搜索
当前位置:首页 >> 研究方向 >> 新闻
论文:Fast escape of hydrogen from gas cavities around corroding magnesium
发表时间:2013-11-21 阅读次数:5188次

Fast escape of hydrogen from gas cavities around corroding magnesium implants 

  • , 
  • , , 
  • , , 
  • , 
  • , 
  • ,
  • , 
  • , , , 
  • , , , 
  • a Department of Chemistry, University of Cincinnati, P.O. Box 210172, Cincinnati, OH 45221-0172, USA
  • b Laboratory for Biomechanics and Biomaterials, Orthopedic Clinic, Hannover Medical School, Anna-von-Borries Strasse 1-7, 30625 Hannover, Germany
  • c CrossBIT, Center for Biocompatibility and Implant Immunology, Hannover Medical School, Feodor-Lynen Strasse 31, 30625 Hannover, Germany
  • d Fraunhofer Institute for Toxicology and Experimental Medicine, Department for Bio- and Environmental Analysis, Nikolai-Fuchs Strasse 1, 30625 Hannover, Germany
  • e Helmholtz Zentrum Geesthacht, Max-Planck Strasse 1, 21502 Geesthacht, Germany
 

Abstract

Magnesium materials are of increasing interest in the development of biodegradable implants as they exhibit properties that make them promising candidates. However, the formation of gas cavities after implantation of magnesium alloys has been widely reported in the literature. The composition of the gas and the concentration of its components in these cavities are not known as only a few studies using non-specific techniques were done about 60 years ago. Currently many researchers assume that these cavities contain primarily hydrogen because it is a product of magnesium corrosion in aqueous media. In order to clearly answer this question we implanted rare earth-containing magnesium alloy disks in mice and determined the concentration of hydrogen gas for up to 10 days using an amperometric hydrogen sensor and mass spectrometric measurements. We were able to directly monitor the hydrogen concentration over a period of 10 days and show that the gas cavities contained only a low concentration of hydrogen gas, even shortly after formation of the cavities. This means that hydrogen must be exchanged very quickly after implantation. To confirm these results hydrogen gas was directly injected subcutaneously. Most of the hydrogen gas was found to exchange within 1 h after injection. Overall, our results disprove the common misbelief that these cavities mainly contain hydrogen and show how quickly this gas is exchanged with the surrounding tissue.

Keywords

  • Magnesium alloys; 
  • Hydrogen; 
  • Biodegradable implants; 
  • Gas cavities; 
  • Amperometric hydrogen sensor

1. Introduction

Metals have been used as internal fixtures to aid the healing of fractured bones and tissue for more than 100 years . Today commonly used metals for these types of implants are stainless steel, Ti and Co–Cr alloys . While these permanent implants are invaluable and generally biocompatible, they can cause problems such as stress shielding and the release of toxic metal ions through corrosion over time  and . Therefore, research groups are developing biodegradable (temporary) metallic implants, many of them focusing on Mg-based materials. Although Mg materials have traditionally been used for structural applications in the automotive and aerospace industry, they have gained attention in the orthopedic and biomedical engineering fields  and . Their unique properties, which include physical and mechanical properties close to those of bone, make them promising candidates for biodegradable implants. Furthermore, these materials are generally non-toxic, light in weight and corrode rapidly in aqueous environments and . During corrosion Mg is oxidized to Mg2+ as water is reduced to H2 and OH. While the human body buffering system can compensate for the release of OH and some increase in Mg2+ is non-toxic , little is known about the fate of H2 in vivo. The evolution of H2 gas after adding Mg and its alloys to aqueous solutions has been extensively observed  and , as has the formation of gas cavities in vivo , and . Two studies conducted over 60 years ago used techniques available at that time to analyze the gas composition of these cavities. McBride  reported in 1938 that gas samples aspirated from a cavity 40 days after implanting a band of Mg alloy showed a gas composition of 5.6% CO2, 6.5% O2, 7.3% H2 and 80.6% N2. However, he did not state how the gas composition was determined. In 1942 McCord et al. used an interferometer to analyze the composition of gas samples drawn from sites of gas gangrene formed in rats 5 days after Mg powder implantation. Their results showed a gas composition of 1.3% CO2 and 15.2% O2, and they calculated that the H2 concentration must have been 2.2% using a method described by Edwards . Additionally, their efforts to ignite the gas sample failed, which led to the conclusion that the sample must have contained less than the 4.1% H2 required for ignition in air. More recently Witte et al. also tried to ignite samples withdrawn from gas cavities, but no combustion was observed. While these studies suggested that H2 might not be the major component in these gas cavities, they did not directly measure H2 nor did they use current analytical methods to determine the concentration of H2. More importantly, while these studies remain the only ones that have attempted to analyze the H2 concentration, it is often still assumed that gas cavities formed during Mg material corrosion in vivo contain primarily H2,,  and . Nevertheless, highly selective analytical techniques to measure H2 are available. An electrochemical sensor for H2, analogous to the commonly used Clark O2 sensor, has previously been used in vivo to determine local blood flow using the H2 washout technique . This amperometric sensor detects H2 by selectively oxidizing H2 to H+. Although this sensor is not implantable, it would enable measurements on the surface of the gas cavities. Mass spectrometry is another commonly used analytical technique that would allow direct analysis of H2 and other gases in the cavity.

Here we report a method to directly analyze the H2 concentration and the gas composition of cavities formed during in vivo degradation of subcutaneously implanted rare earth-containing Mg alloy disks. We used an amperometric H2 sensor and mass spectrometry to analyze the gas in the cavities over the course of 10 days, thereby providing a time course for H2 behavior in vivo. After the experiment we analyzed the response of the skin covering the alloy disks and the corrosion layer on the alloys. Our results are especially significant in that they alleviate concerns about H2 gas accumulating in the bodies of implant patients.

2. Materials and methods

2.1. Mg alloy preparation

The Mg alloy Mg–4 wt.%Y–0.5 wt.%Gd–2 wt.%Nd–0.5 wt.%Dy was prepared using pure elements by direct chill permanent mold casting according to the standard procedure as describe by Elsayed et al. . All casting operations were performed under a protective gas (Ar with 2% SF6). Pure Mg was melted in a mild steel crucible. At a melting temperature of 730 °C the alloying elements Y, Nd, Gd and Dy were added. The alloy was stirred for 20 min at 150 rpm. Afterwards the melt was poured into a thin walled (3 mm thickness) mold made of mild steel. The mold was then kept in a holding furnace at 680 °C for 1 h to homogenize the melt. After holding, the mold was dipped into flowing water (15 °C) to solidify the material and to produce a casting. Mg alloy disks (8.0 mm diameter, 1.5 mm thick) were machined from the cast material, polished with SiC emery paper (up to 4000 grit), briefly etched and cleaned in 100% ethanol in an ultrasonic bath. All disks were γ-ray sterilized with 27 kGy of 60Co radiation and had an average weight of 141.2 ± 1.3 mg before implantation.

 

2.2. Animal model

The animal experiment was conducted under an Ethics Committee approved protocol in accordance with German federal animal welfare legislation (Ref. No. 33.9-42502-04-08/1499) and in accordance with theNational Institutes of Health Guidelines for the Use of Laboratory Animals. 10 female hairless mice from the Charles River Laboratories (Crl: SKH1-h) aged 12–24 weeks were used in this study. These mice are hairless but immunocompetent. Their fur is normal for up to 10 days, and then the hair is gradually lost, starting around the nose. Around day 20 the fur is lost completely. The average weight of the mice used was about 26 g. After implantation each mouse was housed individually and was fed a standard diet (Altromin1324) and water ad libitum. The animal husbandry rooms were illuminated by artificial light 14 h a day starting at 7 a.m. The mice were anaesthetized by intraperitoneal injection of 2% xylazine (10 mg kg body weight−1, Rompun®, Bayer Health Care, Leverkusen, Germany) and 10% ketamine (100 mg kg body weight−1, KetaminGräub®, Albrecht GmbH, Aulendorf, Germany). In order to avoid cooling of the body mice were placed on a custom made heating plate during surgery and measurements. The dorsal skin was cleaned according to surgical guidelines. Two longitudinal incisions (one in the shoulder region, one in the lumbar region) of 1 cm each were made in the median line through the full thickness of the skin. Subcutaneous pockets between the fascia of the dorsal muscles and the subcutaneous tissue were created by blunt dissection with scissors. The implants were placed in these pockets. The skin was closed with resorbable surgical suture material (Vicryl, Ethicon, Johnson & Johnson GmbH, Germany).

For the H2 injection experiments (99.999%) was injected subcutaneously into hairless mice (three mice per time point). The gas was withdrawn with gastight syringes 1, 2, 4 and 12 h after injection and analyzed by mass spectrometry as described in Section .

2.3. H2 measurements

Amperometric H2 measurements were performed using a H2 microsensor (50 μm tip diameter) connected to a multimeter (both from Unisense, Aarhus, Denmark) polarized at +800 mV for at least 1 h . After a stable current in the low picoampere range was established, the amperometric sensor was ready to be used. The sensor was calibrated by adding known amounts of H2-saturateddeionized H2O to a known volume of deionized H2O (according to the manufacturer’s recommendations  and ). Measurements were taken for 3 min on the skin on top of the gas cavities and subcutaneously in an incision on top of the gas cavities. As a control, measurements were taken on top of the skin in an area without any gas cavities. Standard deviations (error bars in the graphs) were calculated from the averaged first 100 data points taken during each of the 3 min measurements. Microsensor readings were converted into vol.% for comparison with the mass spectrometry data. The manufacturer reported the limit of detection to be 0.02% (0.1 μM) in H2O.

For mass spectrometry, the gas samples were withdrawn from the cavities using 2.5 ml gas tight syringes (Hamilton Messtechnik GmbH, Höchst, Germany). The needles were covered with a septum until the samples were analyzed using a SMart Nose® volatile organic compound analyzer system (former Smartnose SA, Switzerland), which was fitted with a special injection device and a capillary for gas phase transfer purposes. The analysis was performed with as-received gas samples, without any further preparation. About 1–2 ml of each sample were introduced into the injection port heated to 22 °C. Before injection, the N2 flow was stopped for 20 s and acquisition was started simultaneously with opening of the gas outlet. Data were acquired at 70 eV, from 1 to 60 m/z, conducting four cycles within 150 s. Averaged cycles 2–4 were used for data analysis. After each injection the entire system was purged with N2 at a flow rate of 80 ml min−1 for 2 min. The system was calibrated with pure gas samples of H2 (99.999%), O2(99.5%), and CO2 (99.995%) as well as wish by gas mixtures of H2 and CO2 injected in amounts from 0.1 to 1.0 ml under similar conditions.

2.4. Histology

The excised tissue samples were fixed in 3.7% commercial formalin (Otto Fischer, Saarbrücken, Germany), then embedded and sectioned in paraffin. Then they were stained with Mayer’s hematoxylin (Merck KGaA, Darmstadt, Germany) and 1% eosin (Merck KGaA, Darmstadt, Germany) and mounted in Eukitt (Labonord, Mönchengladbach, Germany) according to standard procedures. We performed at least five histomorphometric measurements per hematoxylin and eosin (H&E) cross-section of each mouse biopsy and analyzed 10 skin biopsies from 2 and 10 day mice as well as four full skin biopsies from the controls.

2.5. XPS Material analysis and corrosion rate determination

X-ray induced photoelectron spectroscopy (XPS) experiments were carried out in a Kratos AXIS Ultra DLD system (Manchester, UK) equipped with a 15 kV X-ray gun using monochromatic Al Kα radiation. The analyzed area size was 700 × 300 μm positioned in the center of the sputter area (2 × 2 mm). Energy resolution was set to a pass energy of 20 eV for high resolution region measurements. A charge neutralizer was applied to correct for the chemical shifts caused by the non-conductive corrosion products. For depth profiling Ar ions (Ar+ at 4 keV) were used to etch the samples. The applied sputter rate was approximately 25 nm min−1. The core levels Mg 2p, O 1s, C 1s, N 1s, Na 1s, Ca 2p, P 2p, Gd 3d, Y 3d, Nd 3d and Dy 4pwere analyzed to determine the depth distribution. Curve fitting and deconvolution of the O 1s spectra were performed with CasaXPS 2.315 software (Casa Software Ltd, UK). The binding energy was estimated to an accuracy of ±0.3 eV. The corrosion layer thickness was measured on micrographs of metallographic specimens of the corroded Mg alloy samples after 2 and 10 days subcutaneous implantation using ImageJ software.

The corrosion rate was determined after the explanted Mg alloy disks were treated with 180 g l−1 CrO3 at 22 °C for 20 min to remove the corrosion products. Differences in weight before and after implantation were used to calculate the corrosion rate in mm year−1 as described elsewhere .

3. Results and discussion

We implanted fast corroding rare earth element-containing Mg alloy disks (8.0 × 1.5 mm, permanent mold cast, machined and polished Mg–4Y–2Nd–0.5Gd–0.5Dy) with an average weight of 141.2 ± 1.3 mg subcutaneously in 10 hairless mice (Crl: SKH1-h) for up to 10 days. We chose this alloy since it corrodes rapidly and has previously been shown to produce large gas cavities in vivo. Two Mg alloy disks (one for each type of measurement) per mouse were placed subcutaneously under sterile conditions. H2 evolution was monitored over 2 days (2 day group) and 10 days (10 day group), with five mice per group. The mice started showing clinically visible subcutaneous gas cavities about 24 h after implantation () and the visible size of these cavities increased during the course of the experiment.

Fig. 1. 

A hairless but immunocompetent mouse is shown 1 day after the implantation of rare earth element-containing Mg alloy disks. Arrows indicate the development of clinically visible subcutaneous gas cavities around the implants.

A H2 microsensor from Unisense with a 50 μm tip diameter, calibrated following the manufacturer’s recommendations, was used for the amperometric measurements. The mice were anesthetized and 1 cm long incisions close to the gas cavities around the disk implanted in the left shoulder region were made using a sterile surgical scalpel. The anesthetized mice were immobilized on a heated swiveling table and the microsensor was positioned with a micromanipulator (). Measurements were taken on the skin on top of the gas cavities, in the incision on top of the fibrous tissue layer engulfing the gas cavities and on the skin in a control area. We were able to measure H2 on the skin on top of the gas cavities. This indicated that the H2at least partially exchanges through the skin, which could be facilitated by enrichment of H2 in the fatty tissue or oily layer of the skin, as H2 is more soluble in non-polar than in polar solutions. However, the results from these measurements were generally slightly lower compared with measurements in the incision. Measurements inside the gas cavities were done in preliminary experiments by piercing the tissue engulfing the gas cavities with the microsensor (data not shown). These experiments showed that there was no significant difference between measuring inside the cavity compared with measuring on the top of the tissue engulfing the cavity. Since it was not always possible to pierce the tissue with the microsensor and there was no significant difference between the measurements, we decided to only measure in the incisions on top of the tissue. The H2 concentration in the incisions ranged from 97 ± 46 μM (0.2 ± 0.1 vol.%) to 305 ± 32 μM (0.69 ± 0.07 vol.%) 1 day after surgery and from 95 ± 34 μM (0.22 ± 0.08 vol.%) to 428 ± 35 μM (0.97 ± 0.08 vol.%) 2 days after surgery for the 2 day group (a). On average, a minimal increase in H2over the course of 2 days was observed. For the 10 day group (b) the results varied considerably among mice. After 2 days the H2 concentration in the incisions ranged from 66 ± 32 μM (0.15 ± 0.07 vol.%) to 611 ± 44 μM (1.4 ± 0.1 vol.%) and from 258 ± 36 μM (0.59 ± 0.08 vol.%) to 391 ± 31 μM (0.89 ± 0.07 vol.%) after 5 days. The experiment was completed 10 days post-surgery due to animal welfare concerns associated with the very large size of the gas cavities. At that point (b) the H2 concentration in the incision ranged from 361 ± 91 μM (0.8 ± 0.2 vol.%) to 507 ± 38 μM (1.15 ± 0.09 vol.%), with two data points above the upper limits of the calibration curve (>805 μM (>1.8 vol.%), data points not included on graph). The variation in the 10 day group could be due to free movement of the Mg alloy disk inside the gas cavity. During free disk movement the protective corrosion layer is destroyed, thus resulting in Mg corrosion depending on the mouse activity level.

 
Fig. 2. 

Experimental apparatus for amperometric measurements. An anesthetized mouse is positioned on a swiveling table, which is connected to a peristaltic pump and heated to 37 °C by a water bath. A H2microsensor is mounted in a micromanipulator and positioned in an incision in the skin of the hairless mouse to record a blank reading before implantation.

Fig. 3. 

Results of the amperometric H2 sensor measurements for the (a) 2 day and (b) 10 day groups. Measurements were taken on the skin on top of the gas cavities (skin), in the incision on top of the gas cavities (incision) and on a control area (control). The results shown are not blank corrected.

We used mass spectrometry (MS) to confirm the results from the amperometric microsensor and to analyze the composition of the gas. Gas samples were withdrawn from the cavities in the right lumbar region with a gastight syringe and injected into a volatile organic compound analyzer system (Smart Nose®). The H2concentration was between 0.05 ± 0.04 and 0.12 ± 0.04 vol.% 2 days after surgery (a) and the CO2concentration was between 0.15 ± 0.04 and 0.18 ± 0.04 vol.%. 10 days after surgery (b) the H2concentration ranged from 0.06 ± 0.04 to 0.30 ± 0.04 vol.% and the CO2 concentration was between 0.08 ± 0.04 and 0.12 ± 0.04 vol.%. These results confirmed the amperometric measurements. Additionally, there were no significant differences on O2 concentration and no other gases besides N2 were detected. One should note that N2 could not be quantified since it was also used as a carrier gas.

Fig. 4. 

Mass spectrometry. Results of the mass spectrometric gas analysis of samples withdrawn from cavities (a) 2 and (b) 10 days after implantation.

To analyze how quickly H2 exchanged, we injected H2 subcutaneously into hairless mice (three mice per time point). After 1, 2, 4 and 12 h the gas was withdrawn and analyzed with a Smart Nose® system (). The H2 concentration dramatically decreased to as low as 0.22 ± 0.04 vol.% after only 1 h and was almost completely exchanged (0.02 ± 0.04 vol.%) after 12 h. This indicates that the exchange of H2 occurred very rapidly, while the gas cavity did not visually change in size.

Fig. 5. 

H2 injection. The results of the H2 injection experiments with three mice per time point showed that the H2concentration 1 h after injection was below 0.6 ± 0.04 vol.%. At the 12 h time point it had further decreased below 0.2 ± 0.04 vol.%. The CO2 concentration remained constant throughout the experiment.

Tissue samples were taken from the skin covering the implants for histological analysis to gain insight into the gas exchange process. The H&E stained histology showed a slightly thicker fibrous capsule (204 ± 18 μm) around the gas cavity after 2 days (a), which decreased in thickness to 120 ± 17 μm after 10 days (b). This is in the normal range of physiological mouse fascia (157 ± 23 μm), as shown in control samples (c) taken from the same mouse. One could argue that the thicker fibrous capsule after 48 h (a) could have kept more H2 inside the cavity, since the diffusion distance through the skin is greater than after 10 days (b). However, the H2 concentration was lower after 2 days than after 10 days. On the other hand, the thicker tissue layer would allow more H2 to saturate the surrounding tissue, thus leading to a lower concentration of H2 in the cavity. This could mean that gas exchange through the skin is not the major path of exchange. Additionally, the difference in size of the vacuoles seen in the skin sections of the test and control samples are a normal observation in this type of mouse model, as described elsewhere .

 
Fig. 6. 

H&E stained histology of skin samples. Full cross-sections of mouse skin that covered the gas cavities (a) 2 and (b) 10 days after surgery and (c) a control sample are shown. , fibrous capsule; *, physiological dermal cysts.

To better understand this result, we analyzed the corroded Mg disks by XPS and determined the corrosion rate from the weight loss of the explanted Mg alloy disks after removing the corrosion products with chromic acid, as described elsewhere . The XPS analysis revealed that the bulk matrix primarily determined the corrosion properties of the alloy, while intermetallics like Mg41Nd5 or Mg24Y5 () revealed slightly modified degradation properties, as described by Liu et al. . The results of the XPS measurements () were for two disks implanted in the lumbar region for 2 and 10 days, respectively. To analyze the corrosion products the composition of body fluids that were in contact with the material and the most common inorganic compounds (,  and Ca2+ compounds) other than water  needed to be considered. The determined elemental distribution represented only the outer 5 μm of the corrosion layer, not the elemental distribution through the whole thickness of the corrosion layer. The sample from the 2 day group (a) had a 1 μm thick corrosion active dissolution interaction zone consisting of Mg(OH)2. The high resolution scan of the O 1s state shows the different excitation states. Below the dissolution region different amounts of , O2−,  and other high order compositions (, marked by X) were found. These results are comparable with Abidin’s work . The shift in electrochemical potential and corrosion currents depends on the amounts of intermetallic phases, while the quantity of reacting rare earth elements depends on their reactivity . For example, Y2O3 and Y(OH)3 formed are comparable with the Mg(OH)2 zone in Mg–O–X compounds. Other probable compounds are hydroxyapatites based on Ca or P with aqueous bonding as well as rare earth compounds like NdPO4. The substantial increase in thickness of the corroded layer after 10 days is displayed in the depth profile (b). Organic compounds (e.g., from proteins) interacting at the disks could account for the high concentration of carbon atoms. The O 1s level showed a OH zone down to approximately 5 μm. For longer periods in vivo, the influence of rare earth elements slows down, acting as a degradation brake, as explained by Yao in the case of Y . Shortly after implantation strong H2 evolution occurred, after which the degradation rate stabilized. Due to interaction with the environment and passivation, stable compounds were formed, which enhanced growth of the corrosion layer. Finally, a very inhomogeneous surface structure arose and this mixture of organics,  and OHdetermined the disk degradation properties. It seems that the elemental composition of the outer corrosion layer changed during the time course of the 10 day implantation. Analysis of the corrosion layer thickness performed after the XPS analysis revealed that the thickness increased from 27.5 ± 10.5 μm after 2 days to 201 ± 107 μm after 10 days. Overall, this supports the findings of the corrosion rate determination from weight loss, which revealed that the 10 day implants corroded faster (20.7 ± 13.1 mm year−1) than the 2 day implants (8.6 ± 9.9 mm year−1).

Fig. 7. 

The microstructure of a Mg alloy implant. Large grains and their growth directions are shown. The Mg–rare earth intermetallics Mg41Nd5 or Mg24Y5 are visible in interdendritic regions. The Mg–rare earth phases and the bulk matrix are in competition with regard to corrosion surface reactions.

 
Fig. 8. 

XPS analysis of the corrosion layer on the alloy disks. Elemental depth profiles of disks implanted in the lumbar region for (left) 2 and (right) 10 days are shown. Elemental depth profiles (top) and high resolution scans of the O 1s state (bottom) give information about the distribution of different compounds (12,000 s sputtering is ∼5 μm).

Based on the results of our study, the known physical properties of H2 and the low concentration of H2 in the atmosphere and the body,  we propose the following potential exchange mechanism, in which the concentration gradient between the gases in the initially formed gas cavity, the adjacent tissue and the atmosphere seems to be the main driving force behind the exchange. Corrosion of the alloy disks starts immediately after implantation and the initial evolution of H2 drives the formation of the cavity, pushing the surrounding tissue layers apart. Simultaneously, H2 is expected to saturate the adjacent tissues, although the solubility of H2 in cells and aqueous media such as biological fluids is low  and . While the alloy disks continue to corrode and the size of the cavities increases, the evolving H2 in the growing cavities continuously exchanges with dissolved gases like N2, O2 and CO2 from the neighboring tissues and blood vessels . Diffusion of H2 through the skin also occurs, since we were able to measure about the same levels of H2 on top of the skin as directly on the bubble. However, this type of exchange may be too slow to be the main path of exchange. As the corrosion behavior of the alloy changes over the course of 10 days, the size of the gas cavities as well as the H2 concentration in the cavities increased. This can be explained by an increased corrosion rate that must be slightly higher than the exchange rate of H2. Previous studies also reported that the cavities disappeared several months after implantation , which means that the corrosion rate of the implant material and thus the evolution rate of H2 falls below the exchange rate of H2.

4. Conclusion

In this study we were able to directly measure the concentration of H2 in subcutaneous gas cavities with an amperometric H2 microsensor over the course of 10 days. The results were confirmed by mass spectrometry, which was also used to determine the gas composition. The results show that the observed subcutaneous gas cavities contained only a low concentration of H2 even shortly after implantation, which led to the conclusion that H2 is readily exchanged in vivo. Furthermore, we subcutaneously injected H2 to determine how quickly it is exchanged. With these experiments we were able to show that H2 is exchanged very quickly after implantation, as there was only a very low concentration detected 1 day after surgery and within 1 h after subcutaneous injection of H2. We also detected H2 on top of the skin, which suggests that H2readily exchanges through the skin and/or accumulates in fatty tissue, as it is more soluble in non-polar solvents compared with body fluids. Over 10 days the size of the cavities as well as the H2 concentration increased. The increase in corrosion rate and the corrosion layer after 10 days could explain the increase in size of the gas cavities, as well as the slight increase in H2 concentration, given that H2 exchanges at a constant rate. Although the exact nature of this exchange and the formation of the cavities is not clear, diffusion through the skin as well as diffusion into capillaries and transport by the vascular system might be essential, as outlined in the proposed exchange mechanism. These findings apply to subcutaneous gas cavities, but further research is needed to evaluate whether this exchange behavior still applies when more tissue layers cover the cavities (as found after implantation in bone). Our conclusions support the hypothesis drawn by McBride and McCord over 60 years ago that H2 is absorbed as rapidly as it is formed when oxidation of the metal is slow  and , even though in their experiments H2 was not directly measured and the measurements were done 40 days after implantation.

These results suggest that toxicity resulting from high concentrations of H2 accumulating in tissue adjacent to an implant consisting of Mg or an Mg alloy may not be an insurmountable problem in the practical application of these materials.

Acknowledgements

The authors thank Maria Brauneis and Maike Haupt of CrossBIT, Sophie Müller, Heike Achilles and Mattias Reebmann of the Laboratory for Biomechanics and Biomaterials at the Hannover Medical School as well as Bianca Lavae-Mokhtari of the Fraunhofer Institute for Toxicology and Experimental Medicine in Hannover for their technical support. Additionally, the authors thank Gerald B. Kasting from the Department of Pharmaceutics and Cosmetic Science at the University of Cincinnati for helpful discussion about the gas exchange processes. The study was supported by the NSF Engineering Research Center for Revolutionizing Metallic Biomaterials ().

 

Appendix A. Figures with essential color discrimination

Certain figures in this article, particularly Figs. 1, 2, 4, 6–8, are difficult to interpret in black and white. The full color images can be found in the on-line version, at .

 

 

References

    • H.K. Uhthoff, P. Poitras, D.S. Backman
    • Internal plate fixation of fractures: short history and recent developments

    • J Orthop Sci, 11 (2006), pp. 118–126

    •  | 
    • F. Witte, N. Hort, C. Vogt, S. Cohen, K.U. Kainer, R. Willumeit et al.
    • Degradable biomaterials based on magnesium corrosion

    • Curr Opin Solid State Mater Sci, 12 (2008), pp. 63–72

    •  |   | 
    • M.P. Staiger, A.M. Pietak, J. Huadmai, G. Dias
    • Magnesium and its alloys as orthopedic biomaterials: a review

    • Biomaterials, 27 (2006), pp. 1728–1734

    •  |   | 
    • C. Di Mario, H. Griffiths, O. Goktekin, N. Peeters, J. Verbist, M. Bosiers et al.
    • Drug-eluting bioabsorbable magnesium stent

    • J Interv Cardiol, 17 (2004), pp. 391–395

    •  | 
    • G.L. Song, A. Atrens
    • Understanding magnesium corrosion: a framework for improved alloy performance

    • Adv Eng Mater, 5 (2003), pp. 837–858

    •  | 
    • G.L. Song, A. Atrens
    • Corrosion mechanisms of magnesium alloys

    • Adv Eng Mater, 1 (1999), pp. 11–33

    •  | 
    • R.A. Kaya
    • The effects of magnesium particles in posterolateral spinal fusion an experimental in vivo study in a sheep model

    • J Neurosurg Spine, 6 (2007), pp. 141–149

    •  | 
    • C.P. McCord, J.J. Prendergast, S.F. Meek, G.C. Harrold
    • Chemical gas gangrene from metallic magnesium

    • Ind Med, 11 (1942), pp. 71–75

    •  
    • J.D. Edwards
    • Application of the interferometer to gas analysis

    • Chem Metall Eng, 21 (1919), pp. 560–565

    • F. Witte, V. Kaese, H. Haferkamp, E. Switzer, A. Meyer-Lindenberg, C.J. Wirth et al.
    • In vivo corrosion of four magnesium alloys and the associated bone response

    • Biomaterials, 26 (2005), pp. 3557–3563

    •  |   | 
    • G.L. Song
    • Control of biodegradation of biocompatable magnesium alloys

    • Corros Sci, 49 (2007), pp. 1696–1701

    •  |   | 
    • N. Erdmann, A. Bondarenko, M. Hewicker-Trautwein, N. Angrisani, J. Reifenrath, A. Lucas et al.
    • Evaluation of the soft tissue biocompatibility of MgCa0.8 and surgical steel 316L in vivo: a comparative study in rabbits

    • Biomed Eng Online, 9 (2010), p. 63

 

Copyright @ 2013  上海交通大学生物医用金属材料与器械研究组  沪交ICP备20131313

技术支持: