Changes in morphology over time due to denaturing agents in an ethanol and camphor solution

1 Changes in morphology over time due to denaturing agents in an ethanol and camphor solution Hanne Jacobs1, Garin Cael2, Stefanie Bauvois1 1 University of Antwerp, Blindestraat 9, Antwerp, Belgium 2 Royal Museum for Central Africa, Leuvensesteenweg 13, Tervuren, Belgium Corresponding author: Hanne Jacobs ([email protected]) Copyright: © Hanne Jacobs et al. This is an open access article distributed under terms of the Creative Commons Attribution License (Attribution 4.0 International – CC BY 4.0). Research Article Abstract The ethanol collections of the Africamuseum require an approved denaturing agent, that can be used to top-up containers in which, historically, camphor-denatured ethanol has been used. Denatonium benzoate, diethyl ether and isopropyl alcohol were evaluated on their ability to preserve Lithobates catesbeianus (Shaw 1802) in the long term. There was no visual difference between the colors of the preservation liquids over time, but precipitates were formed in almost all of the test containers. The snout-urostyle length and the length of the thigh of the test specimens were assessed over time. After artificial aging, specimens in a mixture of ethanol denatured with isopropylalcohol shrunk less than those in other mixtures, even the control group, albeit by only a small fraction. Key words: Amphibian collections, Conservation studies, Denaturing ethanol, Fluid preservation, Morphology Introduction The Africamuseum in Tervuren, Belgium (RMCA) possesses a large collection of preserved animals (Africamuseum 2025). A substantial part of the vertebrate collections has been stored in containers containing 70% diluted ethanol, denatured with 0.35 g camphor per liter of the alcohol and water mixture, for decades. However, the European Union doesn’t list the use of camphor as a denaturing agent for ethanol (EUR-Lex 2018). Over time, the preservative fluid in the container will evaporate and need topping up, otherwise the specimens will dry out and deteriorate (Simmons 2014). Refilling the containers of existing specimens with ethanol that is denatured with camphor is no longer allowed. Ideally, the Africamuseum would use non-denatured ethanol, as this is recommended for collections, hence the search for another denaturing agent, that can be added to an ethanol and demineralized water mixture, to top up the containers without provoking a harmful chemical reaction with the existing preservation fluid. In this paper, alternative denaturing agents were evaluated that could be added to the specimens preserved in ethanol with camphor. Three denaturing agents were tested: denatonium benzoate, diethyl ether and isopropyl alcohol. The safety and health of the museum staff was taken into account. Each of these compounds are permitted as denaturing agents for ethanol by the Academic editor: Luis Ceríaco Received: 11 June 2025 Accepted: 29 September 2025 Published: 10 October 2025 Citation: Jacobs H, Cael G, Bauvois S (2025) Changes in morphology over time due to denaturing agents in an ethanol and camphor solution. Natural History Collections and Museomics 2: 1–12. https://doi.org/10.3897/ nhcm.2.161606 Natural History Collections and Museomics 2: 1–12 (2025) DOI: 10.3897/nhcm.2.161606 2 NHCM 2: 1–12 (2025), DOI: 10.3897/nhcm.2.161606 Hanne Jacobs et al.: Changes in morphology due to denaturing agents European Union (EUR-Lex 2018). Diethyl ether and isopropyl alcohol are often used on their own as preservation liquids, making them a suitable alternative (Simmons 2014). Denatonium benzoate has no history as a preservation liquid but is of interest. It is a bittering agent used in a wide variety of household products to prevent digestion or as a denaturant for alcohol without changing the contents (Berning et al. 1982). Twenty fresh specimens of Lithobates catesbeianus were preserved with 70% ethanol, camphor and one of the three denaturing agents mentioned above, to evaluate the preservation properties. The samples were assessed bimonthly to quantify the impact of the denaturing agent on the specimen. The snout-urostyle length (SUL) and the length of the thigh (THL) were measured (Fig. 1), as the mixture can shrink the specimen up to six percent (Deichmann et al. 2009). This trait is undesirable in a scientific collection that is regularly used for taxonomic research. The shrinkage is caused by dehydration because alcohol replaces water in the tissue of specimens. All alcohols will cause shrinkage but the amount of shrinkage can vary between types of alcohol (Ciferri 1971). Additionally, the color and transparency of the preservation liquid was visually monitored to ensure that the preservation mixture would not become opaque. Methods Theoretical compatibility and safety It was important to check the long-term compatibility of each component before making the mixtures. Camphor is soluble in ethanol (National Center for Biotechnology Information 2024) and will therefore also be dissolved by other alcohols like isopropanol. However, the functional carbonyl group of camphor will react with several aldehydes, ketones and strongly reducing or strongly oxidizing compounds (Mahdy et al. 2022). Denatonium benzoate is chemically inert (Berning et al. 1982) and should be stable in the presence of air or light (ThermoFisher Scientific 2024). Both isopropylalcohol and ethanol have low reactivity. Ether is a base and can react with several acids and can photo-oxidize to an explosive compound, diethyl ether peroxide (Arora and Gautam 2025). Working with fluid collections raises concerns on the health and safety of the staff dealing with these collections (Van Dam 2003; Cloth Workers Foundation 2012; Neumann et al. 2022). In this paper we have used the criteria defined by Van Dam to assess the flammability and toxicity of the mixture. However, there were almost no differences between the mixtures when it evaluated these parameters because of how similar the composition of the mixtures was. The base of each mixture consisted of 70% ethanol and 0.35 g camphor per Figure 1. Morphological measurements: the snout-urostyle length (SUL) and the thigh length (THL). The measurement of the thigh was taken by measuring the length between the cloacal gap and the external thigh joint. 3 NHCM 2: 1–12 (2025), DOI: 10.3897/nhcm.2.161606 Hanne Jacobs et al.: Changes in morphology due to denaturing agents liter water and ethanol; each mixture had a 1% difference in composition. The denaturing agents would most likely only either slow down or speed up any reactions caused by ethanol, camphor or the specimen. Therefore, we would theoretically expect all results to be similar. None of the tested mixtures were safe enough according to the categories in van Dam (2003). The large amount of ethanol in each mixture immediately fails each mixture for categories “Toxicity” and “Flammability”. In order to succeed in the category the compound had to have a score equal to 1 or lower on the Hazard Rating F (Flammability), as defined by the National Fire Protection Agency (NFPA). As the aim of this research was to evaluate which denaturants could be mixed with ethanol and/or each other, the presence of ethanol was indispensable, which means the mixture immediately fails the categories “Toxicity” and “Flammability”. The mixtures will be referred to as the following from this point on in the paper: Experimental setup and statistical methodology For this experiment the species Lithobates catesbeianus was selected as a suitable test subject. This frog species is invasive in Belgium and specimens are regularly euthanized. The frogs used in this study were part of the LIFE-3n Bullfrog project.1 Furthermore, we assumed that because frogs have a semi-permeable skin, they would absorb the preservation fluid easily and might be more prone to shrinkage. Twenty fresh specimens of Lithobates catesbeianus were fixed by injection of 4% formaldehyde buffered with 0.5–1.5% methanol and were wrapped in a cheesecloth permeated with a 4% formaldehyde solution for 24 hours. In order to simulate the current situation in the Africa museum twenty glass containers, each containing one specimen, were filled halfway with ethanol denatured with camphor to reflect samples in the current situation. They were divided into four sets of five specimens. Each set was topped up using ethanol denatured with one of three denaturing agents (isopropyl alcohol, denatonium benzoate, or diethyl ether). A fourth test set was topped up with the base solution of ethanol and camphor, and was used as a control group. Unfortunately, no pure ethanol was available for this experiment. The ethanol used was already denatured with methyl ethyl ketone and denatonium benzoate. The consequences of the use of this specific product were elaborated upon in chapter “Discussion”. 1 The project aims to sustainably reduce populations by releasing sterilized males. These frogs were fertile and no longer of use in the study. They were donated by Prof. Dr. Descamps (University of Hasselt) for this specific project. Table 1. Composition of each mixture. Name Components ECI Ethanol, camphor, isopropyl alcohol ECD Ethanol, camphor, denatonium benzoate ECE Ethanol, camphor, diethyl ether EC (Reference mixture) Ethanol, camphor 4 NHCM 2: 1–12 (2025), DOI: 10.3897/nhcm.2.161606 Hanne Jacobs et al.: Changes in morphology due to denaturing agents Every two weeks the snout-urostyle length (SUL) and the length of the thigh was measured, based on Watters et al. (2016), using a Mitutoyo analog caliper, with a precision of 0.02 mm, to quantify the shrinking of the specimen. These measurements were chosen because both the body and the thigh are generally measured in taxonomical research. A total of seven measurements were made over a period of three months, including a pre-preservation measurement as baseline. Additionally, pictures were taken bimonthly to visually check differences in color and opacity of the liquid throughout time. All pictures were taken under the same (artificial) light source using the same white background. A Calibrite Color Checker Passport Photo was used to ensure that any further changes in lighting could be fixed afterwards (Fig. 2). All measurements were taken by the first author, to ensure consistency, and to allow the measurements to be compared to each other. A nonlinear mixed effects model was proposed to quantify the change in both lengths throughout time: Length = Elow * (1 + Ehigh * exp(-t * log(2)/Thalf)) In this model Ehigh represents the maximum amount of loss of length as a percentage. Elow represents the size of a frog after it would have theoretically experienced this maximum amount of loss of length, t is time in days since the start of the experiment, and Thalf quantifies the time required to achieve half of the maximal shrinkage and can therefore be associated with the rate to achieve the maximal shrinkage. The model itself is a modified version of an exponential decay model as used in nuclear physics (Pommé et al. 2024) or pharmacokinetic modeling (Gabrielsson 2001). The modification here is that there’s an exponential decay from the initial level to the new, stable level. The model described the average shrinking effect over time but it also accounted for the individual deviation of each frog. Note that in typical statistical analyses and modeling, one assumes independence of the observations. In this setting, data of the same set of frogs was assessed multiple times over time, hence, no independent samples. Therefore, a mixed effects model (Pinheiro et al. 1995; Verbeke and Molenberghs 1997) was utilized. This accounts for the Figure 2. An example of a picture taken of ECD2. Please note the slightly yellowed hue in the liquid. 5 NHCM 2: 1–12 (2025), DOI: 10.3897/nhcm.2.161606 Hanne Jacobs et al.: Changes in morphology due to denaturing agents heterogeneity between the different longitudinal profiles by incorporation of an independent random effect at Elow and Ehigh (thigh measurement), and correlated random effects at Elow and Ehigh as well as an independent random effect at Thalf (SUL measurement) (Davidian and Giltinan 2000), i.e., quantifying the size of the frogs at population level, while accounting for animal specific deviations. Artificial aging At the end of the three-month follow-up, all twenty specimens were aged in their container based on the method described by Von Endt et al. (2000). In this paper, the long term stability of feathers and hair in 70% ethanol was investigated by heating to 180 °C for 1 and 2 days. This method was adapted for this research: the specimens in their respective containers were heated in a hot water bath to 40 °C for seven hours. The increased temperature will accelerate any chemical or biophysical reactions that normally would take place over a longer period of time. The main reason for lowering the temperature was the presence of diethyl ether which could form dangerous diethyl ether peroxide compounds when heated. Each specimen was measured right before and right after the aging process in order to assess the long-term effect of the denaturants. The purpose of the artificial aging was to demonstrate an absence of further deterioration of the samples. The difference between before (day 118) and after (day 126) artificial aging was analyzed using equivalence testing (Chow and Liu1999; Patterson and Jones 2017). The data was pairwise identified and logarithmically transformed with a linear mixed effect model where each frog served as its own reference and test observation. The model included an indicator for pre-, versus post-aging. The effect of aging was assessed by quantifying the difference before and after the artificial aging process on a logarithmic scale. Results Analysis of the data The size of SUL and thigh was measured bimonthly over a period of 3 months. Each specimen initially experienced a rapid exponential loss of length in the first few weeks, which slowed over time. Measurements were depicted in Figs 3, 4. The data was modeled while the difference between the different test compounds was assessed both at the parameters Ehigh and Thalf. Incorporation of the test compounds at the level of Thalf fit the data best. The model fit the raw data of the SUL and the size of the thigh of each specimen adequately. The SUL in the EC-reference group decreased 4.69% with a Thalf of 14 days. The length of the thigh in the EC-reference group decreased 4.46% with a Thalf of 20 days. The change in Thalf of the other compounds relative to the reference is depicted in Table 2. Note the independence of the maximal amount of shrinkage to the different compounds used. The maximum loss of length for the thigh consisted of 4.46 ± 1.26%. The loss for SUL after preservation in EC is 4.93 ± 1.26%. In this setting the reduction was 11.4, -14.0 and -6.66 for ECD, ECE and ECI, respectively relative to EC. 6 NHCM 2: 1–12 (2025), DOI: 10.3897/nhcm.2.161606 Hanne Jacobs et al.: Changes in morphology due to denaturing agents Figure 3. The SUL of each set was depicted at population level in a predictable distribution graph. Each point is an individual measurement of a frog at one specific moment. The population average is depicted by the black line. Colored intervals were used to display the prediction intervals. From the outer color to the inner color, each color represented 90%, 80%, 70%, 60% and 50%. Figure 4. The length of the thigh was represented in the same way as the SUL. For explanation of the graph, see Fig. 1. 7 NHCM 2: 1–12 (2025), DOI: 10.3897/nhcm.2.161606 Hanne Jacobs et al.: Changes in morphology due to denaturing agents Note that the estimated differences were not statistically significant, which was likely due to the small sample size of each of the compound groups. Interpretation of these results can be found in the chapter “Discussion”. Artificial aging The effect of aging on each set was estimated by quantifying the difference before and after the artificial aging process on a logarithmic scale. The exponents of this difference resulted in the average shrinkage in percentage that took place during the artificial aging process (Table 3). Visual differences During the first month a precipitation reaction was noticed in almost all the glass containers. The precipitate was not analyzed. Because of the unknown nature and unknown consequences of the precipitate, this reaction was undesirable. Therefore each tested set was judged on how much precipitation was formed during the entire experiment. Each test specimen was divided into four categories based on the amount of precipitation at the bottom of the container (Table 4). Discussion Two important sidenotes need to be made about the execution of the experiment. First, no pure ethanol was obtained prior to the research. Only ethanol that was already denatured with methyl ethyl ketone and denatonium benzoate was available for use. As the base mixture was the same for each category, we assumed that the differences between the groups would primarily be influenced by the difference in denaturants, and therefore the results would still reflect the suitability of each denaturant. However, any observed reactions might have also have been influenced by any reaction between the ketone and other Table 2. The relative change in Thalf compared to the reference compound for the SUL and thigh. SUL Thigh ECD 64% faster 340% slower ECI 67% faster 70% faster ECE 86% faster 30% faster Table 3. The average shrinkage of the snout-urostyle length and the size for the thigh of each mixture. Note that the 90% confidence intervals were omitted for readability reasons. SUL Size of the thigh EC 1.98% smaller 0.0302% smaller ECD 3.15% smaller 0.0366% smaller ECI 0.797% smaller 0.0464% smaller ECE 1.98% smaller 0.0221% smaller 8 NHCM 2: 1–12 (2025), DOI: 10.3897/nhcm.2.161606 Hanne Jacobs et al.: Changes in morphology due to denaturing agents components. Camphor has the tendency to react with different ketones, but no reaction between, specifically, methyl ethyl ketone and camphor has yet been noted in research (Mahdy et al. 2022). Secondly, it is also important to remember the small sample size used in this experiment (due to time constraints and limited lab space). The limited number of specimens used meant that the conclusions were statistically underpowered in contrast to studies with a larger sample size (Hayek and Heyer 2004). Concretely, an increased number of specimens would decrease the uncertainty surrounding results found in this research because any deviant data would become less significant. Nevertheless, the following conclusions could be made from the data. All specimens shrunk during the experiment, except one sample treated with ECI, whose thigh increased by 1.30 mm in the first two weeks of the experiment. There are several possible explanations for this. Although this could be due to a measuring error during the experiment, this phenomenon has been observed before in a study by Lee (1982).The SUL of six Rhinella marina (Linnaeus 1758) expanded by an average of +1.9% for male specimens and +1.2% for female specimens, while eight specimens shrunk. An increased sample size or a repeat of the experiment could either confirm or negate this phenomenon. During the experiment each specimen shrunk during the first few weeks. The shrinking caused by the ethanol, camphor and denaturing agent stabilized throughout time. Although the time to reach this maximal change differed between the compounds, it was considered to have limited practical impact: The preservation of the samples is intended for multiple decades, whereas the estimated half-lifes were in the magnitude of one month, as seen in Figs 1, 2. Attaining the maximal shrinkage after 5 half-lifes would therefore confirm the restricted relevance of the estimated differences and suggested only initial impact in the transition process from denaturants with camphor. Note that no statistically significant differences were found (based on likelihood ratio test: Table 4. The different categories of the amount of precipitation and which specimen belonged to which category. Score 0 1 2 3 Description There is no visible precipitate. There is little visible precipitate. There is enough precipitate to fill the entire lower ring in the container A large amount of precipitate obstructs the view of the bottom of the container. Example EC 3 ECE 1 ECE 5 ECD 2 Specimen belonging to the category ECI 4 ECI 2 ECI 5 ECI1 EC2 ECI 3 ECE 2 ECD 2 EC3 ECE 1 ECE 4 ECD 4 EC4 ECE 3 ECE 5 ECD 3 EC 1 ECD 5 EC 5 9 NHCM 2: 1–12 (2025), DOI: 10.3897/nhcm.2.161606 Hanne Jacobs et al.: Changes in morphology due to denaturing agents p = 0.06064359 and p = 2.501712e-07 for SUL and thigh, respectively) in the shrinkage between the sets before the artificial aging process, however, this might be related to the limited sample size in the experiment. Note that this study was not a priori powered to detect a prespecified biologically relevant effect, hence all p-values should therefore be considered as hypothesis generating rather than hypothesis confirming. When considering the long-term changes, a minor difference in average shrinkage was noted after the artificial aging process. The SUL of set ECI shrunk the least of all sets (including control set EC) with a shrinkage of only 0.797%. Specimens of set EC shrunk by 1.98%.Set ECE performed similar to set EC in this category. Set ECD performed the worst with a shrinkage of 3.15%.The length of the thighs of set ECI shrunk on average by 0.0464%. Despite it being the largest shrinkage of the length of the thighs, this shrinkage was small enough to be considered of limited biological relevance. The thighs of the specimens most likely shrunk less than the SUL because it contains less soft tissue. No difference between color of the set was found. The liquid in each container turned to a slight yellowish hue, but each specimen was still clearly visible. Any variation in color could have been caused by a difference in fat and blood present in each specimen, which could have been transmitted to the liquid. Another issue in the containers was the formation of precipitates because of its unknown composition and ability to cloud the specimen. This influenced the opacity of the liquid. After visual inspection, testset EC appeared to have the least amount of precipitation (Table 3), respectively followed by testset ECI, testset ECE and, finally, testset ECD. The exact composition of the precipitation was unknown, but possibly caused by the previously mentioned addition of methyl ethyl ketone and denatonium benzoate. Considering that set ECD had the most precipitation and that denatonium benzoate was present in all sets due to the impurity of the ethanol, the precipitation was most likely to blame on interactions between camphor and denatonium benzoate despite the latter being considered chemically inert. Conclusions It should be reiterated that the experimental set-up was restricted to a single species, Lithobates catesbeianus. Different species might react differently to the same experiment. Regardless, repetition and/or amplification of the experiment is required to lessen statistical error. Pure ethanol would be needed as a base of the solutions. Furthermore, it might have been interesting to omit some specimens from the rapid aging procedure, to check how relevant the rapid aging was in the overall picture. There were minimal differences in shrinkage and color change between the tested sets. Each specimen shrunk within the first few weeks of testing, except for the thigh of ECI1. This shrinkage stabilized throughout time. The three tested denaturing agents―denatonium benzoate, isopropyl alcohol and diethyl ether―did not cause a significant difference in shrinkage during the first few months of testing. However, after artificial aging, specimens in a mixture with isopropyl alcohol shrunk the least on average even while excluding the