Molecular Dynamics Simulation of Silicone Oil: Degradation upon Oscillatory Testing

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study “Molecular Dynamics Simulation of Silicone Oil: Degradation upon Oscillatory Testing” by Pascal Puhlmann and Dirk Zahn reports on the theoretical study of polysiloxanes upon oscillatory compression/decompression runs. The authors investigate chain scission, reassembly and cyclization mechanisms from ns-scale molecular dynamics simulations and they compare the degradation reactions of linear and cyclic silicone chains. Overall, the study is good and it can be interesting for the journal readership, however, some improvement should be done. My comments / suggestions are as follows:

1. The choice of Reactive Force Field method for this study should be explained in more detains in the Introduction section.
2. Why was D5 rings chosen for a model fluid?
3. In the lines 126–127 the authors state: “The ageing of silicone oils (fortunately) is a rather slow process, and the underlying time scales drastically exceed the ns scale inherent to our molecular dynamics simulations.”. How long does it take for silicone to age (this should be added in the manuscript)? What can increase the ageing process of polysiloxanes?
4. A possible practical application of the study should be emphasized in the Introduction and Conclusion sections.
5. In the lines 268–270 the authors state “Upon heating up to 1000 K and vaporization of the products, it is however quite intuitive to expect increased production of the entropically favored small-molecule species such as cyclotri- and cyclotetrasiloxanes, respectively.” The references should be added.

Author Response

comment: The choice of Reactive Force Field method for this study should be explained in more detains in the Introduction section.

reply: our model was chosen consistent to previous simulation work on silicone formation. This is now described more clearly at the end of the introduction: "Using simulation models of hundreds of silicone oil molecules, we offer unbiased comparison of Si-O bond dissociation, formation and concerted reorganization ranging from scission and fusion of linear PDMS strands to cyclization processes. For this purpose, we take use of a recently presented reactive modelling approach to describing silicone oil formation [20]. To this end, both formation and degradation of silicone oil are treated within analogous simulation setups."

comment: Why was D5 rings chosen for a model fluid?

reply: D5 is a common base oil used in cosmetics etc. This is now stated in the methods section: "In the present work, we adopt these models and compare this to a fluid of 100 cyclopentasiloxane species, namely 5 membered rings of dimethylsilicone (D5) - which is often used as a base oil in cosmetics and hair care applications. "

comment: In the lines 126–127 the authors state: “The ageing of silicone oils (fortunately) is a rather slow process, and the underlying time scales drastically exceed the ns scale inherent to our molecular dynamics simulations.”. How long does it take for silicone to age (this should be added in the manuscript)? What can increase the ageing process of polysiloxanes?

reply: This is now described in more detail. The new text reads: "The ageing of silicone oils (fortunately) is a rather slow process, and the underlying time scales drastically exceed the ns scale inherent to our molecular dynamics simulations. We are bound to rather drastic testing setups as compared to the established experimental procedures - which may induce silicone oil aging by irradiation or heating to several hundred °C in hundreds to thousands of hours  [6,7,8,9,14,15]."

comment:  A possible practical application of the study should be emphasized in the Introduction and Conclusion sections.

reply: We gratefully pick up this suggestions. An additional paragraph was added to the conclusion section: "We argue that our molecular simulation setup offers an additional perspective to understanding the ageing of silicone oil formulations - and thus provide a step forward to rationally designing oil formulations with tailor-made properties. For example, model extension to exploring the role of additives or contaminations is straight-forward. Likewise, different temperature/pressure conditions could be contrasted to provide molecular scale insights into the resulting ageing products. "

comment: In the lines 268–270 the authors state “Upon heating up to 1000 K and vaporization of the products, it is however quite intuitive to expect increased production of the entropically favored small-molecule species such as cyclotri- and cyclotetrasiloxanes, respectively.” The references should be added.

reply: We now explain the thermodynamic consideration in more detail: "Upon heating up to T=1000 K and vaporization of the products, it is however quite intuitive to expect increased production of the entropically favored small-molecule species (since entropy S scales with the logarithm of the number of freely mobile molecules). To this end, the formation of cyclotri- and cyclotetrasiloxanes, respectively, arguably stems from the extreme temperatures applied as the entropy gain may compensate for the less favorable formation energy of these molecules compared to D5. "

 

Reviewer 2 Report

Comments and Suggestions for Authors

The paper "Molecular Dynamics Simulation of Silicone Oil: Degradation upon Oscillatory Testing"reports molecular dynamics simulations on the silicone oils behavior upon oscillatory testing. The information reported here is of interest to silicone chemistry and  industry research, but certain aspects related to the silicones in the simulated conditions are not fully presented.

Abstract: - First sentence is ambiguous, it's unclear which varieties of silicone oil formulations are considered in the study. Silicones are among the most stable polymers, therefore a rapid degradation in oscillation tests according to the scenarios presented is difficult to demonstrate in practice due to the unstable radical formation and condensation reactions which may also involve the silicone ring formation!

Introduction section - Some of the properties of silicones are well addressed, but their origin from the particularities of the silicon atom are not mentioned, precisely these must be added!

Figure 1. The silicone radical/ cation presented here are stable? I suggest to be mentioned the information related to their stability. Also the references referring to the radicals should be added to the Figure 1 legend.

-page 2, line 50- what does it mean "Si-O bond reorganization"

-page 2, line 72 - what "a variety of cyclization processes" refers to?

-Section 2- the abbreviations QM/MM must be explained

-Section 3- Figure 2- images must be enlarged and the details explained!

-Does the composition of the formulation influence the mode of radical production and cyclization?

-page 4, line 129- briefly the experimental procedures at references [6,7,8,9,14,15] must be described!

-Conclusions- reference [20] must be inserted in section 2. The same observation for line 263, the references there should be moved to section 3.

Based on these observations my recommendation for this paper is Major Revision.

 

 

 

Author Response

comment: Abstract: - First sentence is ambiguous, it's unclear which varieties of silicone oil formulations are considered in the study.

reply: we rephrased the first sentence to "The fate of a selection of linear and cyclic silicone oil formulations in heavy-duty fluid dampers is studied from molecular dynamics simulations. "

comment: Introduction section - Some of the properties of silicones are well addressed, but their origin from the particularities of the silicon atom are not mentioned, precisely these must be added!

reply: While avoiding a length review on silicone, we now added some more details regarding the robustness of Si bonding at the beginning of the introduction: "Silicones oils are important alternatives to hydrocarbon-bases lubricants with silicones outperforming alkanes in terms of chemical stability, bio-compability and fluidity behavior under high pressure [1,2,3,4,5]. Much of this is owed to the great stability of the Si-O bonds as compared to C-C bonding in alkanes. This robustness is particularly appealing for heavy-duty application such as lubrication at high temperature or for damping fluids exposed to extreme numbers of testing cycles."

comment: Figure 1. The silicone radical/ cation presented here are stable? I suggest to be mentioned the information related to their stability. Also the references referring to the radicals should be added to the Figure 1 legend.

reply: Indeed, this point deserves clarification. We now provide a more detailed caption to figure 1: "Illustration of silicone degradation via i) Si-C bond and ii) Si-O bond cleavage. Both of the illustrated intermediates are rather metastable and subject to i) radical reactions yielding methane, ethene etc. (which can be traced by gas sensors) and ii) addition to adjacent silicone moieties leading to the reorganization of the network of -Si-O-Si- bonds. "

comment: page 2, line 50- what does it mean "Si-O bond reorganization"

reply: Si-O bond reorganization according to mechanism ii) is now explained more clearly throughout the whole manuscript. See also fig.1 and related text.

comment: page 2, line 72 - what "a variety of cyclization processes" refers to? 

reply: we meant to express the possibility that cyclization may lead to differently sized rings. Anyway, we agree that there is only one conceptual processs of linear->cycle reoragnization mechanism and thus call this simply "cyclization".

comment: Section 2- the abbreviations QM/MM must be explained

reply: done.

comment: Section 3- Figure 2- images must be enlarged and the details explained!

reply: the image size & resolution is quite strong in our submitted version, implementation issues must be removed my the editorial staff. In turn, the caption of figure 2 is now extended to: "Upper panel: exemplary series of snapshots from mechanical testing of the silicone oil formulation derived from 90/10 DMS D/TMS. Lower panel: schematics of the alternating compression/elongation cycles applied to the simulation cell. Upon completion of subsequent half-cycles a linear silicone strand experiences coiling/elongation moves that eventually provoke Si-O bond reorganization in favor of extruding a cyclopentasiloxane species. Note that the overall number of bonds is unchanged and the coordination of Si and O atoms remains balanced. Yet the reaction energy is not exactly zero, but (slightly) endothermic because of the intramolecular strain inhered to the ring formation. For better visibility only Si (yellow/cyan) and O (red) atoms are shown, whilst the remaining oil molecules are illustrated in grey, respectively. "

comment:  Does the composition of the formulation influence the mode of radical production and cyclization?

reply: Yes, our work reports on how the different formulations lead to different size distribution of the cyclic siloxanes produced during ageing - see figs 6 and 7 for the underlying statistics. This is now pointed out more clearly in the captions of figs. 6+7. Radical formation was not observed in our cyclic testing setup (running at 300 K and without irradiation or oxidative stress).

comment: page 4, line 129- briefly the experimental procedures at references [6,7,8,9,14,15] must be described!

reply: to maintain the flow of our introductory text, this is now implemented very briefly in the main manuscript - and in th ecaption of figure 1.

comment: Conclusions- reference [20] must be inserted in section 2. The same observation for line 263, the references there should be moved to section 3.

reply: ref. 20 is cited many times, including section 2.

 

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors provided detailed responses and very good arguments for the suggestions made. The paper can be accepted for publication in this form.