INVESTIGATION of the COMPATI Blllty of GOLD BRAZE ALLOYS with AEROZINE 50 Report 3332 8 November 1967

https://ntrs.nasa.gov/search.jsp?R=19680001044 2020-03-16T16:37:11+00:00Z

INVESTIGATION OF THE COMPATI BlLlTY OF GOLD BRAZE ALLOYS WITH AEROZINE 50

Report 3332 8 November 1967

Y I INVESTIGATION OF THE COMPATI BlLlTY OF GOLD BRAZE ALLOYS 3 WITH AEROZINE 50

Report 3332 8 November 1967

AE AL CORPORATION A SUBSIDIARY OF THE GENERAL TIRE & RUBBER COMPANY This experimental study, "Investigation of the Compatibility of

Gold Braze Alloys with AeroZINE 50," was conducted for the NASA Manned

Spacecraft Center by the Aerojet-General Corporation at Sacramento,

California. The investigation was performed under Contract NAS 9-7567

during the period from 29 September to 30 October 1967.

The experimental investigation was conducted by the Advanced

Propulsion Research Department, Dr. C. M. Beighley, Manager. The

Aerojet Program Manager was Dr. W. R. Fish. Contributors to the program were Dr. E. Me Vander Wall and Mre J. J. Bost of Fuels and

Combustion Research, and Mr. J. A. Curtis of the Analytical Chemistry

Laboratory.

The NASA Technical Monitor was Dr. W. R. Downs of the Structures

and Mechanics Division, Manned Spacecraft Center, Houston, Texas.

J i Y .. J c

TABLE OF CONTENTS r)

-Page I. Introduction 1

A. Objective 1

B. Technical Background 1

C Technical Approach

11. Summary

111. Technical Discussion

A. Materials and Methods

1. Braze Alloys

2* AeroZINE 50 3. Metal Content of AeroZINE 50 Exposed eo Alloys 7 4. Decomposition of AeroZINE 50 8

B. Experimental Results 11

1. Metal Content of AeroZINE 50 Exposed to Alloys 11 2. Decomposition of AeroZINE 50 13 C Discussion 14 PV Conclusions 1-7

TABLE AND FIGURE Table 1. Summary of Experimental Results 12 Figure 1, Manometric Gas Evolution Vessel 9

iv I. INTRODUCTION

A. OBJECTTVE

The objective of this investigation was to determine the

compatibility of two gold-based braze alloys with AeroZINE 70 rocket

I fuel. The heat exchanger used to heat helium wikh AeroZmE 50 in the

propulsion system of the Apollo Lunar Module is constructed of Type

347 stainless steel. The nickel-based braze alloy originally used in

the fabrication of this unit WE/S found to be unsuitable, and the gold- based alloys, Palniro 7 and Nioro, were selected as appropriate replacements from a structural standpoint. The investigation reported

here was undertaken to determine the appropriateness of the selection

from the standpoint of propellant compatibility.

B. TECHNICAL BACKGROUND Palniro 7 and Nioro are trade names of braze alloys of the following cQmposition:

Palniro '7 70% gold, 22% nickel, 8% palladium Nioro 82% gold, 18% nickel Because of the tendency of gold, nickel and palladium to form unstable

complex ions with ammonia and hydrazine, it is of considerable practical

1 importance to establish whether or not such ions can appear in AeroZINE 509 I c '9 a rocket fuel containing approximately equal parts of hydrazine and

unsymmetrical dimethylhydrazine, when it is exposed to the alloys. This

fuel mixture normally contains small amounts of ammonia and chloride ion,

the presence of which might be favorable to the formation of unstable

ammines and their hydrazine analogs.

-1- The following structures have been assigned to the ammines of gold, nickel and palladium chlorides:

The analogous hydrazine compounds are also shown. Hydrazinakes of these metal salts containing two or more molecules of hydrazine are known or theoretically possible. The complexes of gold with ammonia and hydrazine are of particular concern in the use of gold-based metals in AeroZPNE 50 because of the explosive nature of "fulminating gold", which can be made by the action of ammonia on auric chloride. It should be borne in mind, however, that the synthetic routes for the ammonia complexes of the metals are not via the metals themselves, but by the reaction of ammonia with their salts. The likeli- hood that explosive compounds of gold, nickel and palladium will appear in AeroZlNE 50 exposed to alloys containing these elements is therefore not strong. In addition, it should be kept in mind that the salts of gold and palladium are reduced to the metal by hydrazine; the salts of nickel are also reduced but at a very slow rate under ordinary circum-

stances. On the other hand, when such catalysts as palladium are present in alkaline ammoniacal nickel solutions, the formation of metallic nickel takes place rapidly. In view of these opposing reactions, it is

impossible to predict whether or not soluble compounds of gold, nickel and palladium will appear in AeroZPNE 50 exposed to alloys containing

these metals.

-2- C. TECHNICAL APPROACH The compatibility of the braze alloys with AeroZINE 50 was investigated by two methods. In each case experiments were conducted with 0.5% ammonia solutions in AeroZm 50 and with 0.05$ chloride solutions as well as with specification grade AeroZINE $0. One method was to determine the total quantities of gold, nickel and palladium in

AeroZINE 50 and in the ammonia and chloride solutions after they had been exposed to specimens of the alloys for 48 hours at 150°F. No attempt was made to identify the form in which the metals were present.

The purpose of the experiments was to establish whether or not the metals appeared in the fuel in sufficient quantities to constitute a potential hazard if they were in the form of explosively unstable compounds.

In the second approach experiments were conducted to determine the decomposition rate of the AeroZINE 50 and the solution$ during

0 their exposure to specimens of the braze alloys at 1.50 F. The rates

at which gaseous decomposition products were evolved by the fuel samples

stored in manometric vessels were used for this purpose. The principal

decomposition products of AeroZINE 50 are nitrogen, hydrogen, ammonia and

alkyl amines. The ammonia and amines have an appreciable solubility In

the fuel, but because the nitrogen and hydrogen are essentially insoluble, this manometric approach is a very sensitive method for the observation

of decomposition. Because finely divided metals and metal ions are

known to catalyze the decomposition of hydrazine and its'deriuatives,

the investigation of fuel stability in the presence of the alloys was

considered to be an important adjunct to the measurements of dissolved

metals.

-3- 1x0 SUMMARY A two-part experimental study was conducted to determine the compatibility of two gold-based braze alloys with AeroZINE 50. In the first part the total quantities of gold, nickel and palladium were determined by emission spectrography in AeroZINE 50, in AeroZINE 50 containing 0.5% ammonia, and in AeroZINE 50 containing 0.5$ chloride ion after these fuels were exposed for 48 hours at 150°F to specimens of Palniro 7 foil (70% gold, 22% nickel, 8% palladium) and Nioro foil (82% gold, 18% nickel). Neither gold nor palladium was present in detectable amounts in any of the fuel samples. Nickel was found in detectable amounts in all the fuel samples, including those not exposed to the alloys (controls) ., The amounts of nickel in the fuels exposed to Palniro 7 were not significantly different from those found in the controls, but significantly higher amounts were found in the fuels exposed to Nioro. The difference in the results is attributed to the protection of the nickel by the palladium present in the Palniro 7. In the second part of the etudy the compatibility of the two alloys with the three types of fuels was investigated by observing changes of pressure developed in a manometric vessel due to the evolution of gaseous decomposition products of the fuels. The rates of gas evolution at 150°F were less than the sensitivity of the method when the fuels were exposed to Palniro 7. In the parallel experiments with Nioro measureable gas evolution rates were observed which were attributed to the dissolved nickel cornpound(s)

-4 - The amounts of dissolved nickel compound(s) measured in the first series of experiments and the gas evolution rates observed in the second series were so small that it was concluded that exposure of AeroZINE 50 and its ammonia and chloride solutions to the gold- based braze alloys presents no hazard through the formation of unstable metal complexes with ammonia or hydrazine or through the loss of fuel by decomposition. The data obtained in this study provide no basis

for concluding that either alloy is incompatible with AeroZINE 50.

-5- Y

111. TECHPTICAL DISCUSSION A. MATERIALS AND METHODS 1. Braze Alloys

The braze alloys were used in both types of experiments in the form of 3-mil foil. After confirming its identify by emission spectrograph, each of th samples of foil was cut into specimens of equal sizeo The Palniro 7 specimens had a total surface area of 3.8 square inches, and the total surface area of the Nioro specimens was'

3.6 square inches. The specimens were then degreased with acetone and rinsed in electronic grade methanol (Fed. Spec, O-M-232d). They were then wrapped in tissue paper and placed in a vacuwn flask. After they had been subjected to a vacuum for 20 minutes, dry nitrogen gas was introduced into the flask and the specimens removed for use in the experimental work

2. AeroZINE 50

A T-liter sample of AeroZINE 50 was analyzed in accordance with Specification MIL-P-27402. The following values show that the fuel complied with the requirements of this specification:

N2H4 5106% mMH 47 05%

Water 0 06$

These values were obtained by gas chromatography. By the method used at Aerojet the water content is determined simultaneously with the hydrazine and UDHM, avoiding the calculation of the water content by difference. The balance is therefore ammonia and amine impurities. In a separate analysis by gas chromatography the ammonia content was shown to be 0.1%. The remaining 0.2% was probably aniline, a normal contaminant remaining from the azeotropic distillation of hydrazine, and traces of dimethyl and other alkylamines.

The chloride ion content was determined to be 0.008% by a turbidometric method calibrated with standards prepared by the dilution of an AeroZINE 50 solution of hydrazine hydrochloride with water.

Difficulties were encountered in the use of aqueous sodium chloride standards for the calibration of the turbidometric method.

A 3-liter sample of AeroZINE 50 containing 0.5% (nominal) of ammonia was prepared from the stock sample by adding gaseous ammonia.

The actual content was found in duplicate gas chromatographic analyses to be 0.48 and 0.49%. Hydrazine hydrochloride was added to 1.6 liters of the AeroZINE 50 stock sample to prepare a 0.05% (nominal) solution of chloride ion. The actual concentration was found by the turbidometric method to be 0.0495%.

3. Metal Content of AeroZINE 50 Exposed to Alloys

The specimens of Palniro 7 and Nioro were exposed to the AeroZW 50 and to the ammonia and chloride solutions in 25O-c~

Florence flasks fitted with ground glass stoppers. Before use the flasks were cleaned and passivated with a 50% aqueous solution of hydrazine. The flasks were filled with the solution and heated in a water bath at 175-195'F for one hour. They were then removed from the bath and allowed to stand overnight at ambient temperature. After the passivating solutions were removed, the flasks were rinsed 5 tinies with

-7- distilled water and 3 times with electronic grade methanol. !They were

then dried by evacuation to about 0.5 mm Hg pressure for 30 minutes.

The vacuum was broken with dry nitrogen gas, which was retained in the

flasks until they were used.

The rectangular braze alloy specimens were folded twice

to give them a W-shape, placed in the flasks, and covered with 50-cc

quantities of AeroZINE 50. After the stoppers were taped in position

to insure their retention, the flasks were placed in a constant

temperature bath at 15OoF for 48 hours. At the end of this period the

fuel samples were transferred to beakers and evaporated to dryness on

hot plates over a period of abut 5 hours. The resultant residues were

then dissolved in 5-cc quantities of 15% aqua regia, which contained

50 micrograms of molybdenum per milliliter as an internal standard.

One-cc portions of these solutions were analyzed for gold, nickel and

palladium by emission spectrography.

4. Decomposition of AeroZINE 50

The effect of the gold-based braze alloys on the

chemical stability of AeroZIXE 50 and of the ammonia and chloride

solutions was determined by the measurement of pressure developed in

the ullage space of the manometric vessel shown in Figure 1. The use

of this apparatus permits the measurement of the rate at which gaseous

products of decomposition of the fuel are generated, providing an index

of the catalytic effect of surfaces or dissolved materials that are

present in the fuel. The apparatus used is 8 modification of one

recommended by the Gel Test Methods Sub-committee of the ICRPG Working Manometric Gas Evolution Vessel

Figure 1 i Page 9 3 Group on Liquid Propellant Test Methods in CPLA Publication No, 123, October 1966. The quantity of gas evolved per pound of fuel per minute is expressed in cubic centimeters corrected to 77OF and one atmosphere of pressure

Before use the manometric vessels were calibrated with respect to total volume and then cleaned and passivated by the procedure described in the previous section. The subsequent preparations were carried out in a nitrogen-filled dry-box. The rectangular alloy specimens were coiled on a stainless steel rod and inserted into the

100-cc sample containers of the vessels. Then 50-cc volumes of the fuels were pipetted into the sample containers and 1.5- to 2.0-cc volumes were injected into the expansion bulbs. The latter measure was taken to balance the vapor pressures of the fuels at both ends of the manometers. The final operation, which was carried on outside the dry-box, was to seal the sample containers and the open ends of the manometers with an oxygen-gas torch. After the vessels were allowed

0 to equilibrate in a constant temperature bath at 1.50 F, manometer readings were made with a cathetometer every 24 hours. Since the ullage volumes were known from data on the total volumes of the vessels and the volumes of added fuels at the bath temperature, it was possible to calculate the quantities of evolved gases from the observed pressure changes, which were expressed in cc's corrected to 770 F and 1 atmosphere pressure and converted to rates by dividing by the weights of the fuel samples (lb) and by the elapsed times (min) in the bath. The limit of

-10- sensitivity of this method is about 1.0 x cc/lb-min, and the maximum uncertainty of the rate data due to temperature gradients in the baths is about 2 x cc/lb-min.

B. MPERIM~~NTALIRESULTS 1. Metal Content of AerQZINE 50 Exposed to ALloys The specimens Qf Palniro 7 and Wioro were exposed in triplicate experiments to 50-cc samples of the fu$ls, AeroZTNE 50, AeroZINE 50 containing 0.5% of ammonia, and AeroZIITE 50 containing

0.05% chloride ion. Two control experiments with 50-c~volumes of the respective fuel alone were conducted with each group of six experiments with the alloys. The exposure time and temperature in all cases were h8 hours and lfSO°F. This and other information on the conditions of the experiments are summarized in Table I, along with the total quantities of gold, nickel and palladium found in each case at the end of the exposure periods.

Neither gold nor palladium was pmsent in detectable amounts in any of the fuel samples. The limit of detection was 0.1 picrogram, which is equivalent to 2.2 x lQ-3 ppm in 45 gm of fuel

(50 cc x 0.899 gm/cc at 77'F). On the other band, nickel was found in all the fuel samples, including the controls, The average value for the six control aamples was 3.8 micrograms and the average value (3.3 micrograms) for the fuels exposed to Palniro 7 was not significantly different. These quantities are equivalent to a concentration of about

8 x lo1* ppm OF nickel in the fuel.

-11- 33-7 393 5-73000 T9-7000 000 000 rlrlrl rlrlrl rlrlrl rlrlrl xxx xxx xxx xxx MrlO 999 UfcOcucucu 999 ~rlcu rlrlrl rlrlrl A- vvv vvv

-12- The quantities of nickel found in the fuel samples exposed to Nioro were significantly higher than those found in the controls and in the fuel samples exposed to Palniro 7. In the experiments with Nioro and AeroZINE 50 the quantities of nickel varied widely, and it is possible that the largest of these was due to contamination.

The smallest of these, however, was 5-8 times the amount of nickel found in the control samples and in the fuel samples exposed to Palniro 7. In the experiments with Nioro and the fuel containing added ammonia the results were less variable, averaging 3lmicrograms, a value 8 times the average for the controls. The fuel containing added chloride ion contained the largest quantities of nickel. The average of the three values wa$ 440 micrograms, which is 100 times the average for all the controls.

2. Decomposition of AeroZINE 50

In this series of experiments specimens of Palniro 7 and Nioro were exposed in triplicate to 50-cc samples of each of the three fuels. Two control experiments with 50-cc volumes of AeroZINE $0 were conducted simultaneously. Controls experiments with the fuel

samples containing added ammonia and chloride ion were not conducted in this series. The exposure times in these experiments were 14 and 0 18 days; the temperature was 150 3’. The results of these decomposition

experiments are presented in Table 1 in parallel with the results of

the metal analyses discussed in the previous section.

The data on gas evolution rates are in accord with the data on nickel content. In the two control experiments and in seven of the nine in which Palniro 7 was exposed to the fuels, the rates were below the limit of sensitivity of the method. In the two experiments

-13- with Palniro 7 in which measureable rates were observed, one rate approached the limit of sensitivity and the other had a very low value. On the other hand, in the experiments with Nioro under conditions which led to the appearance of relatively high amounts of nickel in the fuels, relatively high gas evolution rates were also observed. In only one of the nine experiments with Nioro was a low rate observed. It is

clear from the results that the decomposition of AeroZINE 50 is caused

by the nickel-containing compound(s) formed by the attack of the fuels

on the palladium-free alloy, but there is no direct correlation between

the amounts of nickel found and the decomposition rates. When AeroZINE 50

containing added ammonia was exposed to Nioro, slightly higher rates of

decomposition were observed than when the Nioro was exposed to AeroZPNE 50

itself; however, even though the largest quantities of nickel were found

when Nioro was exposed to AeroZm 50 containing added chloride ion,

the decomposition rates of this material appeared to be no higher than

when AeroZSNE 50 was used.

C DISCUSSION

The data summarized in Table 1 show that of the metals under

consideration only nickel was present in the AeroZINE 50 used in this investigation. The source of this element is undoubtedly the stainless steel alloys widely used in the manufacture, transport and.handling of

this fuel and its components. Its presence shows not only that this

element is attacked by AeroZINE 50 but also that the compound(s) in

which it is present is not reduced to the free metal by the fuel. When the AeroZINE 50 was exposed to the palladium-free Nioro

(82% gold, 18%nickel) there was attack on the nickel in this alloy.

It is not clear that the addition of ammonia to the fuel had any effect, but it is obvious that the attack on the nickel was enhanced by the presence in the fuel of added chloride ion. On the other hand, the experiments with Palniro 7, which contains 8% of palladium, show that the palladium protects the nickel from attack, even when the fuel contains added amounts of ammonia and chloride ion.

The basis for the protection of the nickel by the palladium is not known, but this type of behavior has been observed previously,

For example, the addition of minor amounts of platinum to titaniun reduces the corrosive attack on this metal by sulphuric acid. For single phase systems such as the gold-based alloys studied in this investigation, the presumption is made that the attack by the AeroZINE 50 takes place at crystal boundaries where the nickel is locally con- centrated and anodic to the bulk of the metal. A means of explaining the protective effect of the palladium is to assume that the palladium alters this relationship by making the grain boundaries less anodic, or cathodic, to the metal matrix. Under these circumstances none of the three metals is attacked by the fuel.

Even though the nickel in the palladium-free alloy was attacked by the fuel, the extent of attack and the concentration of nickel in the AeroZINE 50 are of no consequence. The data obtained in this study provide no basis for concluding that either alloy is incompatible with the fuel. Consider, for example, the average amount of nickel (289-g) found in the AeroZINE 50 expcsed to Nioro in terms

-15- of the rate of corrosion. If the rate of attack of 289 ,ug/48 hrs were to continue for a year (-53 mg/yr) the corrosion rate of the Nioro would be about 5,6 X 10"' mils/year. The rate for the Nioro in AeroZXNE 50 containing 0.05$ of chloride ion would be only 8.5 X lo-' milslyr. When these amounts of nickel are considered in terms of concentrations in the fuels, 289 fig in the AeroZINE 50 and 440 Fgin the chloride solution exposed to Nioro are equivalent to 6.4 and 9.7 ppm, respectively.

If the nickel were in solution in the form of the trihydrazinate of nickel

C12, the equivalent concentrations for the complex would be 25 and 38 ppm. The presence in AeroZPrJE 50 of an unstable compound in concentrations of this magnitude is not considered to constitute a hazard, especially in view of the relative stability of the ammines of nickel salts compared to that of the analogous gold complexese

The data on the evolution of gaseous products indicate that the loss of AeroZIJKE 50 through catalytic decomposition is inconsequential

0 even at 1-50 F. If for purposes of simplicity it is assumed that the evolved gases are due to the decomposition only of hydrazine, the more unstable of the two components of AeroZPNE 509 it is seen that a gas evolution rate of 28 X 10 -4 cc/lb-min is equivalent to an annual loss of about O.25$ of the fuel. This is based on the following equation:

3N H -+- 2EJH: += 2N2 t- 3H2 24 3 and the assumption that all of the ammonia remains in solution in the fuel., A loss of this magnitude is within the error of current analytical methods for hydrazine and hydrazine-type fuels.

-16- IV e CONCLUSIONS

When AeroZINE 50 is exposed at 150°F for 48 hours to the two gold-based braze alloys, Palniro 7 (70% gold, 22% nickel, 8% palladium) and Nioro (82% gold, 18% nickel), neither the gold nor the palladium is attacked by the fuel to a detectable degree. This result is not affected by the presence of 0.5% of ammonia or 0.05% of chloride in the fuel. In the case of Nioro the nickel is attacked, especially by the fuel containing added chloride ion, but the palladium in the

Palniro 7 protects the nickel when this alloy is exposed to AeroZIm 50 and the ammonia and chloride solutions.

The amounts of nickel found in the fuel samples are so small as to be inconsequential, especially in view of the relative stability of nickel complexes containing ammonia and hydrazine. The concentrations of nickel in the fuels are less than 10 ppm and the corrosion rates of the Nioro are less than 0.1 mil/yr under the experimental conditions used in this study.

Measureable gas evolution rates were observed for fuel samples exposed to Nioro and these are attributed to the presence of nickel compounds. However, the highest rate observed is equivalent to an annual fuel loss rate of about 0.25%.

The data obtained in this investigation lead to the conclusion that the braze alloys studied are compatible with AeroZINJ3 50 even when excessive amounts of ammonia or chloride ion are present in the fuel.

The selection of Palniro 7 and Nioro for fabrication of the heat exchanger of the Apollo Lunar Module is appropriate from the standpoint of fuel compatibility.

-17-