Skip to navigation – Site map
Dossier

Testing phenol formaldehyde composite insulation material for conservation purposes

Tabea Rude

Abstracts

This investigation related to the degradation of phenol formaldehyde based electrical insulation materials in close mechanical contact with CZ108 brass in a Post Office Type 36 time transmitter, an electromechanical clock. Through testing of the two material types, the project exposed differences in degradation properties due to different manufacturing methods of the plastic component. Preventive conservation advice was then synthesised from the outcome.

Top of page

Full text

I would like to thank my tutors Lorna Calcutt and Matthew Read for their insightful comments and continuous support throughout the project. My sincere thanks also goes to Professor Norman Billingham, Dr James Nye, Derek Bird, John Butt, Simon Taylor, Kenneth Kobb and Jon Colombo for sharing their knowledge, insulation materials and/or documents with me.

Introduction

1An increasing amount of industrial heritage is being incorporated into museum collections. New, material-related problems occur as a result. Some of these arise from the diversity of materials in combination with plastics. These are often stored in poorly-ventilated or effectively sealed environments.

2The project used the Type 36 time transmitter (hereafter referred to as Type 36) as a case study object to explore previously identified problems. The Type 36 is a British, large-scale production item made between 1925 and 1981 (COWSILL 1997:2).

3Degradation and long-term inter-material effects appear especially prominently where phenol formaldehyde-based insulation material is in close contact with copper zinc alloys (brasses). Degradation affected both materials and presents itself as corrosion on metal and crystals on plastic.

4The project investigated the degradation of phenol formaldehyde-based plastics reinforced with three different materials; paper, fabric and wood flour. These plastics were tested in close mechanical contact with the copper zinc alloy CZ108.

5Original Type 36 electrical insulation materials were donated by several clock collectors for destructive testing. This provided samples of the three differently reinforced types of plastic material from twelve different clocks, manufactured over four decades (1930s, 1940s, 1970s and 1980s). As a subsidiary question, this project also considered the oak clock case micro-environment as a contributor to the corrosion processes.

6The Type 36 provided one-second, 6-second and 30-second impulses for up to 10,000 telephone exchanges and numerous clock dials at a time (BRITISHTELEPHONES.COM 2015; KOBB 2015). During their service life, component parts were exchanged and updated regularly to ensure reliability. Today, Type 36 components are starting to indicate wear and material fatigue due to age, long-term use, and storage.

Title: Fig. 1 The Post Office Type 36 case study object

Title: Fig. 1 The Post Office Type 36 case study object

On the left hand side the illustration indicates the possible connection to numerous clock dials and telephone exchanges. On the right hand side the phenol formaldehyde based insulation materials SRBWF, SRBP and SRBF are shown in blown-up view.

Credits: Tabea Rude (photo), Michael Goldrei (editing).

7The project’s intended aim was to identify the materials, to carry out practical experimentation, and to identify the degradation factors, notably the corrosion products and their effects on identified materials.

8The expected outcome was to formulate advice for conservators on options for treatment of existing corrosion products, and support preventive conservation strategies such as removal of cumulated corrosion products that further accelerate anticipated corrosion and leaving the clock case open to counter off-gassing and control relative humidity.

Phenol formaldehyde composites

9Phenol formaldehyde resin is used in the Type 36 as an electrical insulator. It is either paper-reinforced resin [Synthetic Resin Bonded Paper], fabric [Synthetic Resin Bonded Fabric] or wood flour [Synthetic Resin Bonded Wood Flour], forming a ‘fibre reinforced plastic’[FRP] (SHODHGANGA.INFLIBINET.AC.IN 2015)

10The paper and fabric in SRBP and SRBF offer continuous fibre reinforcement (PRESSPAHN.COM 2015). This material was produced in sheet or rod form. A disadvantage of using this material is that machining or cutting exposes reinforcing fibres to air, like end-grain in wood allowing the ingress of water vapour. SRBWF is a particle reinforced resin which is moulded specifically for application (SHASHOUA 2008:62).

11In manufacture ‘The first stage of the poly-condensation reaction [of phenol formaldehyde] involves the reactions of formaldehyde at the ortho- and para positions of the phenol ring to make mixtures of methylol (-CH2OH) phenols, known as resols:’(BILLINGHAM 2015) which are made with phenol and a molar excess of formaldehyde. The two components are mixed with water and a catalyst.

Fig. 2 The first stage of the poly-condensation reaction of phenol formaldehyde

Fig. 2 The first stage of the poly-condensation reaction of phenol formaldehyde

Reactions of formaldehyde at the ortho- and para positions of the phenol ring.

Credits: Norman Billingham

12‘At this stage fillers can be added and the mixture can be impregnated into paper, fabric, wood flour or other reinforcements and moulded to a desired shape. Further heating under pressure leads to the formation of a highly crosslinked network in which the phenol rings are joined by methylene (-CH2-) and methylene ether (-CH2OCH2-) bridges’ (BILLINGHAM 2015).

Fig.3 The second stage of the poly-condensation reaction of phenol formaldehyde

Fig.3 The second stage of the poly-condensation reaction of phenol formaldehyde

Under heat and pressure the phenol rings are joined by methylene and methylene ether bridges.

Credits: Norman Billingham

13The choice of catalyst and molar ratio influence degradation properties (SALTHAMMER ET ALIA 2010:7). Morphology, molecular weight and degree of polymerisation can be tailored through catalyst and molar ratio choices and desirable properties can be targeted (PLENCO.COM 2015).

14In degradation, the polymerisation process reverses. Chain scission occurs and the crosslinked structure resolves to a linear structure comprising a complex combination of monomers. Having free reactive molecules, the monomers react with other molecules (e.g. water or oxygen). Hydrolysis and oxidation cause further degradation. The cellulose-based filler (e.g. paper, wood flour or fabric) swells through hydrolysis. Cracks or distortions emerge (SHASHOUA 2008:157).

15Modern phenol formaldehyde resins are believed to be stable, owing to highly crosslinked molecular structures. Earlier manufacturing techniques did not benefit from the same understanding, explaining possible flaws in mid-20th century resins.

16Degradation is possible not only in the resin, but also its reinforcements, notably on cut edges of SRBP and SRBF.

Copper zinc alloys in the multi-material object Type 36

17A survey as part of preliminary investigations of the Type 36 showed that one type of brass was used predominantly. This aligns with the British Standard CZ108 (modern name: CuZn37), containing 63% copper and 37% Zinc (COPPER DEVELOPMENT ASSOCIATION 2004).

18If a brass contains 15% or more zinc, the alloy is susceptible to dezincification under circumstances of elevated oxygen and moisture levels. Dezincification is an ‘intergranular corrosion’. Corrosion starts at the grain boundaries of copper and zinc. It results in a selective removal of zinc as the least noble element of the alloy and leaves porous copper behind (SELWYN 2004:32).

19In respect to the Type 36, volatile compounds or susceptible polymers from any of the components in the clock can react with the copper ions in the brass and cause its corrosion. The copper acts as a catalyst, accelerating the process of its corrosion and the degradation of neighbouring materials such as phenol formaldehyde insulation material.

20As a result, the long polymer chains in the Type 36 phenol formaldehyde-based electrical insulation material experience chain scission and lose molecular weight. The degradation of organic material accelerates, and the materials in contact become fragile and brittle.

21The corrosion product, copper formate, formed of copper and formic acid, deposits at the material borders. Its properties attract water and therefore cause further degradation (see Fig. 4).

Fig. 4 Detail of an electrical clock

Fig. 4 Detail of an electrical clock

Phenol formaldehyde composite degradation and its effects on brass.

Credits: Mostyn Gale

22Although not hermetically sealed, emissions are contained within the closed oak clock case. Thickett and Lee state that concentrations of trapped gases increase over time and further instigate degradation (THICKETT AND LEE 2004:3).

Accelerated degradation as a method of investigation

23Stability of a material can be assessed by exposing it to exaggerated environmental conditions, i.e.: elevated levels of moisture, heat or light, etc. This process is called accelerated or forced degradation.

24The project focussed on the development of chemical properties during accelerated degradation. Results were also used to give further indications on physical change and inter-material effects.

25Degradation can be artificially induced by photochemical, thermal or hydrothermal accelerated degradation. During preliminary investigations all three of these methods were tested. To refine the project, it was decided to concentrate on the latter, accelerated hydrothermal degradation (exposure to elevated temperature and relative humidity levels). To avoid molecular changes in the tested material phenol formaldehyde, its glass transition temperature of 170°C was taken into account and all tests were carried out well below that value.

26Two different types of artificially accelerated hydrothermal degradation experiments were carried out. The first one concentrated on the material itself, while the second one also took the identified corrosion product copper formate into account.

27The aim of the project was to investigate whether degradation differs in materials from different decades of manufacture. Samples were exposed to controlled conditions of elevated temperature and relative humidity levels. Degradation then was assessed with FTIR spectroscopy before and after the accelerated degradation experiment.

28Fourier Transform Infra-Red spectroscopy with Attenuated Total Reflection accessory was used to obtain control samples of the three different types of phenol formaldehyde based composite plastics used in the Type 36 for plotting them against the degraded samples. Following accelerated degradation experiments a second spectrum was obtained of every sample in order to compare and assess chemical changes in the material.

Assessing chemical change through testing

  • 2 Not all material was used in the final tests.

29For the first experiment, material was sourced from twelve different Type 36 manufactured between 1930 and 1980.2

Fig. 5 The 1970s Type 36 and its three main testing materials

Fig. 5 The 1970s Type 36 and its three main testing materials

The flake samples of the testing materials SRBP, SRBF, SRBWF are visible in the blown-up images, with the lines pointing at the points of where they were harvested from the case study object.

Credits: Tabea Rude

30Sampling was carried out by scraping phenol formaldehyde composites with a scalpel. Flakes were placed in a glass petri dish. Three samples of each composite material with a minimum of 1.5mm diameter and a thickness of ca. 0.5 – 0.75mm each were analysed by FTIR spectrometry and compared to the corresponding control sample.

31Samples were then divided into three glass vials and three brass sample dishes. A minimum of ca. 0.1 gram sample material was placed in each vessel.

  • 3 For the 1930s only one type of Fibre Reinforced Plastic (SRBP) could be sourced, three jars with si (...)

32Six samples from each available material (see Fig. 6) were generated, documented and prepared in the same way as described above resulting in three samples held in glass vials and three samples held in brass sample dishes. Respectively, one glass vial and one brass sample dish containing the same sample material was placed into a wide mouth soda lime glass ointment jar fitted with an urea cap. The materials SRBP, SRBF and SRBWF from the same decade were placed in the same three jars.3

  • 4 Test materials were chosen due to their availability and validity, therefore some decades are repre (...)

33One jar for each decade (two jars for the 1980s decade4) was equipped with 5 g calibrated silica gel to create a dry environment; one jar contained 0.05ml of deionised water; to create an average of 30-40% relative humidity and the third jar contained 10ml of deionised water to create 100% relative humidity.

Fig. 6 Tested phenol formaldehyde composites from the 1930s, 1940s, 1970s and 1980s.

Fig. 6 Tested phenol formaldehyde composites from the 1930s, 1940s, 1970s and 1980s.

Clocks are named after their owner (Nye clock, Butt clock), their manufacturer (Magneta, Gents) and/or year of manufacture (1930s, 1970s).

Credits: Tabea Rude

34The 15 jarswere then placed into a Size 2 Gallenkamp Hotbox Oven with a circulation fan and remained there for 12 weeks at a constant 70 °C. The chosen time frame was the maximum test period available within the project constraints. The temperature was chosen after discussion Professor Norman Billingham and David Dorning (West Dean science tutor).

Fig. 7 Sampling example of the 1940s jar exposed to high RH

Fig. 7 Sampling example of the 1940s jar exposed to high RH

SRBP, SRBF and SRBWF material flakes in three glass vials and three brass sample dishes from 1940 in high RH (achieved by glass vial filled with 10ml deionised water).

Credits: Tabea Rude

35The aim was to assess chemical change(s) by FTIR, in light of supporting literature (DERRICK ET ALIA 2015; HASLAM AND WILLIS 1965) and correspondence (BILLINGHAM 2015). Furthermore, chemical changes were correlated to physical changes by the author, supported by polymer scientist Professor Norman Billingham in order to generate further information on practical conservation treatments which could be applied to multi-material objects like the Type 36.

Investigating the corrosion product

36Initially the corrosion products taken from three different substrates (bond line areas of copper and copper zinc with SRBP and SRBF) on the Type 36 were analysed by FTIR spectroscopy and compared with a spectrum of pure copper formate. This provided confirmation that the corrosion products were in fact copper formate, as initially indicated by their blue colour.

37In order to investigate what type of damage copper formate causes, and how fast it occurs and migrates, for the second experiment, copper formate was placed in equal quantities of 0.1 gram into purpose made shallow recesses (sample dishes) in both SRBP (three samples), brass (three samples, see Fig. 8) and combination test sample dishes of SRBP and brass riveted together (three samples, see Fig. 8, left hand side).

Fig. 8 Sampling example for the copper formate accelerated degradation experiment

Fig. 8 Sampling example for the copper formate accelerated degradation experiment

Combination test sample dishes made by riveting SRBP and brass with steel rivet and fill recess with 0.1 gram of copper formate on the left hand side. Jar No. 6 illustrated on the right hand side showing combination test sample dish and SRBP dish each filled with 0.1 gram of copper formate in mason jar with glass vial holding 10ml deionised water to create high humidity.

Credits: Tabea Rude

38The samples were then placed into Kerr™ wide mouth mason jars (560 ml). Two of the jars (Fig. 9: Jar 1 and Jar 4) were equipped with 5 g calibrated silica gel to create a dry and acid-free environment, two jars contained 0.05ml deionised water to create a medium-dry environment (Fig. 9: Jar 2 and Jar 5) and the last two jars were equipped with a glass vial holding 10 ml of deionised water to create a very humid environment (Fig. 9: Jar 3 and Jar 6). Brass samples (Fig. 9: Jars 1, 2, 3) were placed in separate jars to investigate how copper formate reacts with brass on its own, without emissions from SRBP. The SRBP sample dishes and the combination test sample dishes were placed in Jars 4 (low RH), 5 (medium RH), 6 (high RH) (see Fig. 9). The six mason jars were then placed into the Size 2 Gallenkamp Hotbox Oven with fan and remained there for 30 days at constant 70 °C temperature.

Fig. 9 Set-up of copper formate accelerated degradation experiment

Fig. 9 Set-up of copper formate accelerated degradation experiment

Photos of results without magnification and using a Nikon© AZ100 zoom microscope are shown below the corresponding jar. Microscopic views of riveted brass and SRBP sample dishes were extremely similar and therefore are not shown in this table.

Credits: Tabea Rude (photos), Michael Goldrei (editing)

39As chemical changes of phenol formaldehyde materials were already under investigation in the first accelerated degradation test, this test was intended to focus on physical changes that may be visible to the naked eye. The test was also intended to provide information about the effects of copper formate on the surface of the substrate materials (SRBP and brass) rather than the material structure, leading to inform the conservator about types of damage which can be expected from the reaction between formaldehyde and copper and, how to identify it.

Experiments: Assessing chemical change

40The experiment was set up as described above. Problems with sample preparation were encountered, as any exposure to air movement risked sample loss because the flake samples were very low mass. Flake sampling technique did allow the use of very small amounts of sample material, which reduced the damage and loss of original material. This sampling method resulted in good quality spectra. (DERRICK ET ALIA 1999:88).

41The oven temperature, which was checked on a regular basis remained stable at 70°C. However, during week six, most of the urea lids cracked, which meant that the calculated degree of humidity was lost. As it was not intended to achieve a specific age in terms of translating the experiments duration into real time ageing, the results were still valid.

42After 12 weeks the experiment was terminated. All samples were exposed to FTIR spectroscopy again.

43The brass sample dishes were scanned with a Canon™ CanoScan LiDE 110 flatbed photo & document scanner (2400 x 4800 dpi resolution). This was done to investigate whether a colour difference could be detected between brass sample dishes exposed to different humidities. The scanner was chosen as a tool to ensure equal lighting conditions. Further colour measurements would have exceeded the time frame of this project.

44When opening the jars after the experiment, a distinct formaldehyde smell emanated from the jars exposed to high RH indicating significant off-gassing. The jars exposed to medium RH also off-gassed to a more limited extent. The sample jars equipped with silica gel did not smell at all. These observations confirmed that high RH increases the amount of off-gassing and the use of a scavenger/absorbent such as silica gel does not only buffer the relative humidity, but may absorb also the formaldehyde emissions to an extent. This was verified by Shashoua (SHASHOUA 2008:199).

45The colours of the brass sample dishes containing the flake sample material were compared visually after scanning under equal conditions. The brass sample dishes exposed to high RH were tarnished more darkly than the brass sample dishes exposed to low or medium RH. A distinct difference between low and medium RH could not be confirmed.

46The following spectral analysis and interpretation was carried out with the support of Professor Norman Billingham. Firstly, it was found that the analysed resins ‘are consistent with the expected structure. The broad intense peak in the region 3600 cm-1 to 3100cm-1 is associated with the OH-group attached to the phenyl ring in the resins and to OH groups from unreacted methylol groups. The peaks between 3000 and 2800cm-1 are due to stretching vibrations of C-H groups both aliphatic and aromatic. The peak at around 1200cm-1 is associated with the vibration of ether linkages. The double peak at around 1600cm-1 arises from vibrations of aromatic rings and bands between 1600 and 1200 cm-1 are skeletal vibrations of substituted phenols’ (BILLINGHAM 2015) (Fig. 10).

Fig. 10 Normalised spectra of SRBWF held in brass sample dishes from 1980s clock

Fig. 10 Normalised spectra of SRBWF held in brass sample dishes from 1980s clock

SRBWF exposed to low (red graph, S37b), medium (pink graph, S38b) and high (blue graph, S39b) humidity plotted against a control sample taken before artificially accelerated degradation experiment (black graph, SC 37, SC38, SC39). X-axis showing infra-red light transmittance in % versus frequency on the y-axis in reciprocal centimetres (cm-1).

Credits: Tabea Rude

47Additionally, spectra of early phenol formaldehyde composites (1930s and 1940s) shown in illustration 11 also show strong absorptions just above 1000 cm-1. A band at 1010cm-1 is attributable to methylol groups and its presence suggests incomplete cure of the resin (PILATO 2010). ‘However the bands here are too broad and intense to be typical of phenol formaldehyde resin and it seems likely that they come from some sort of filler rather than the resin itself. Silicone dioxide bonds in materials like asbestos show broad and strong absorptions in this region and it is tempting to speculate that there is some such filler present.’ (BILLINGHAM 2015). In the future these assumptions could be verified using GC-MS analysis, which was not available in the set time frame of this project.

Fig. 11 Spectra of SRBP samples produced in the 1940s

Fig. 11 Spectra of SRBP samples produced in the 1940s

Loss of OH groups visible in SRBP samples held in brass dishes and exposed to high humidity (compare to ill.6). Control (red graph, SC4), material exposed to low humidity (black graph, S4b), medium humidity (pink graph, S5b) and high humidity (blue graph, S6b). X-axis showing infra-red light transmittance in % versus frequency on the y-axis in reciprocal centimetres (cm-1)

Credits: Tabea Rude

48These findings were confirmed by literature, as access to XRD, TEM and Raman spectroscopy was impossible at the time of testing. The use of asbestos as a cheap and fire-resistant filler was common in early phenol formaldehyde resins (ASBESTOS.COM 2015).

49For spectral analysis and interpretation, the spectra taken from one single material divided into three different brass sample dishes, each of them exposed to a different relative humidity, were plotted against each other and the same procedure was carried out for every single material held in a glass vial. In a second stage, the control sample of each material was plotted against a sample contained in glass and in brass. This was executed for all three relative humidities in order to spot possible differences of using different sampling substrates.

Results: Assessing chemical change

50The accelerated degradation experiment showed that the phenol formaldehyde-based plastics are indeed chemically relatively stable, as little change in intensity of absorbance bands and peaks were observed. It therefore confirms the properties expected from a highly crosslinked polymer structure.

51However, one sample material used in the clock dating from the 1940s showed a significant change in the OH-NH region between 4000 and 3000 cm -1 (see circled area in Fig. 11). A ‘broad-envelope-type band centred around 3400cm -1 (DERRICK ET ALIA 2015:93), which was visible in the control spectrum, disappeared in all six aged samples.’ ‘The fact that the peak is large in the control sample and disappears on ageing shows that some material absorbing in this range is being lost on ageing’ (BILLINGHAM 2015). It is tempting to interpret this phenomenon as a loss of absorbed water, which could be due to incomplete curing of the phenol formaldehyde resin. Incorrect curing can occur locally through ‘contaminants or poor mixing, but it is more likely to occur throughout the whole bond line, because of incorrect formulation or mixing, or thermal exposure’ (TWI-GLOBAL.COM 2015). To confirm this theory further analysis with analytical tools other than FTIR are necessary.

52Though the majority of the composite samples proved to be chemically stable, preliminary tests described in the methodology showed the problem of its delamination, due to swelling of the reinforcements paper, fabric or wood flour or malfunctions during polymerisation process. The swelling does not necessarily cause any chemical changes in the highly crosslinked resin, but it does present a physical problem, as delamination can mean malfunction. These problems occur especially on composites with continuous fibre reinforcements such as SRBP and SRBF. Their fibres are exposed to the environment as soon as they are machined (see Fig. 1) opposed to SRBWF, which is moulded specifically for its application.

Practical applications for preventative conservation measures

53In terms of conservation practices, the risk of delamination of the reinforcement materials can be reduced by keeping the temperature and RH stable and thus reducing fluctuations in volume of the reinforcement materials (paper, fabric, wood flour). Shashoua recommends a relative humidity of 55 ± 3 per cent and a temperature of 18 ± 2°C, in addition to ‘good ventilation’ (SHASHOUA 2008:195).

54Even though the material used in the Type 36 is not exposed to physical stress and has only a passive function, a delamination would increase the risk of material loss. Furthermore, an advanced stage of delamination could limit its electrical insulation properties.

55Despite the known and confirmed stability of phenol formaldehyde resins it is reported that corrosion products on bond lines of copper zinc and copper with plastic do appear on a significant percentage of electromechanical clocks containing these composites. ‘Since these products are usually blue, they would appear to be copper salts (or copper-zinc salt mixtures) produced by acid attack on the brass. It was tempting to speculate that these salts may be hydrated formates formed from formic acid deriving from oxidation of formaldehyde released from the composite material. Formaldehyde evolution from phenolic composites and adhesives has been a significant concern in recent years and modern composites are made under conditions which minimise formaldehyde release. However, historic materials were produced with far less care for complete cure and release of free formaldehyde’ (BILLINGHAM 2015)

56An increase in air movement in the clock case could reduce an accumulation of formaldehyde vapours and any other potentially harmful indoor pollutants or gasses. The ventilation of a clock case should especially be considered for objects that already show corrosive products on phenol formaldehyde composites and copper alloys. In practice, this could be achieved by keeping the clock door ajar or open overnight on a regular basis.

57Health and safety issues due to formaldehyde in a room with average ventilation are unlikely, increasing levels can be detected easily by its pungent smell and reduced by increased ventilation.

Testing the corrosion product

58Before the second thermohydrolytic accelerated degradation experiment corrosion products observed on the Type 36 brass and SRBP/SRBF insulation material were compared with pure copper formate using FTIR spectrometry.

59Following the identification of the corrosion product, the experiment was carried out as described above. Results were obtained visually and using a Nikon© AZ100 zoom microscope.

60The accelerated degradation experiment very clearly showed to the naked eye that elevated relative humidity (>50%) causes acceleration of degradation of both brass CZ108 and SRBP. While the brass sample dishes exposed to low and medium relative humidity barely showed a darkening of the brass colour, the SRBP sample dishes appeared to have lost surface gloss (see Fig. 9). The brass exposed to high relative humidity changed to a grey colour and appeared to have developed a friable crystalline crust.The SRBP sample dishes exposed to high RH turned into a dull grey brown colour. Additionally, the colour of the copper formate held in the recesses changed. The strong, light blue colour faded in samples exposed to low RH and darkened in samples exposed to medium RH. The high RH exposure showed a colour change towards light green (see Fig. 9).

61When observed under the microscope, the structure of brass seemed unchanged in brass sample dishes exposed to low and medium RH, while the samples exposed to high RH showed a hard crystalline crust, which could only be lifted with a scalpel blade. The SRBP sample dishes showed a gradual increase of surface scratching, with the SRBP sample dishes exposed to high RH having a dramatically changed surface from the original. The combination test sample dishes under the microscope showed similar changes.

Practical applications for conservation treatments

  • 5 Copper formate is a hydrophilic salt, which means it attracts and traps water. With a pKa value of (...)

62For the phenol formaldehyde electrical insulation this means that resistance of the material can decrease, due to the accumulation of salt, moisture, dirt, oil or corrosive vapours penetrating the insulator surface.5 As electrical insulators are statically charged, they tend to attract contaminants. Low resistance and leakage of current is the result, and the insulator turns into a partial conductor. Consulted literature does not state to what extend this influences its insulating properties.

63Once one contaminant accumulates, it becomes easier for other contaminants to become attached. This also has an effect on neighbouring metals, as the highly reactive copper alloys catalyse under moist conditions further and continue the reaction with formic acid (BILLINGHAM 2015).

64Hence, it is important to remove all contaminants regularly to avoid an accumulation of contaminants of any kind. As the removal techniques have not been investigated in this project, it would be advisable to test methods on small test pieces before treating the object (SHASHOUA 2008:209).

Conclusion

65Through practical research and literature review, the challenges of the conservation of this group of multi-material objects became apparent. Previously rarely addressed conservation issues can be demonstrated and investigated using the Type 36 as a suitable case-study. Correspondence with collectors and horologists, and additional review of horological conservation literature supported the need for inhibitive/preventive conservation methods, as literature, treatment and exhibition choices generally revealed an unawareness of harmful factors: Type 36 objects showing signs of degradation are operated and on display apparently without any well-defined preventive conservation measures to reduce micro-environmental or multi-material issues. If applied, these may avoid discarding of original materials such as the phenol formaldehyde based insulation materials used in the Type 36; or accelerated degradation through unsuitable storage conditions. Additionally, the identification of the possibility of asbestos as a filler in the early electrical insulation from the 1930s and 1940s helps to take necessary precautions during conservation processes. Phenol formaldehyde resin using wood flour as reinforcement is a more common composite and has been mentioned in plastic conservation literature, but is not regarded as a problematic material compared to other faster degrading polymers (BILLINGHAM 2015). Problematic micro-environments are a well-known problem in display cases and other disciplines of conservation but have not been commonly addressed by horologists. Horological conservation literature focusses strongly on mechanical issues, friction, movement and corresponding material loss of the movement and mostly ignores case and micro-environments.

66The comparison of degradation through artificially accelerating degradation of insulation materials originating from 12 different Type 36 clocks produced between 1930 and 1980 showed changing properties of the phenol formaldehyde materials over time. These differences are most likely due to different processing techniques and composition.

67The interpretation of chemical changes in tested materials proved to be challenging and not as detailed as a FTIR spectrum potentially allows. Relationships made between interpreted chemical bond changes and physical changes are often superficial. To go into depth of FTIR spectra interpretation, a more extensive chemistry knowledge and sufficient training by FTIR specialists would have been needed.

68The accelerated degradation test with copper formate applied to brass and Synthetic Resin Bonded Paper gave clear indications of the destructive effects of the corrosion product. The copper formate causes severe degradation of the resin, which spreads quickly over the whole material when exposed to high relative humidity.

69The visual examinations would have been complemented very much by gloss metry, colour measurement and SEM investigations, which were not possible in the given time frame. In the future, methods of copper formate removal could also be further investigated to widen the practical treatment options.

70A conclusion regarding practical conservation practices can be made:

71Developed corrosion products on Type 36 parts should be removed to avoid spreading and causing further damage of the surfaces of both phenol formaldehyde composites and brass. This allows to avoid or reduce the risk of degradation and especially delamination of phenol formaldehyde insulation in the Type 36. The relative humidity should be kept stable within recommended range of 55 ± 3 per cent which also decreases the risk of delamination (SHASHOUA 2008:195). Regular air exchange could reduce accumulation of gasses and pollutants in an enclosed clock case such as the Type 36.

Top of page

Bibliography

Books

Feller, R. (1994) Accelerated Aging. Marina del Rey, CA: Getty Conservation Institute

Haslam, J. and Willis, H. (1965) Identification And Analysis Of Plastics. London: Iliffe Books

Lavédrine, B., Fournier, A. and Martin, G. (2013) Preservation Of Plastic Artefacts In Museum Collections. Belgium: Bietlot Imprimerie CTHS

Pilato, L. (2010) Phenolic Resins: A Century Of Progress. Berlin: Springer

Selwyn, L. (2004) Metals And Corrosion. [Ottawa]: Canadian Conservation Institute

Shashoua, Y. (2008) Conservation Of Plastics. Amsterdam: Elsevier/Butterworth-Heinemann

Articles

Cowsill, J. (1997) ‘Post Office Clock Survey: Analysis of Results’. AHS Electrical publication No.62

Salthammer, T., Mentese, S. and Marutzky, R. (2010) 'Formaldehyde In The Indoor Environment'.Chem. Rev.110 (4), pp2536-2572

Websites

Amazon.co.uk. (2015) Kerr Wide Mouth Mason Jar-1/2 Pint: Amazon.Co.Uk: Kitchen & Home [online] available from <http://www.amazon.co.uk/gp/product/B0000BYD0F?psc=1&redirect=true&ref_=oh_aui_detailpage_o04_s00> [29 July 2015]

Asbestos.com, (2015) Plastics Containing Asbestos - Brands & Products [online] available from <http://www.asbestos.com/products/general/plastics.php> [27 August 2015]

Britishtelephones.com, (2015) CLOCK No. 36 [online] available from <http://www.britishtelephones.com/clocks/clock36.htm> [20 July 2015]

Copper Development Association. (2004) Copper And Copper Alloys- Compositions, Applications And Properties [online] available from <http://www.copperalliance.org.uk/docs/librariesprovider5/resources/pub-120-copper-and-copper-alloys.pdf> [17 July 2015]

Derrick, M., Stulik, D. and Landry, J. (1999) Infrared Spectroscopy In Conservation Science [online] available from <http://www.getty.edu/conservation/publications_resources/pdf_publications/pdf/infrared_spectroscopy.pdf> [3 August 2015]

Easyflip.co.uk, (2015) Digital Catalogue [online] available from <http://www.easyflip.co.uk/Scientific_labs_catalogue2014_prod/?page=108> [27 July 2015]

Guetta, D. (2006) Acidity, Basicity And Pka [online] available from <http://www.columbia.edu/~crg2133/Files/CambridgeIA/Chemistry/AcidityBasicitykPa.pdf> [4 September 2015]

Halfords.com, (2015) Halfords Battery Top-Up Water 5L [online] available from <http://www.halfords.com/motoring-travel/bulbs-wiper-blades-batteries/car-battery-chargers/halfords-battery-top-up-water-5l> [29 July 2015]

Photolux-luminance.com, (2015) PHOTOLUX , Is A Photometric Measurement System Or A Photoluminance Meter, A Compound Of A Processing Software And A Calibrated Digital Camera Provided With An Fish-Eye Lens, Create, From Photos Of A Camera, HDR Luminances Maps, Make Luminances Statistical Analyses, Calculate The Indices Of Comfort UGR, Calculate The Indication Of Comfort TI [online] available from <http://www.photolux-luminance.com/> [31 August 2015]

Plenco.com, (2015) Phenolic Novolac And Resol Resins - Phenolic Thermosetting Resin [online] available from <https://plenco.com/phenolic-novolac-resol-resins.htm> [17 July 2015]

Presspahn.com, (2015) Presspahn Ltd - Products - SRBP P1 (Paxoline/Paxolin) [online] available from <http://www.presspahn.com/Products/SRBP/P1.htm> [20 July 2015]

Shodhganga.inflibnet.ac.in, (2015) Indian ETD Repository @ INFLIBNET: Home [online] available from <http://shodhganga.inflibnet.ac.in/bitstream/10603/8942/6/06_chapter%201.pdf> [30 June 2015]

Sigmaaldrich.com, (2015) Copper Formate Aldrichcpr | Sigma-Aldrich [online] available from <http://www.sigmaaldrich.com/catalog/product/aldrich/cds013540?lang=en&region=GB> [21 July 2015]

Thickett, D. and Lee, L. (2004) Selection Of Materials For The Storage Or Display Of Museum Objects [online] available from <https://www.britishmuseum.org/pdf/OP_111%20selection_of_materials_for_the_storage_or_display_of_museum_objects.pdf> [14 July 2015]

Traditionaloven.com, (2015) Convert Drops Of Water Drop - Gtt SI Into Milliliters Ml Volume And Capacity For Culinary Practise [online] available from <http://www.traditionaloven.com/culinary-arts/volume/convert-drop-to-milliliter-ml.html> [29 July 2015]

Twi-global.com, (2015) FAQ: What Is The Difference Between Debonding And Delamination [In Adhesive Joints, Coatings And Composites, And Other Defects [online] available from <http://www.twi-global.com/technical-knowledge/faqs/process-faqs/faq-what-is-the-difference-between-debonding-and-delamination-in-adhesive-joints-coatings-and-composites-and-other-defects-found/> [10 August 2015]

UNITED STATES NAVAL ACADEMY ANNAPOLIS, MARYLAND, (1993) DELAMINATION IN COMPOSITE MATERIALS AS OBSERVED USING AN OPTICAL FIBER STRAIN GAGE [online] available from <http://www.dtic.mil/dtic/tr/fulltext/u2/a270803.pdf> [10 August 2015]

Personal correspondence

Billingham, N. (2015) FTIR Interpretation. Email to Tabea Rude, 22.August

Kobb, K. (2015) The Post Office And You. Conversation with Tabea Rude, 21.May

Other

Post Office Telecommunications Headquarters, (1974) Specification For Clock No. 36 (Mark 6) (Clock, Electric Master And Controlling). West Dean College. Instruction/Specification. Chichester

Shashoua, Y. (2015) PCIP Course: Conservation Of Plastics. West Dean College, 16-19.March

Top of page

Notes

2 Not all material was used in the final tests.

3 For the 1930s only one type of Fibre Reinforced Plastic (SRBP) could be sourced, three jars with six samples cover this decade (two samples in each jar). For the 1980s, two clocks’ components were available for testing, therefore this decade is given six jars (six samples in each jar). 1940s and 1970s are covered by each three jars (six samples in each jar)

4 Test materials were chosen due to their availability and validity, therefore some decades are represented by two clocks and other decades by a single clock or not at all.

5 Copper formate is a hydrophilic salt, which means it attracts and traps water. With a pKa value of 3.75, formic acid (HCOOH) is a weak acid. As a result it is ‘only partially dissociated and in aqueous solutions an equilibrium is established: HCOOH ⥃ HCOO-+ H+. When the salt copper formate is dissolved in water it will dissociate completely to Cu2+ and HCOO- ions. The same equilibrium must be established, the free acid is formed by the reaction with water: HCOO- + H2O ⥃ HCOOH + OH-.Thus the solution contains free formic acid. Copper formate is a hygroscopic solid and it is expected that similar equilibria will be set up’ (BILLINGHAM 2015)

Top of page

List of illustrations

Title Title: Fig. 1 The Post Office Type 36 case study object
Caption On the left hand side the illustration indicates the possible connection to numerous clock dials and telephone exchanges. On the right hand side the phenol formaldehyde based insulation materials SRBWF, SRBP and SRBF are shown in blown-up view.
Credits Credits: Tabea Rude (photo), Michael Goldrei (editing).
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-1.jpg
File image/jpeg, 380k
Title Fig. 2 The first stage of the poly-condensation reaction of phenol formaldehyde
Caption Reactions of formaldehyde at the ortho- and para positions of the phenol ring.
Credits Credits: Norman Billingham
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-2.jpg
File image/jpeg, 48k
Title Fig.3 The second stage of the poly-condensation reaction of phenol formaldehyde
Caption Under heat and pressure the phenol rings are joined by methylene and methylene ether bridges.
Credits Credits: Norman Billingham
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-3.jpg
File image/jpeg, 56k
Title Fig. 4 Detail of an electrical clock
Caption Phenol formaldehyde composite degradation and its effects on brass.
Credits Credits: Mostyn Gale
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-4.jpg
File image/jpeg, 148k
Title Fig. 5 The 1970s Type 36 and its three main testing materials
Caption The flake samples of the testing materials SRBP, SRBF, SRBWF are visible in the blown-up images, with the lines pointing at the points of where they were harvested from the case study object.
Credits Credits: Tabea Rude
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-5.jpg
File image/jpeg, 936k
Title Fig. 6 Tested phenol formaldehyde composites from the 1930s, 1940s, 1970s and 1980s.
Caption Clocks are named after their owner (Nye clock, Butt clock), their manufacturer (Magneta, Gents) and/or year of manufacture (1930s, 1970s).
Credits Credits: Tabea Rude
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-6.jpg
File image/jpeg, 756k
Title Fig. 7 Sampling example of the 1940s jar exposed to high RH
Caption SRBP, SRBF and SRBWF material flakes in three glass vials and three brass sample dishes from 1940 in high RH (achieved by glass vial filled with 10ml deionised water).
Credits Credits: Tabea Rude
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-7.jpg
File image/jpeg, 188k
Title Fig. 8 Sampling example for the copper formate accelerated degradation experiment
Caption Combination test sample dishes made by riveting SRBP and brass with steel rivet and fill recess with 0.1 gram of copper formate on the left hand side. Jar No. 6 illustrated on the right hand side showing combination test sample dish and SRBP dish each filled with 0.1 gram of copper formate in mason jar with glass vial holding 10ml deionised water to create high humidity.
Credits Credits: Tabea Rude
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-8.jpg
File image/jpeg, 352k
Title Fig. 9 Set-up of copper formate accelerated degradation experiment
Caption Photos of results without magnification and using a Nikon© AZ100 zoom microscope are shown below the corresponding jar. Microscopic views of riveted brass and SRBP sample dishes were extremely similar and therefore are not shown in this table.
Credits Credits: Tabea Rude (photos), Michael Goldrei (editing)
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-9.jpg
File image/jpeg, 2.2M
Title Fig. 10 Normalised spectra of SRBWF held in brass sample dishes from 1980s clock
Caption SRBWF exposed to low (red graph, S37b), medium (pink graph, S38b) and high (blue graph, S39b) humidity plotted against a control sample taken before artificially accelerated degradation experiment (black graph, SC 37, SC38, SC39). X-axis showing infra-red light transmittance in % versus frequency on the y-axis in reciprocal centimetres (cm-1).
Credits Credits: Tabea Rude
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-10.jpg
File image/jpeg, 100k
Title Fig. 11 Spectra of SRBP samples produced in the 1940s
Caption Loss of OH groups visible in SRBP samples held in brass dishes and exposed to high humidity (compare to ill.6). Control (red graph, SC4), material exposed to low humidity (black graph, S4b), medium humidity (pink graph, S5b) and high humidity (blue graph, S6b). X-axis showing infra-red light transmittance in % versus frequency on the y-axis in reciprocal centimetres (cm-1)
Credits Credits: Tabea Rude
URL http://journals.openedition.org/ceroart/docannexe/image/5263/img-11.jpg
File image/jpeg, 116k
Top of page

References

Electronic reference

Tabea Rude, « Testing phenol formaldehyde composite insulation material for conservation purposes », CeROArt [Online], EGG 6 | 2017, Online since 23 May 2018, connection on 21 August 2018. URL : http://journals.openedition.org/ceroart/5263

Top of page

About the author

Tabea Rude

Tabea Rude is an objects conservator at the Wien Museum in Vienna where she is responsible for their clock and automata collection. Previously she worked at The Clockworks Museum in London where she did conservation work and research on electromechanical clocks. She also worked for private clients on mechanical clocks and related dynamic objects. She graduated in July 2016 from the MA programme in Conservation Studies at West Dean College (University of Sussex). Previously, she completed a three year watchmaking apprenticeship in Pforzheim, Germany.

Top of page

Copyright

Licence Creative Commons
CeROArt – Conservation, exposition, restauration d'objets d'arts est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Top of page
  • Logo Fondation Périer-d’Ieteren
  • Logo DOAJ – Directory of Open Access Journals
  • OpenEdition Journals