Although the chemistry of phosphate esters was originally studied in the 1840s, their potential as fire-resistant fluids was not discovered until towards the end of WW II when it became necessary to increase the safety of hydraulic systems in military aircraft(1,2). Commercial airline applications followed, but in hazardous industrial applications polychlorinated biphenyls (pcbs) were initially preferred. While phosphate esters were used in blends with pcbs to reduce, for example, the very high density of the latter(3), phosphate-based fluids only became more widely used in the 1960s/1970s when pcbs were banned on toxicological and eco-toxicological grounds.
Subsequently two types of phosphates have found widespread use: trialkyl phosphates (e.g. tributyl phosphate) in aircraft hydraulic systems (because of a combination of adequate fire-resistance with very good low temperature properties), and triaryl phosphates for general industrial hydraulics with much better fire-resistance and stability (see Table 1). This is still largely the situation today, although the use of phosphates in some industrial applications has declined on account of their high cost and the availability of other (but more flammable), synthetic non-aqueous fluids (Table 2(4)).
Table 1: Typical Properties of Phosphate Esters Used in Hydraulic Fluid Formulations. *Within the same fluid type and same viscosity grade there may be differences in composition that can lead to different levels of performance.
Response to Changing Market Requirements
In over 50 years of use, the chemistry and technology of application - particularly of the aryl phosphates - has steadily evolved. The same generic description therefore hides changes to their composition and application in the following areas.
1) Base-stock Chemistry and Purity
Aryl phosphates were originally based on ‘phenols’ obtained from the distillation of coal tar, a complex mixture of cresols and/or xylenols. These were reacted with phosphorus oxychloride (POCl3) to produce the triaryl phosphate (e.g.trixylyl phosphate)(4).
Table 2: A Comparison of Fire-Resistance Properties for Non-Aqueous Fluid Types Used in Industrial Hydraulic Applications
In the 1970s, natural gas displaced coal as a domestic fuel and resulted in the closure of many of the tar-distillers. For an alternative feedstock industry turned to ‘synthetic’ alkylated phenols (e.g. isopropylphenols or tertiarybutylphenols) where the phenol was obtained from crude oil. Today, cresols and xylenols are also produced synthetically but all types of feedstock are used as the resulting phosphates have different characteristics leading to their selection depending on the application requirements(4).
Service experience early indicated that fluid life was dependent on new fluid quality, and particularly on initial levels of water, acidity and ‘chlorine’ etc. Refinements in the manufacturing process subsequently led to a steady improvement in purity. For example, while industrial fluids originally contained up to 500 ppm of ‘chlorine’(5), which resulted in servo-valve erosion, commercial products today typically contain about 20 ppm.
The aryl phosphates initially produced from cresylic acids were also neurotoxic. The cause was traced to specific cresol isomers and when these were greatly reduced or eliminated, as with the synthetic alkylated phosphates, the situation improved significantly(4,6). Currently, most phosphate esters meet the toxicity requirements for industrial hydraulic fluids laid down by the 7th Luxembourg Report or its successor, EN 14489(7) and the recommendations for handling are similar to those advised for mineral oil.
In aviation applications, efforts to reduce fluid weight and increase thermal stability have resulted in some changes to the base-stock composition but the need to maintain adequate fire-resistance combined with good low temperature performance has limited the options available.
2) Additive Chemistry
Industrial fluids initially used conventional additive mixtures but have slowly moved towards simpler, or even additive-free, products. The reasons for this included cost; the removal of some additives by the adsorbent solids used for controlling acidity levels, and the fact that some additives help stabilise foam formation in the fluid.
Base-stock chemistry has also played a role in determining the additive package depending on the phosphate structure. For example, some phosphates benefit from the use of antioxidants and metal passivators while other structures are very oxidatively stable, even when additive free.(4,8)
Additives to improve hydrolytic stability have also been tried but need frequent monitoring and replenishment, and in critical industrial applications in situ conditioning (see below) has been preferred.
As industrial fluid packages became simpler, the reverse was happening in aviation fluids. The need to eliminate servo-valve erosion and increase stability resulted in more sophisticated packages which included anti-erosion additives, hydrolysis stabilisers, improved oxidation inhibition etc. At one stage, water was being deliberately added to produce a ‘wet’ fluid in order to increase conductivity and avoid valve erosion!
3) Fluid Conditioning
The high cost of these fluids has resulted in a search for ways to extend their life – particularly in general industrial applications. As the acid formed by hydrolysis catalyses further degradation, controlling the fluid acidity effectively increases their life4. Early attempts at removing acid in situ involved circulating the fluid through fuller’s earth. However, this caused further problems as metal soaps or salts were formed which adversely affected fluid properties and eventually produced sticky deposits that fouled servo-valves etc(9,10,11). Other solids have been tried but the latest technology involving ion exchange resins has been extremely successful(12,13,14 and Figure 1(15)). Using this treatment the life of the fluid can now approach or even exceed that of the equipment in which it is used(16).
The concept of in situ fluid conditioning has not been extended to aircraft due to weight and space restrictions and could really only be applied during overhaul.
Solids treatment is, however, an added cost and ways of reducing the amount of solid used have also been investigated. The effects of drying and de-gassing the fluid have been very successful in assisting degradation control(14,17); the latter because air causes fluid oxidation with the generation of yet more acid.
Figure 1: An example of the reclamation of degraded phosphate ester using ion exchange resin treatment
4) Fluid Monitoring
The need to accurately monitor the condition of the fluid in use requires a variety of tests to examine critical properties(18,19). However, as sampling and external testing is time-consuming and expensive, some on-line procedures are now being used (e.g. water and particulates), but until all the important parameters can be checked on-line it will be still be necessary to test samples at specified intervals. Long term the goal must be to monitor the condition of the fluid remotely - perhaps by the fluid supplier - and for recommendations on fluid maintenance to be sent electronically to the user.
Since their introduction over 50 years ago, significant changes in phosphate ester composition have occurred. These have resulted in fluids that are of higher quality, for example, they are more stable and less hazardous. The use of fluid conditioning to control their degradation and extend fluid life has also been very successful. Although fluid development will continue, the main emphasis in future will be on integrating improved condition monitoring with degradation control in order to further enhance life-cycle costs.
W. David Phillips is a life member of STLE. You can find his contact information in our membership database.
Other articles in this issue: