B. J. ARCHER AND m. W. GIBSON (1953).
!8. W. E. BRADT AND H. B. LINFORD, This Journal, 70, 431 (1936).
19. A. W. LAUBENGAYER AND R. B. EATON, J . Am. Chem. Soe., 62, 2704 (1940).
20. V. A. PLOTNIKOV AND E. B. GUTMAN, J. Appl. Chem. USSR, 19,826 (1946).
21. GRAHAM, CROWLEY, AND ASSOCIATES, INC., Electro- deposition of Zirconium, 1951-53, Contract No. AT(11-1)-173.
22. M. L. HOLT, This Journal, 98, 33C (1951). 23. A. BRENNER AND S. SENDEROFF, ibid, 99, 223C (1952). 24. G. D. P. CORDNER AND H. W. WORNER, Australian J.
Appl. Science, 2,358 (1951). 25. W. A. PLOTNIKOFF, J. Russ. Phys. Chem. Soc., 4, 466
(1902). 26. D. E. COUCH ANn A. BRENNER, This Journal, 99, 234
(1952). 27. H. R. HOEKSTRA AND J. J. KATZ, J. Am. Chem. Soc.,
71, 2488 (1949). 28. E. WIBERG AND R. USON, Z. Naturforsch., 6b, 392,
(1951). 29. G. WILKINSON, P. L. PAUSON, J. M. BIRMINGHAM,
ANn F. A. COTTON, J. Am. Chem. Soc., 75, 1011 (1953). 30. D. F. HERMAN AND W. K. NELSON, J. Am. Chem. Sou.,
74, 2693 (1952). 31. D. F. HERMAN AND W. K. NELSON, ibid., 75, 3877
(1953). 32. D. B. KEYES, S. SWANN, JR., W. KLABVNDE, ANn S.
T. SCmCKTANZ, Ind. Eng. Chem., 20, 1068 (1928). 33. J. BYE AND W. HAEGI, Compt. rend., 236, 381 (1953).
34. F. H. HURLEY AND T. P. WIER, U. S. Pat . 2,446,349, Aug. 3, 1948.
35. S. Z. HAIDER, M. H. KHUNDKAR, AND MD. SIDDIQULLAH, J. Appl. Chem. (London), 4, 93 (1954).
36. D. C. BRADLEY AND W. WARDLAW, J. Chem. Soc. 1951, 280.
37. V. L. HANSLEY, Ind. Eng. Chem., 43, 1759 (1951). 38. J. M. SHERFEY, Paper presented before the New York
meeting of the Electrochemical Society, April, 1953.
A P P E N D I X
PREPARATION OF TITANIUM (III) BOROHYDRIDE FROM TETRABUTYL TITANATE, Ti(OC4Hg)4,
AND DIBORANE, B2H6
Diborane (67.0 millimoles) was passed into a tetra- hydrofuran solution containing 29.0 millimoles of te t rabutyl t i tanate . ' A dark blue solution was formed and blue crystals sett led out. The solution was filtered in a He atmosphere and the crystals washed with petroleum ether. Yield as Ti(BH4)~: 26.1 millimoles. The filtrate was evaporated nearly to dryness, and a high boiling (225~ colorless liquid remained [the reported boiling point of t r ibutyl borate is 229~-230~ (35)]. This liquid had properties similar to those of boron esters, but the amount isolated was too small to s tudy thoroughly. With tetraisopropyl t i tanate, the borohydride was obtained in 90% yield.
Analysis :
Calculated for Ti (BH4)a.C4HsO (%) Found Ti 29.2 29.1 B 19.4 14.5 H 12.2 10.3 C 29.2 33.0
The crystalline blue solid underwent part ial decomposi- t ion at 135~176 and on heating to a higher temperature decomposed completely with the formation of a hright metallic mirror on the walls of the container. I t was in- soluble in the usual organic solvents except for te t rahydro- furan but reacted readily with peroxide-contaminated ethers and active solvents, such as water.
Ti(BH4)3 apparently is stabilized by coordination with tetrahydrofuran, as indicated by the facts that large crys- tals of the compound react rather slowly on exposure to the atmosphere (the finely divided compound occasionally ignites spontaneously), and the compound can be stored in dosed containers for long periods of t ime without noticeable decomposition,
REACTION OF DIBORANE WITH TIC14 AND ZrCh
When TiCl4 or ZrC14 reacted with diborane in tetrahydro- furan, the borohydride was not obtained. Crystall ine Ti and Zr compounds having the approximate empirical formula MC12.6(C4H~O)~.3 were isolated from these solu- tions.
Analyses: Ti compound: %Ti = 14.5, %C1 = 28.1, %C = 37.9, %H = 6.4; Zr compound: %Zr = 24.5, %C1 = 26.1, remainder organic material . The impure compounds con- tained a small amount of B. The Zr and Ti compounds were white and blue, respectively.
Compounds similar to those described here were obtained by reduction of TIC14 in te t rahydrofuran with LiH.
PREPARATION OF Zr(BH4)4
Preparat ion of this compound from KZrF5 and Al(BH4)a has been described by Hoekstra and Katz (27). Ini t ial ly,
1 Obtained from the du Pont Co.
) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66Downloaded on 2014-06-17 to IP
Vol. 104, No. 1 E L E C T R O D E P O S I T I O N O F M E T A L S 27
e ther solut ions of this compound were prepared in situ b y adding to an e ther solut ion the s to ichiometr ic quant i t i es of reac tants , according to e i ther of the following reac t ions :
ZrCI4 + 2AI(BH4)~ = Zr(BH4)4 + 2A1BH4C12 (I)
KZrF5 + 2AI(BH4)~ = Zr(BH4)~ + K F + 2A1BH4F~ (II)
Reac t ion (I) has the d i sadvan tage t h a t b o t h react ion products are soluble in e ther and difficult to separate .
Since Zr(BH~)4 and Al(BH4)~ are soluble, the l a t t e r con- t amina tes the p roduc t of b o t h react ions (I) and (II) if an excess is used. For th is reason react ion (II) was carr ied out wi th a s l ight excess of KZrFs. This K sa l t was used because a comparison of the re la t ive react iv i t ies of NaZrFs , KZrFs, and a p roduc t corresponding to KF-2ZrF4, showed t h a t KZrFs reac ted more rap id ly and completely.
When KZrF6 was used in excess, the yield was much less t h a n expected. An examinat ion of the products showed t h a t the react ion p robab ly proceeded by a stepwise mechanism involving the in te rmedia te format ion of a compound such as Zr(BH4)2F2, and for the reac t ion to go to complet ion, an excess of Al(BH4)3 mus t be used. Since b o t h react ions were unsa t i s fac to ry for the p repa ra t ion of an e thereal solut ion of Zr(BH4)4, o ther methods of prepar ing this compound were inves t iga ted .
I t was found t h a t Zr(BH4)4 could be prepared con- ven ien t ly by pass ing d iborane in to an e the r solut ion of a Zr es ter of the type Zr(OR)4. A p roduc t of th is re- act ion is p robab ly the corresponding boron ester , B(OR)3, as in the s imilar p repa ra t ion of Ti(BH4)a. This me thod is no t sa t i s fac tory for the p repa ra t ion of pure Zr(BH4)~ since the B ester and Zr(BH~)4 are b o t h present in the resul t ing solut ion.
The p repa ra t ion of pure Zr(BH~)4 was accomplished by the dry react ion of LiBH4 and ZrCh at room tempera tu re .
ZrC14 + 4LiBH4 = Zr(BH4)4 + 4LiC1 ( I I I )
This me thod is convenien t and rapid , and the yield is abou t 75% wi th respect to bo th reac tan ts .
Reaction of AI(BH,)a with KZrF~
A mixture of 25.9 millimoles of KZrF~ and 57.8 millimoles of AI(BI-I4)3 (mole ra t io 1:2) in 20 ml e ther was shaken for 16 hr . The solut ion was fi l tered and the resul t ing clear solu- t ion showed no fluoride. By means of hydr id e and Zr analy- ses, t he yield of Zr(BH4)4 was calculated to be 65%. When th is exper iment was repeated using the N a salt , the mole ra t io of 3NaZrFs:4AI(BH4)~, and 24 hr of shaking, the yield of Zr(Btt4)4 was only 30%, indica t ing fo rmat ion of in te r - media te z i rconium borohydrides such as Zr(BH~)_~F~_ owing to the use of insufficient AI(BH4)3
Reaction of Zirconium Tetraisopropylate with Diborane
Diborane (0.14 mole) was passed into a t e t r ahyd ro fu ran solut ion conta in ing 0.02 moles of te t ra iospropyl z i rconate (36). The resul t ing clear solut ion (34 ml) was analyzed for hydr ide and Zr content . I t conta ined 0.6 millimoles of Zr and 14.7 millimoles of hydr ide per mil l i l i ter (determined by evolu t ion of H). The theoret ical hydr ide con ten t for 0.6 millimoles of Zr(BH4)4 is abou t 10 millimoles. The ana ly t i - cal results indicate the solut ion conta ined Zr(BH4)4 and a small amount of diborane.
Preparation of Zr(BH4)4 from LiBH4 and ZrCl~
In an iner t a tmosphere chamber , 67.0 millimoles pow- dered ZrCl~, 272 millimoles powdered LiBH~, and some 0.5 em diameter dry Ni balls (to help mixing) were swirled b y
hand for abou t 15 min at room tempera ture . The mate r ia l in the flask became wet and hea t was evolved. The flask was shaken for about 5 min more unt i l i ts conten ts had be- come paste- l ike and the Ni balls were no longer able to produce mixing.
After cooling to room t empera tu re the flask was placed in a vacuum sys tem and evacua ted for about 30 min. The Zr(BH+)4 was collected in a l iquid N t rap . The yield was 77%.
When an excess of LiBH4 was used, the induct ion per iod was lessened and the yields were increased. Yields as h igh as 90% have been ob ta ined by using an excess of LIBH4 a n d / o r by b reak ing up the LiC1 and evacua t ing again at about 50~ The l a t t e r procedure increased the yield by 6% in one run. Analysis of the product gave 55% Zr, 30% B, and 10% H. The theore t ica l values for Zr(BH4)4 are 60.6% Zr, 28.7% B, and 10.7% H. The mel t ing poin t of the p roduc t was 28.7~ which is ident ical wi th the previously repor ted mel t ing po in t of Zr(BH+)4 (27).
For a metal l ic der ivat ive , the compound showed a sur- pris ing solubi l i ty in organic solvents , inc luding e ther , pe- t ro leum ether , and benzene.
A t t e m p t s to prepare Zr(BH4)4 in th is manne r by use of NaBH4 or KBH4 were unsuccessful.
PR~I',~RATION OF Zr(A1H4)4
To a di lute e ther solut ion of Zr(BH4)4 in a s in tered glass filter funnel was added an e the r solut ion of LiA1H~. The reac t ion was carr ied out in an a tmosphere of He. Af ter a sufficient amount of whi te solid was formed, the e ther solu- t ion was removed rapid ly and the solid was isolated. I t was uns tab le a t room tempera ture , decomposing wi th in several hours in to a pyrophor ie b lack solid. Analysis of the whi te solid ind ica ted a formula corresponding to Zr(A1H4)4.
Resul ts of analysis : wt. sample, 41.9 mg; 62 ml H_~ a t 26~ = 2.52 millimoles hydr ide ; 13.8 mg Zr = 0.15 milli- moles Zr; 17.6 mg A1 = 0.63 millimoles A1. This da t a would indica te an empirical fo rmula : ZrAl4.~Ht6.s.
PREPARATION OF PHENYLZIRCONIUM COMPOUND
Method A: On adding 3.7 millimoles Zr(BH4)4 to an e thyl e the r solut ion conta in ing 13.5 millimoles phenyl l i th ium, a b lack colored solut ion formed wi th evolut ion of heat . The solut ion was di lu ted wi th benzene and filtered. The f i l t ra te was then evapora ted to dryness and pumped under vacuum for I hr a t 50~ Dur ing the drying process a small amoun t of b iphenyl condensed on the upper wall of the conta iner . The resul t ing b lack solid gave a posi t ive Michler ' s ke tone tes t for a carbon to me ta l bond.
Method B: Pheny l l i t h ium was allowed to reac t wi th a s lurry of ZrC14 in e thyl e ther in a manne r s imilar to Me thod A above. The solut ion was evapora ted to a low volume, filtered, and the product (a t a n powder) p rec ip i ta ted by addi t ion of pe t ro leum ether . I t con ta ined 63% of the Zr in i t ia l ly used, gave a posi t ive Michler ' s ke tone tes t and the mel t ing poin t was 175~176
The analysis of the compound prepared by each me thod gave the following resul ts :
Method A Method B Theory for Zr (C~tts)~
Zr 27,5 29.6 28.3 C 49 42 67.0 H 5.3 4.7 4.7
The compound prepared by Method A conta ined abou t 8% Zr(BII,)4 as impur i ty , and the product f rom Method B con ta ined abou t 4% LiBr as impur i ty .
) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66Downloaded on 2014-06-17 to IP
28 J O U R N A L O F T H E E L E C T R O C H E M I C A L S O C I E T Y January 1957
T A B L E V. Reduction of TiCI~ in solution
Reducing agent
1. N a dispersion 2. Na dispersion 3. Na dispersion
4. Na* 5. N a - K t 6. LiH
Solvent
T e t r a h y d r o f u r a n E t h e r 50% te t r ahydro- fu ran 50% ben- z e n e
T e t r a h y d r o f u r a n Anisole T e t r a h y d r o f u r a n
Cone. of Ti in reacted soln. in g/1
Total J Reduced (as triva- lent Ti))
i 1.8 I 2.2
15 .0 i 5.8 i
15.4 24.1 5.0 0,0
10.2 8.3
* In exper iment 4, metal l ic Na was added to a t e t r ahy - d rofuran solut ion of TiCl~ in a steel bomb and the resul t ing solut ion was bal l -mil led a t 100~
t Reduct ion occurred in exper iment 5, bu t reduced Ti was found only in a b lack solid formed as a product of the react ion.
In two cases the reduced Ti con ten t ca lcula ted as t r iva- lent Ti ion was greater t han the to ta l Ti concent ra t ion . F rom this i t would appear t h a t a mix ture of di- and t r iva- lent Ti is formed in these reduced solut ions.
PREPARATION OF LOWER VALENT T i COMPOUNDS
Use of Ti(BItd)a in the Ti-A1 alloy b a t h to give a deposit conta in ing Ti suggests the use of organic ba ths conta in ing di- or t r iva len t Ti compounds. However, the avai lable lower va len t compounds, such as the halides, are no t soluble in ethers. They are soluble in more polar carbonyl and hy- droxyl types of solvents which are too reac t ive for purposes of deposit ion.
Lower va len t Ti compounds in e thers were prepared by reduct ion of TIC14 wi th the following reagents : sodium naph tha l ene (37), diborane, LiH, N a i l , Na, Na -K alloy, and dispersions of N a and Li in hydrocarbons . Only LiH and Na dispersion were used for a large "number of experi- ments .
Sodium naph tha l ene and TIC14 reacted in t e t r ahydro - furan solution. A qua l i t a t ive tes t ind ica ted t h a t t r i va l en t Ti was present . The resul t ing solut ion conducted well, bu t gave only a Na deposit .
Diborane reac ted wi th TiCl , in t e t r ahyd r o f u r an to form a green solut ion of TIC13 which did not conduct electr ici ty. After the solut ion had s tood for several days, a blue solid crystal l ized out [cf. prepara t ion of Ti(Bttd)3].
LiH and TIC14 did not react in a hydrocarbon solvent such as benzene, bu t addi t ion of e thyl e ther caused a re- ac t ion to occur wi th evolu t ion of hydrogen and format ion of a green color which changed to reddish-brown on fu r the r reduction. The green compound could not be isolated. The hydrogen ini t ia l ly present in the hydr ide evolved as gas, and the resul t ing solution, when allowed to react wi th water , gave no evidence of hydride. The reduced compound had propert ies s imilar to a lower va l en t Ti halide and was in no way like a hydr ide compound.
The reac t ion of N a i l wi th T iCh was similar to the re- act ion of L i t t except t h a t i t was much slower. This was p robab ly due to the lower solubi l i ty of N a i l in e thyl e ther . The react ion of Li, Na, or Na-K alloy wi th TIC14 in e ther solut ions appeared to give products s imilar to those ob- ta ined when LiH was used. Na and Li were used as disper- sions to obta in a more rapid and complete react ion.
The reac t ion of T iCh wi th LiH in e thyl e ther resul ted in the fo rmat ion of two l iquid layers. The denser layer was smaller in volume and r icher in Ti; on electrolysis i t con- ducted well bu t gave no metal l ic deposit . The upper layer
was low in Ti and conducted poorly. In the more s t rongly complexing ethers , t e t r ahyd ro fu ran and t e t r a h y d r o p y r a n , a separa t ion of the solut ion in to layers was not observed when TIC14 reacted wi th l i th ium hydride, bu t an insoluble solid usual ly formed. Al though the composi t ion of the solid product formed in these react ions was not de termined, re- duced Ti was detected in it. The concen t ra t ion of t r i va l en t Ti in the resul t ing solut ion was about 8-10 g/1 in b o t h sol- vents . This is close to the con ten t of t r iva len t Ti (7.7 g/ l) in a s a tu r a t ed solut ion of TIC13 in t e t r ahydro fu ran . These solutions yielded no metal l ic deposit on electrolysis al- t hough they did conduct well.
Na and Li in the form of dispersions reac ted wi th TIC14 in various e thers to yield solut ions con ta in ing lower va l en t Ti compounds. This is s imilar to the react ions of LiH as described above. Some of the react ions of Na dispersion with TIC14 in various e thers are l is ted in Table V. Similar results were ob ta ined wi th Li dispersion. Table V is a sum- ma~y of resul ts ob ta ined using various reducing agents and solvents. The mole ra t io of TIC14 to reducing agent is un i ty in all cases.
TIC14 in t e t r ahyd ro fu ran reac ted wi th Na dispersion to form a blue solid. Analysis of th is solid gave 12.7% Ti and 30.5% chloride, which has approximate ly the composi t ion TIC13. (C,HsO)~. A similar blue solid was formed when LiH reacted wi th TiCl4 in t e t r ahydro fu ran , bu t the composi t ion of this blue solid was not determined. The react ion of di- boranc wi th T i C h in t e t r a h y d r o f u r a n also formed a blue crystal l ine solid hav ing the approximate composi t ion TiCl~.~(CdH80)~.T. I t was p robab ly the same compound as above.
In addi t ion to the more act ive reducing agents l is ted above, meta ls such as Ni, Fe, Cu, Ag, or Mg were also ef- fective for reduct ion of TIC14 to t r iva len t Ti, bu t only in solvents such as d imethyl formamide , acetone, d imethoxy- e thane , and t e t r ahydro fu ran .
PREPARATION OF TiCI~
In an iner t a tmosphere , 0.9 mote T i C h was placed in a one l i ter bomb equipped wi th a glass liner. Four moles hy- drogen were added, and the bomb was kep t a t 500~ for 29 hr. The pressure was ma in t a ined a t about 2000 lb / iu . ~, which required t h a t some of the gas be ven ted occasionally.
At the end of the react ion the pressure was released and the bomb taken in to an iner t a tmosphere chamber . A yield of 16.5 g TIC13 as flaky red crystals was obta ined. The yield was 12% based on the amount of TiCl4 in i t ia l ly placed in the bomb. M o s t of the unreac ted TIC14 could have been re- covered.
In the above react ion, the bomb also conta ined small amounts of Pd and P t sponge in tes t tubes , bu t la te r ob- servat ions indica ted t ha t ne i the r of them aided the react ion.
A sample of the product was dissolved in water and found to conta in 1.5% insoluble m a t t e r ; the soluble por t ion was TiCl~. The insoluble mater ia l was bel ieved to be TiOCL The l a t t e r had formed previously when the same react ion had been allowed to cont inue for about an hour a t 450~ and 1650 l b / i n 3 In this exper iment a small amount of a gold colored mater ia l was formed. I t was insoluble in all t he usual acids and was not affected by the a tmosphere . This compounds has been previously described (17).
PREPARATION OF THE SALTS OF PERFLUOROACIDS Ti AND Zr SALTS
Ti and Zr sal ts were prepared b y react ion of the meta l te t rachlor ides wi th t r i f luoroacet ic acid. The p repa ra t ion of the Ti sal t i l lus t ra tes the method used.
To 1.9 g TIC14 was added 4.6 g tr i f iuoroacetic acid. The
) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66Downloaded on 2014-06-17 to IP
Vol. 104, No. 1 E L E C T R O D E P O S I T I O N O F M E T A L S 29
reaction mixture was heated at 65~ and evaporated to dryness. A yield of 4.5 g (90%) of white solid was obtained.
The Zr salt was prepared in a similar manner, but in a solution of ethyl ether. Both salts contained a small amount of Ct. Analysis of the salts for metal content gave: Ti, 17.6%; Zr, 25.2%.
Ni and Ag Salts : These were prepared by reaction of the
metal oxides with acids in aqueous solution and subsequent evaporation of the resulting solution to dryness. The salts were then dried at 150~ for several hours.
Nickel trifluoroacetate, silver trifluoroacetate, and silver perfluorobutyrate were prepared in this manner.
The solubility of silver trifluoroacetate was found to be 17.8 g/106 ml in benzene and 47.1 g/100 ml in ethyl ether
Electrodeposition of Metals from Organic Solutions
IV. Eleetrodeposition of Beryllium and Beryllium Alloys
GWENDOLYN B. WOOD AND ABNER BRENNER
National Bureau of Standards, Washington, D. C.
ABSTRACT
The electrodeposition of Be from nonaqueous media was studied. Be compounds used as solutes were the hydride, alunfinohydride, borohydride, halides, alkyls, and aryls. Their preparation, chemical properties, and electrochemical properties in nonaqueous media were investigated. Electrodeposits of Be (95% pure) and Be alloys, such as Be-B and Be-A1, were obtained from organic solutions. I t is believed that these represent the first metallic deposits consisting essentially of Be that have been obtained from organic baths.
Be has a low density (sp. gr. 1.82~ fairly high melting point (approx. 1280~ good corrosion resistance, and relatively high strength per weight ratio, which make it potentially valuable in structural engineering. However, because of its brittleness the pure metal has had little industrial use. Be is of interest in the field of atomic energy because of its property of slowing down neutrons with little tendency for absorbing them.
The literature concerning Be and its compounds is covered in several reviews (1-4). Be is produced com- mercially in the form of powder, flake, compact rod, or beads by the electrolysis of fused salt baths (5-10), many of which are operated near the melting point of the metal. The production of pure Be metal by thermal dissociation of BeI= or BeBr2 has been reported (11).
Electrolysis of Be salts in aeetamide was reported to give a black amorphous deposit which gave a qualitative test :for Be (12). Beryllium nitrate, sulfate, and halides were dissolved in fused alkylpyridinium halides, but no Be was deposited on electrolysis (13).
Booth and Torrey (14-16) made a comprehensive in- vestigation of the electrodeposition of Be using anhydrous Be halides, sulfate, nitrate, acetylaeetonate, beryllium ba- sic acetate, and sodium beryllium fluoride. They used a
number of nonaqueous organic and inorganic solvents and reported that a majori ty of the solutions conducted poorly. They reported electrodeposition of pure metallic Be only fl'om solutions of its compounds, particularly the nitrate and chloride, in liquid NHa on the basis that the deposit gave a qualitative test for Be. Their deposit did not dissolve in HC1, and they ascribed its nonreactivity to its purity. Without quantitative analytical determinations, conclusions as to the puri ty of deposits are questionable.
Several unsuccessful at tempts to duplicate their work have been made in this laboratory. Thin deposits of black, nonmetallic material soluble in dilute HC1 were obtained.
The purpose of this investigation was to develop methods of electrodepositing coherent coatings of pure Be or Be alloys. This involved a study of the chemistry of known Be compounds, the preparation of new Be compounds, and the investigation of the electrochemical properties of these compounds in nonaqueous solvents. A successful method of electrodepositing metallic Be from nonaqueous media other than from fused salts would have several useful applications, such as: eleetrowinning, improvement of the pm'ity and mechanical properties of the metal, production of eleetroplated coatings for protection of other metals against corrosion, .and eleetroforming of complicated shapes.
Since hydrogen is so much more noble than Be, i t is preferentially discharged from aqueous solutions. Conse- quently, this investigation was limited to organic solvents and fused salt baths.
Electrolytic properties of the following types of Be compounds were investigated: the hydride, alumino- hydride, borohydride, halides, alkyls, and aryls. Beryllium fluoride was not used because of its low solubility in or- ganic solvents. The halides are covalent in nature and :do not carry an electric current when fused. Other types of compounds do not lend themselves to fused salt eleetroly- sis, since they cannot be fused without decomposition.
The apparatus used for electrolytic studies and the purification of the solvents has been discussed (17, 18). The electrolytic investigations and preparation of com- pounds were performed under an atmosphere of dried A or He. A compound is considered soluble if it dissolves to the extent of 0.5 g/l.
) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 128.252.67.66Downloaded on 2014-06-17 to IP
