Crystallinity of Calcium Hydroxylapatite Coatings
Crystallinity of Calcium Hydroxylapatite Coatings |
|
Plasma Biotal ltd has been involved in the development of calcium hydroxylapatite (HA) coating for orthopaedic and dental implants since 1987. During the last five last years coating procedures and measurements have been strictly controlled to obtain maximum crystallinity and the minimum of other phosphates. Over 550 X-ray diffraction (XRD) traces have been interpreted by independent laboratories relating to many thousands of implantable devices. A brief history and an indication of important aspects of quality assurance is given together with analysis of the major findings with respect to HA relative crystallinity and the presence of other phosphates.
The best available technique for analysis of HA powders seemed to be based upon comparison with the 'pure crystalline HA standard' XRD pattern published by the JCPDS (Joint Committee on Powder Diffraction Standards). The most sensitive indicator for thermally stable HA was found to be the presence of beta calcium orthophosphate peaks (Whitlockite or bTCP) in the XRD trace after heat treatment.
From the XRD traces it is possible to identify the presence of other phosphates and to obtain an indication of the relative proportions of crystalline and amorphous HA. A pure crystalline HA which matches the JCPDS is used as a standard reference. Care, and a strict adherence to designated procedures, is necessary in the preparation of the sample prior to XRD analysis as particle size and particle size distribution can effect the apparent crystallinity. Unless the conditions of measurement are fully documented and carefully controlled the results cannot be compared.
No quantitative statements were available on the precise meanings of 'as low as possible' or 'as high as possible'. In the absence of definitive data on these important criteria the company undertook a research programme to determine what was achievable using commercially available HA. Within 6 months it became apparent that many products that were called HA were, in fact, mixtures of HA with other phosphates. These mixtures were invariably thermally unstable and could not produce coatings of anything approaching 'pure HA'.
An investigation of the properties of existing commercially available HA coatings showed that generally the relative crystallinity averaged around 50% with the presence of proportionally large fractions of the amorphous phase. Some showed very large quantities of other phosphates present in their coatings. Possibly this was a design requirement but it is more likely that it was a result of the choice of an unsuitable starting material, or the use of incorrect settings whilst spraying, or a combination of both.
According to the literature (American Ceramics Society (1964) phase diagram) an HA which contained the correct stoichiometric ratio between calcium and phosphorus should be thermally stable to at least 900°C. The search for a suitable starting product was complicated by the then lack of any accepted standard for HA intended for plasma spraying.
As the search for a suitable starting material progressed it became apparent that certain HA products were stable for at least 1 hour at 1350 deg C as judged by the absence of detectable bTCP peaks after treatment. This test formed the basis for screening potential suppliers of HA powder and is still one of the obligatory criteria for acceptance of HA intended for plasma spraying.
Having obtained a thermally stable HA it soon became apparent that even this could be easily degraded by inappropriate setting of the many variables, (circa 30), associated with plasma spraying.
By 1988 the position regarding the coating of implants with HA was perceived by the company as follows :-
Given that only clinical data would finally determine the actual requirements for optimum coatings and that this would of necessity take some considerable time, the company made the decision that until evidence should be produced to the contrary then the objective should be to achieve the highest possible relative crystallinity and the lowest attainable levels of other phases. Should it prove necessary, eventually, to provide coatings with lower crystallinity or other phosphates present, this would be very easy to achieve.
The implications of this decision were incorporated within a Quality Assurance Manual and were thus formally included within the quality control of the company. The following were the major points relating to crystallinity and phosphates :-
It was recognised that the above provisions only covered limited areas of the total specification required for HA coatings. The measurement of the shear strength of the coating was also considered of major importance. From preliminary experiments it appeared that the precise testing method employed could greatly alter the measured results. A method based upon the British Standard (BS. 5350 part C15 1982), was chosen as the most appropriate. This method was incorporated within the quality assurance manual (QAM 10). This then became the standard method employed by the company in all shear strength testing. It is not claimed that the method chosen is necessarily the best, or that it will provide results comparable with other methods. What it does provide is a consistent 'baseline' to judge the relative performance of individual coatings. An analysis of shear strength is not described here but all coatings here included have shown greater than 25 MPa when measured by the described method.
Since all XRD traces show some degree of background noise, which consists of a multitude of miniature peaks, for the purposes of definition an 'unidentifiable peak' is deemed to be one which is significantly greater than the general background noise as judged by a qualified observer. To some extent this imposes a stricter control upon high purity coatings than upon lower purity samples as the higher background associated with poor coatings tends to mask spurious peaks.
Different applications required that other variables needed to be quantified and incorporated within the Quality Assurance manual. In order to do this in a coherent manner a section of each QA manual was set aside for the definition of customers exact requirements. Variables covered within this section include coating thickness, particle size of spray powder, surface roughness etc.
Since that time over 550 spray runs have been conducted (As of Dec 1997, 1750 Spray runs have been completed). These runs cover many thousands of orthopaedic devices. Every single one is traceable back to a specific XRD trace. All these XRD data are supplied to the customer as part of the release documentation or stored in case a request for the results are received. Copies of the data are also kept at the independent laboratory where it may be inspected by customers if required.
With the growth of the orthopaedic business the necessity to have assured access to consistent high purity HA became paramount. In order to undertake the production of HA powder of suitable quality and to ensure continuity of supply a new company, Plasma Biotal Ltd., was formed on 01 June 1991. All the existing HA coating work passed over to this company from Plasma Coatings Ltd. on that date. Both these companies form 'Plasma Group'.
In all 564 XRD analyses were undertaken with respect to devices intended for implantation in the period Jan 1989 to April 1994. From these 564 a total of 10 were rejected as failing to conform to the required standards as defined in QAM10 Iss 1 ie. crystallinity less than 75% (1 sample) and/or whitlockite greater than 2.0% (5 samples). The associated components were re-worked before release and despatch.
| % of all samples conforming | 98.2% |
| Average crystallinity of all conforming samples | 89.9 % |
| % average Whitlockite of all conforming samples | 0.12% |
If clinical data become available on a significant proportion of the thousands of devices which have been coated with HA which are covered by the above XRD analyses then further statistical analyses may yield interesting results. An indication of the desirability of high crystallinity and the presence or otherwise of other phosphates in HA coatings would be a major advance.
The results considered include all XRD analyses, (whether or not conforming as defined in QAM 10), undertaken for devices which were intended for implantation from the date of issue of QAM10 Iss1.
| Period | 1989-94 | 1989 | 1990 | 1991 | 1992 | 1993 | 1994 |
| No of samples | 564 | 81 | 88 | 83 | 130 | 146 | 48 |
| 1. Ave Crystallinity | 89.9 | 91.1 | 88.9 | 90.2 | 90.0 | 89.4 | 87.5 |
| Highest | 99 | 98 | 96 | 96 | 99 | 96 | 93 |
| Lowest | 70 | 79 | 70 | 74 | 63 | 75 | 79 |
| Range | 29 | 21 | 29 | 22 | 26 | 21 | 14 |
| 2. Beta Whitlockite (TCP) (%w/w) | 0.12 | 0.11 | 0.08 | 0.39 | 0.01 | 0.01 | 0.00 |
| Number with Nil Beta TCP | 497 | 63 | 66 | 67 | 118 | 145 | 48 |
| % of samples with Nil Beta TCP | 88.1 | 77.7 | 75.0 | 80.7 | 90.7 | 99.3 | 100.0 |
| Highest Beta TCP(%w/w) | 2.9 | 0.5 | 0.8 | 2.9 | 2.0 | 1.0 | 0.0 |
| 3.Samples containing Trace of Calcium Oxide phosphate(COP) | 339 | 21 | 60 | 50 | 60 | 102 | 46 |
| 4.Samples containing Trace of Alpha (TCP) | 435 | 1 | 61 | 59 | 126 | 146 | 42 |
| 5.Samples containing Trace of Calcium Phosphate (CP) | 113 | 0 | 29 | 8 | 50 | 23 | 3 |
6. Other phosphates present in trace quantities.- Trace quantities in this context are defined as identifiable peaks which represent quantities of less than 0.5% w/w. No samples contained more than trace quantities of any phosphates other than bTCP.
Phase Diagram - American Ceramics Society (1964)
J.C.P.D.S. HA data (Joint Committee on Powder Diffraction Standards)
Sintering Density versus Temperature data. A.Brandwood et al - Dept Biomedical Eng University N.S.W. Australia
Dehydration of HA versus Temperature data. A.Brandwood et al - Dept Biomedical Eng University of N.S.W. Australia.
Methodology for sample preparation of HA for Crystallinity analysis. D.Lowe, C.Roome Dept of Physics Staffordshire University.