Home Energy Physics Nuclear Power Electricity Climate Change Lighting Control Contacts Links



By Charles Rhodes, P.Eng., Ph.D.

The term Molten Salt Reactors (MSR) is used to refer to several different types of molten salt cooled reactor. There are theoretical molten salt "paper reactors" containing fixed geometry fuel in fuel tubes within a coolant salt bath such as is contemplated in the following Moltex Video and the Moltex related Ph.D. Thesis and there are molten salt reactors with fuel dissolved in the molten salt coolant, commonly referred to as liquid fuel reactors, a prototype of which was actually built and operated 1965 - 1969 and then summarized by Moir-Teller in 2004.

The most important motivation for pursuit of molten salt reactors is for implementation of the fertile Th-232 into fissile U-233 breeding nuclear fuel cycle. When Th-232 absorbs a neutron it becomes Pa-233 which naturally decays into U-233 after about a month. However, the Pa-233 has a high neutron absorption cross section. If the Pa-233 remains embedded in the Th-232 and is permitted to absorb another neutron to become Pa-234 and the loss of this neutron makes the breeding fuel cycle no longer sustainable. In order to realize sustainable conversion of Th-232 into U-233 it is necessary to continuously extract the Pa-233 from the Th-232. The practical way to achieve this extraction is to have Th-232 in the blanket of the reactor in the form of molten ThF4 or ThCl4 and to continuously circulate that molten blanket salt through a Pa-233 extraction apparatus.

Thus practical realization of Th-232 to U-233 breeding requires molten salt reactors.

A laboratory prototype molten salt reactor (MSR) with liquid fuel and fluoride salt was tested in the 1960s. However, it turns out that although MSRs are simple in concept there are major material corrosion and degradation challenges in their practical implementation.

The purported advantages of molten salt cooled reactors as compared to liquid sodium cooled reactors are:
1) Higher operating temperature than liquid sodium cooled reactors;
2) Much greater fire resistance than liquid sodium cooled reactors;
3) Ability to support the fertile Th-232 to fissile U-233 fuel breeding cycle.

Recently it has become apparent that liquid sodium cooled reactors fitted with molybdenum fuel tubes depleted in Mo-95 can potentially operate at the same high temperatures as molten salt reactors without the attendant material corrosion problems.

A major disadvantages of molten salt reactors are:
1) The high melting point of the salt, which has numerous practical reactor maintenance implications.
2) Numerous material problems that lead to enclosure, heat exchanger and fuel tube corrosion;
3) High sensitivity to trace impurities in the molten salt.

Thus the major motivation for use of molten salt reactors is implementation of the Th-232 - U-233 breeding cycle. This fuel cycle is particularly important in China and India which both have major thorium deposits but comparatively little uranium.

If a reactor with molten fluoride salt coolant is to be operated without fuel breeding the salt component ThF4 should be replaced by ZrF4. An excess of Zr in the salt is helpful in corrosion mitigation.

In the USA there are major political obstacles to thorium based molten salt reactor (MSR) development and deployment. These obstacles are well summarized in the video MSR Political Problems

Many MSR design versions have been proposed involving both thermal neutron and fast neutron spectrums. In a 2020 video Kirk Sorensen reduced the thermal spectrum MSR design versions to a single common theme of thermal neutron breeding of Th-232 into U-233, a graphite moderator, < 700 degree C operation and BeF2-LiF salt coolant, of which the Li must be mono-isotopic Li-7.

Thermal neutron MSRs have a power constraint related to pumped liquid fuel residence time in the moderator which arises from use of delayed neutrons for power control.

Fast spectrum MSRs have comparable constraints. They generally require fuel tubes to confine the nuclear fuel to a central region of the coolant to prevent the fast neutrons impinging on the enclosure walls and the intermediate heat exchange bundles. Fast neutron spectrum MSRs generally must operate at < 700 degrees C and usually rely on KCl-LiCl salt coolant of which the Li must be monoisotopic Li-7 and the Cl must be monoisotopic Cl-37. Note that in a fast spectrum reactor Cl is preferred over F due to its higher atomic weight.

A practical constraint on fast spectrum MSRs is that the fuel must be highly enriched due to its lower neutron absorption cross section for fast neutrons as compared to thermal neutrons. Another practical reactor design constraint in a reactor supporting a sustainable fuel cycle is that the fuel geometry must be adjustable to compensate for fuel aging.


The liquid fuel molten salt reactor designs come in two versions. One version is fluoride salt based and is usually intended for operation with thermal neutrons. The other version is chloride salt based and is usually intended for operation with fast neutrons. In recent years both types of MSRs have received theoretical attention. A thermal neutron liquid fuel MSR was prototyped during the early 1960s. However, to the best of this author's knowledge, no one has ever actually built a large fast neutron MSR.

This author emphasizes that in reactor design safety and material quality control must take priority over theoretical reactor economics. Thus many "paper reactor" designs must be tempered by both reactor safety and by material availability considerations. The material considerations are particularly relevant to specified isotopic materials such as Cl-37, Li-7, Mo-96 to Mo-98, etc. which can only be obtained via difficult and expensive isotope separations.

One contemplated molten chloride fast neutron reactor uses a NaCl - UCl3 - CaCl2 salt mixture. The chlorine is highly enriched in Cl-37. One of the issues with this reactor design is neutron absorption by Cl-35 forming Cl-36 which has a half life of 300,000 years and conversion of Ca-40 into Ca-41 which has a half life of 80,000 years. This salt mixture, which is highly water soluble, poses very challenging long term waste disposal issues. The contemplated mechanical configuration is a Tube-In-Shell Gen IV Reactor.

A molten salt fast neutron reactor likely needs its fuel salt contained in fuel tubes to keep the fuel salt geometry fixed during an earthquake. A large fast neuton reactor not having fuel tubes is subject to fast neutron degradation of the enclosure and heat exchanger and might potentialy go prompt critical due to unpredictable fuell salt wave motion during an earthquake.

A thermal neutron molten salt reactor typically uses a LiF - UF3 - XF fuel salt mixture. The lithium is highly enriched in Li-7. Lithium is used to achieve a relatively low salt melting point. However, one of the issues with lithium is conversion of Li-6 into tritium. That conversion also releases highly chemically corrosive unbonded fluorine. Another issue is that typically X = Be, which has severe cost, chemical toxicity and neutron induced corrosion issues. In a thermal neutron MSR the major issues are corrosion and limited operating life of the graphite moderator.

In molten salt breeder reactor designs different salt mixtures are used in the core and blanket regions which are separated by a neutron transparent pressure barrier, such as described by LeBlanc. Typically in thermal neutron reactors the blanket region contains Th-232 while the core region operates using U-233. Side arm chemistry is used to immediately remove Pa-233 (half life = 27 days) from the blanket salt, age it for several half lives to allow Pa-233 to decay to U-233 and then to inject it into the reactor core. The Pa-233 cannot be left in the blanket because it absorbs neutrons. Other side arm chemistry is needed to separate U-233 from fission products in the reasctor core region. Eventually too much U-232 accumulates in the reactor core and the core fuel must be converted into FNR fuel for potential energy recovery and disposal. To achieve sustainable energy production surplus fissile atoms produced by the FNR must be used to make new MSR start fuel.

There are a legion of potential corrosion problems in Molten Salt Reactors, some of which are discussed in the following video and in the paper titled: Corrosion in the molten fluoride and chloride salts and materials development for nuclear applications.

An important mechanism for corrosion minimization is electrolysis which is used to change the oxidation state of uranium salt. For example, if a chloride salt is used there is side arm chemistry which is constantly removing chlorine from the salt so that when active chlorine is released as a result of a fission the active chlorine is immediately absorbed by UCl4 which converts to U3Cl16. Then electrolysis in a side arm apparatus is used to extract the chlorine to restore the salt to UCl4.

A problem with application of this corrosion suppression technique is that this technique relies on continuous salt circulation. That circulation does not exist inside fuel tubes.

The corrosion resistant alloy 617 is defined by alloy 617 composition specification and by alloy 617 composition specification.

In 2019 ASME approved alloy 617 for use in high temperature nuclear reactors. The physical properties of alloy 617 are available from the following 2015 report. However, as of 2019 there were still uncertainties with respect to MSR enclosure materials and component integrity.

Heat transport and heat exchange issues related to Molten Salt Reactors are set out in the report titled: Process Heat Exchanger Options for Fluoride Salt High Temperature Reactor.

The state of the art in thermal neutron MSRs is reviewed in the paper titled: Recent Research of Thorium Molten Salt Reactors.

Some of the theoretical economic advantages of MSRs are set out in the paper: Market Basis for Salt-Cooled Reactors.

Some of the theoretical technical advantages of MSRs are set out in the paper: Molten Salt Reactors (MSRs): Coupling Spent Fuel Processing and Actinide Burning

A company named ThorCon contemplates assembly of molten salt reactors using modular ship building technology and then towing the reactor to the end user site. This is an ambitious plan that is more concerned about reactor first cost than about fuel sustainability. ThorCon relies on selection of nuclear power plant sites that have sufficient natural heat sink water circulation for waste heat removal but are not subject to large tsunamis. ThorCon contemplates replacing the reactor can every four years due to graphite moderator degradation. Each reactor can must spend yet another four years on the reactor site before removal to allow fission product decay. Thus in practice every such reactor in the field requires three cans. ThorCon has yet to demonstrate a prototype. The initial capital cost is significant due in part to the various regulatory approvals and special purpose ships that are required.

The ThorCon MSR implementation plan is described in the ThorCon Video. The ThorCon heat source is a basic liquid fuel MSR but it is not an advanced reactor.

The historical development of MSRs during the 1950s and 1960s is described in molten salt reactor historical documents.

More recent development of MSRs is described in Review of R & D of thorium molten salt reactor

The major feature of molten salt reactors is that they:
a) provide a practical means for obtaining fission energy from thorium, which is several times more abundant than uranium, as described at Recent Research of Thorium Molten-Salt Reactor from a Sustainability Viewpoint;
b) They provide heat at a higher temperature (750 degrees C) than liquid sodium cooled reactors (500 degrees C) or water cooled reactors (320 degrees C).

However, many practical MSR chemical and material corrosion issues remain to be resolved. The main interest has been in a subset of MSRs known as LFTRs which can theoretically operate in a fuel sustaining mode by thermal neutron breeding of Th-232 into U-233. A major LFTR constraint is physical deterioration of the graphite moderator which limits the graphite working life to about 4 years. Other constraints are C-14 accumulation and the complexity of the required on-going salt chemistry. A lead proponent of MSR technology is Kirk Sorensen of Flibe Energy Inc. in Huntsville, Alabama. (Flibe stands for the MSR salt components: F Li & Be)

There are other companies such as Elysium and Moltex that contemplate operating MSRs with a fast neutron spectrum which avoids use of graphite, minimizes fuel chemistry issues and improves fuel burnup. However, use of fast neutrons causes rapid containment material deterioration and potentially enables prompt critical related safety problems related to the geometric instability of liquid fuels. If there is no moderator the required fuel enrichment level is very high.

A review of MSR technology from the UK perspective as of 2015 can be found at: EPD MSR Review Feasibility Study July 2015 1.02

A typical Molten Salt Reactor promotional presentation is the one by Leslie Dewan. A company promoting inexpensive limited life MSR technology is ThorCon.

A strong argument in favor of Molten Salt Fast Chloride Reactors (MSFCRs) is presented at Why MSFCRs are the Best Form of GenIV Reactor

1. Due to the reduced requirement for cooling water a MSR can be sited much further above the local water table and surrounding bodies of water than the siting of a direct water cooled CANDU reactor, thus enhancing system safety in rare but severe events such as earthquakes, tsunamis, hurricanes, meteorite strikes, etc.

2. The primary coolant of a MSR operates at a relatively low pressure which simplifies many reactor design, construction, operation and maintenance issues. However, if fast spectrum neutrons are used the MSR designer must take into account possible severe pressure transients caused by the rapid molten salt thermal expansion required to suppress local reactivity excursions into prompt neutron criticality;

3. Unlike water cooled and moderated reactors MSRs do not produce high pressure radioactive steam. Hence MSR designs do not need to be concerned about radioactive steam containment. If the turbine steam pressure becomes too high that steam can be directly vented to the atmosphere because it is not radio active.

4. A near term opportunity for MSRs lies in the chemical and hydrocarbon fuel reformation industries, but those industries need a high fossil carbon emissions tax to make conversion to MSRs financially viable.

5. An issue is that the lithium component of the molten salt should be mono-isotopic Li-7 because Li-6 absorbs neutrons and forms tritium.

6. A further issue is the absorption of neutrons by Li-7 which converts it into He-4 leaving highly corrosive fluorine atoms free to attack both the containment wall and the heat exchange tubes. To minimize this corrosion surplus Li metal must be continuously added to both the core salt and the blanket salt. The analogous problem with Be is that neutron absorption by Be converts it into He and hydrogen isotopes that release further corrosive unbonded fluorine or HF.

7. An extensive list of molten salt reactor technical references is contained in the: SAMOFAR Report.

The theoretical advantages of liquid fuel molten salt reactors (MSRs) over a rigid fuel liquid sodium cooled FNRs are:
1. A liquid fuel MSR with good neutron conservation can be fuelled by thorium provided that the start fuel contains sufficient enriched uranium.
2. A liquid fuel MSR delivers heat to the thermal load at a high temperature (750 degrees C). This high temperature heat is of particular importance in production of ammonia and in reforming of hydrocarbons to make energy dense hydrocarbon fuels;
3. The high operating temperature of a liquid fuel MSR potentially reduces the cooling water volume requirement per kWh for electricity generation which in turn reduces impact on marine ecology;
4. The molten salt is much less of a fire hazard than is liquid sodium;
5. A liquid fuel MSR avoids the hundreds of thousands of metal fuel tubes that are required by a rigid fuel power reactor;
6. In a liquid fuel MSR the fuel is relatively uniform and hence the reactor is theoretically less complex to control than a rigid fuel reactor as long as there are no liquid level gradiants, waves or vorticies. However, in real high power liquid fuel MSR the issues of potential wave and vortex formation in the liquid fuel is of extreme importance.

The disadvantages of a liquid fuel MSR as compared to a rigid fuel liquid sodium cooled FNR are:
1. The higher melting point temperature of the salt as compared to sodium triggers difficult metallurgy and corrosion problems;
2. Salts that have lower melting points generally involve beryllium, lithium or chlorine that have major isotope separation and/or nuclear waste disposal issues;
3. Neutron absorption by beryllium and lithium causes formation of tritium and He-4. Then companion fluorine or chlorine atoms in the salt aggressively corrode metal heat exchange tubing and containment vessel walls. This corrosion limits both liquid fuel MSR and rigid fuel MSR metal component working life;
4. The molten salt containment vessel walls and intermediate heat exchange tubes should beisolated by a thick molten salt guard band layer in which neutrons are not released to prevent cumulative fast neutron damage to the containment vessel walls and the intermediate heat exchange tubes. This fast neutron shielding requirement increases the reactor size for a specific thermal power output and forces use of a core zone moderator. Then, as discussed below, the required liquid fuel residence time within the moderator limits the reactor thermal performance;
5. The intermediate heat exchange tubes have only a short working life due to formation of low thermal conductivity fission product deposits on the heat exchange tube surfaces. These heat exchange tubes are a major source of maintenance and decommissioning waste;
6. The side arm fuel recycling chemistry of a liquid fuel thermal neutron MSR is complex and causes serious operational problems and costs for distributed reactors. Typically molten salt reactor designs minimize the chemistry problems by using two different salts, one for maintaining the nuclear reaction and the other breeding. The two different salts are separated by a metal wall with a low neutron absorption cross section;
7. The start fuel supply situation for liquid fuel thorium reactors (LFTRs) is difficult because LFTRs cannot breed new fuel fast enough to rapidly expand the reactor fleet.
8. The required shutdown frequency for containment vessel, intermediate heat exchanger and moderator replacement is much higher for a liquid fuel MSR than for a rigid fuel liquid sodium cooled FNR. In a practical nuclear power station two fully redundant MSR assembies are required to manage this issue, each with a projected working life of 4 to 6 years.
9. The high temperature heat produced by a liquid fuel MSR can potentially be replaced by electricity supplied by rigid fuel liquid sodium cooled FNRs. This potential for high temperature heat replacement caps the allowable capital cost of liquid fuel MSRs.
10. All fission reactors rely on < 1% delayed neutrons for power control. Normally a reactor operates with > 99% prompt neutrons but relies on delayed neutrons for power control. In a liquid fuel thermal neutron MSR the fuel is in motion and must have a residency time of longer than 3 seconds in the core region of the reactor (in proximity to the graphite moderator) for thermal power control stability. This residency time requirement limits the molten salt flow rate and hence the reactor thermal power output. The full implications of this liquid fuel flow rate limit are seldom adequately appreciated by liquid fuel MSR proponents.
11. If in order to operate at a higher power the liquid fuel residency time in the core zone is reduced the reactor shifts toward a dangerous prompt critical state. If the residency time is increased to minimize this problem the molten salt differential temperature rises, which raises the maximum molten salt temperature, which increases the corrosion rate of the metal components and introduces severe thermal stress issues.
12. In a MSR some gas space is required to keep inert gas fission products at a low pressure. If the gas space over the liquid salt is small then in the event of an earthquake with a 3 g horizontal acceleration component the tangential stress in the molten salt containment vessel walls will become very large. Accomodation of this wall stress while simultaneously thermally insulating the wall sets a practical maximum on the containment vessel diameter.

If there is a large gas space over the molten salt liquid fuel the potential for unforeseen fuel geometry changes increases as the diameter of a liquid fuel MSR containment vessel increases. If an unforeseen event such as a hydraulic surface oscillation, vortex, accident or earthquake occurs which quickly changes the liquid fuel geometry from being 99.2% to 100+% critical on prompt neutrons the reactor could potentially explode due to internal pressure waves caused by sudden thermal expansion of the molten salt.

Thus the maximum size of liquid fuel fast neutron MSRs may be limited by containment vessel fabrication issues related to making the reactor earthquake tolerant.
13. In a thermal neutron MSR a provision must be made for molten salt transfer into a dump tank and then back again to allow practical replacement of the containment vessel, intermediate heat exchanger and the graphite moderator.
14. A liquid fuel MSR dilutes its fissile fuel with chloride and/or fluoride salts. This dilution generally means that the fissile fuel must be highly enriched to maintain sufficient core zone reactivity. It may also mean that Th-232 - U-233 fueled reactors must be supported by U-238 - Pu-239 fueled reactoe to achieve ongoing conversion of the fertile fuels U-238 andTh-232 into the fissiloe fuels Pu-239 and U-233.
15. A liquid fuel thermal neutron MSR relies on a graphite or zirconium hydride moderator to achieve core zone reactivity. Hence realizing the hard fast neutron spectrum required for consuming all the transuranium actinides in a thermal neutron reactor is impossible. Thus liquid fuel thermal neutron MSRs generally must rely on external liquid sodium cooled FNRs for disposal of transuranium actinides.
16. The practical complexity of chemically dealing with the spectrum of corrosive and radioactive fission products mixed with the salt in a MSR should not be under estimated.

1. MSRs have persistent chemical corrosion problems that can be mitigated but for which there are presently few good long term solutions. Due to corrosion and fast neutron degradation MSRs need frequent major component replacement. Hence liquid sodium cooled FNRs will likely be more economic than MSRs for public electricity generation.

2. It may be that MSRs will not be cost competitive in the foreseeable future because equivalent high temperatures can be attained by generating electricity with a liquid sodium cooled FNR and then using that electricity to make high temperature heat.

An experimental thermal-neutron MSR operated for >17,000 hours until Dec. 1969. The engineering/chemistry documentation is at:
Experimental MSR Design Information

One point to note is that in the experimental MSR the molten salt pumping rate was such that a significant fraction of the delayed neutrons typically appeared outside the graphite moderator core. A MSR designer must aware of this issue and avoid any reactor flow geometry that might lead to consequent prompt neutron criticality.

Today the Chinese have a vigorous MSR program aimed at development of Th-232 - U-233 breeder reactors. This program is summarized in the video:
Chinese MSR Development Progream as of 2012.

A rigid fuel MSR is conceptually similar to a liquid sodium cooled FNR except that the liquid sodium primary coolant is replaced by a molten salt. The main advantages of a molten salt cooled rigid fuel reactor over a liquid sodium cooled FNR are a reduced fire hazard and a higher operating temperature. However, those advantages are offset by the disadvantages of more complex intermediate heat exchanger and containment vessel metallurgy. There are also problems with neutron absorption by the molten salt. For example, neutron absorption by Li-6 causes formation of H-3 and He-4. Then the corresponding Cl or F ion in the salt corrodes reactor metal components. This problem causes containment and heat exchanger material corrosion and hence frequent major equipment replacement.

A representative solid fuel molten salt reactor design is a Pebble-Bed Fluoride-Salt-Cooled High-Temperature Reactor. A fundamental problem with this reactor design is that the nuclear fuel waste is almost impossible to reprocess and is unsuitable for fuel breeding.

Liquid fuel molten salt fast neutron reactors theoretically offer a multiplicity of fuel processing advantages. They preferentially use chloride rather than fluoride salts due to reduced neutron moderation by the higher atomic weight chlorine nucleus. However, the fast neutrons rapidly destroy containment vessel and heat exchange tube materials, so at this time liquid fuel molten salt fast neuton reactors are just "paper reactors". A relevant video is Elysium Molten Salt Fast Chloride Reactor. Another potential complication in large liquid fuel molten salt fast neutron power reactors is maintenance of reactivity uniformity through the reactor core zone.

Various claims have been made relating to future liquid fuel power reactors. However, an emerging safety issue is prompt neutron criticality in large liquid fuel power reactors. One facet of this issue is that in an earthquake or other accident circumstance fissionable molten salt liquid fuels are inherently unstable and may rapidly change their fuel geometry and hence their local reactivity. A rapid change in local reactivity as small as 0.5% in a fast neutron MSR might cause local prompt neutron criticality and hence a large molten salt thermal expansion pressure pulse sufficient to cause an enclosure wall failure. This problem can be avoided by use of fuel tubes with plenums which normally maintain fuel geometry but which permit rapid fuel expansion to suppress a prompt critical condition. It is likely that fissionable liquid fuel fast neutron reactors will be shown to be only safe for use in relatively small rigid containment vessels in which the wall stress related to a transient thermal expansion pressure pulse is manageable.

Fast neutron MSR designs use mono-isotopic Cl-37 to achieve a faster neutron spectrum than is possible with fluorine which helps in consuming transuranium actinides. However, that performance advantage is offset by the more complex corrosion control chemistry related to use of chlorine and by the cost of separating the isotopes Cl-35 and Cl-37.

Due to material problems liquid fuel molten salt fast neutron reactors are presently just "paper reactors". Such reactors have never been successfully built and operated due to lack of suitable materials. However, there are promotional videos which address the theoretical fuel advantages of the molten salt fast neutron reactors if the material and pump seal problems can be solved. These videos include:
Elysium Industries Molten Chloride Fast Reactor by Ed Pheil at Thorium Energy Alliance Conference 8 (TEAC 8)
Elysium Industries Molten Chloride Fast Reactor by Ed Pheil at Thorium Energy Alliance Conference 10 (TEAC 10)
Ed Pheil and Carl Perez (co-founders of Elysium Industries) engaged in an October 2019 discussion related to the Elysium Technologies reactor.

The peak wall stress issue is much less severe in a thermal neutron liquid fuel MSR than in a fast neutron liquid fuel MSR, in part due to localization of the nuclear reaction within the graphite moderator of a liquid fuel thermal neutron MSR.

An advantage of a molten salt liquid fuel thermal neutron reactor over a molten salt liquid fuel fast neutron reactor is that nuclear reactions only occur in the proximity of the central graphite moderator. Hence the reactor can be designed with a wide guard band of molten salt in whch no nuclear reactions occur. This guard band protects the containment vessel walls and the heat exchange tubes from neutron damage. However, a major problem with this type of reactor is graphite deterioration over time, which can potentially cause the reactivity to become dangerously unstable.

Molten salt thermal neutron reactors generally use fluoride salts to achieve lower melting points and to avoid formation of the long lived radio isotope Cl-36.

A liquid fuel thermal neutron reactor generally relies on the presence of a graphite moderator in the core zone for criticality. Then the nuclear reaction can be stopped by transferring the liquid fuel to a companion dump tank which contains no moderator. In an absence of station power situation this automatic transfer can be accomplished using a freeze plug and gravity drainage. The dump tanks must have sufficient heat transfer capability to remove fission product decay heat.

Comparison of FNRs and LFTRs

Within the category of liquid fuel molten salt thermal neutron reactors is the important subcategory of liquid fuel thorium fluoride reactors (LFTRs). These reactors can breed Th-232 into U-233 with thermal neutrons. The basic concepts of a power LFTR are set out in ORNL-3996. To make the system neutronics work it is necessary to promptly chemically extract the intermediate element protactinium isotope (Pa-233), which has a large thermal neutron absorption cross section, as soon as it is formed. This chemical extraction is simplified by use of separate blanket and core salt fluids. The reactor core contains U-233 or U-235 dissolved in salt. There should be a thin low neutron absorption barrier between the reactor core and the reactor blanket. The reactor blanket contains Th-232 dissolved in molten salt.

The extracted protactinium (Pa-233) decays with a half life of 27 days into fissile U-233. The U-233 can then be inserted into the reactor core. The U-233 generally contains a sufficient fraction of U-232 which makes this fissile material unsuitable for weapon production. However, U-232 decays emitting a hard gamma photons which require extra radiation shielding.

Another parallel running process must be used to continuously extract neutron absorbing fission products from the reactor core salt. One of the important fission products is Xenon-135 which has a large thermal neutron cross section.

Generally the LFTR core contains graphite which acts as a moderator and hence reduces the neutron kinetic energy and confines the neutron flux to the proximity of the moderator. The moderator minimizes high energy neutron impingement on the containment walls and on the heat exchange tubes. However, the projected graphite working life is only 4 to 7 years.

There must also be provisions for circulating the core salt through an external heat exchanger for thermal energy extraction. There is a limit on that circulation rate imposed by the required residency time in the graphite moderator. Otherwise there is a risk of prompt criticality.

Liquid Fuel Thorium Reactors (LFTRs) offer the potential benefit of an almost unlimited future fuel supply. However, LFTRs have complications related to neutron conservation, corrosion, fuel chemistry, size limitations and long lived waste generation. In principle the waste generation complications can be circumvented by use of companion uranium fuelled Fast Neutron Reactors (FNRs).

A liquid fuel MSR has flowing liquid fuel rather than fixed position solid fuel. The average reactivity is controlled by the fissile fuel concentration in the molten salt. A liquid fuel MSR usually relies on thermal expansion of the molten salt to give the reactor a negative slope power versus temperature reactivity characteristic. The reliance on delayed neutrons for power stability places an additional contraint on liquid fuel MSR design. If the fuel atom residence time in the moderated reactor core zone is not much greater than the delayed neutron arrival time (approximately 3 seconds) the fraction of delayed neutrons available for thermal power regulation substantially decreases causing a liquid fuel MSR to operate much closer to prompt neutron criticality than does a liquid sodium cooled FNR or a molten salt solid fuel FNR.

Consequently at high liquid fuel flow rates a liquid fuel MSR is sensitive to small changes in core zone fuel geometry. A reactivity increase can cause the fission power to rapidly grow before the reactivity control system has time to suppress the power rise. This issue is of major concern because in a liquid fuel MSR the fraction of delayed neutrons participating in criticality can be reduced to dangerously low levels in an effort to increase the reactor's power output for economic reasons. Thermal expansion of the liquid salt improves the liquid fuel MSR power stability by reducing the fissile fuel density in the core zone with increasing temperature. However, generally reactivity control rods are still required for safety shutdown purposes.

Haynes Alloys brochure.

Here are two possible Haynes alloys for MSR's HAYNES HR-160 and HAYNES-556.

Most MSR corrosion mitigation relies on changing the oxidation states of uranium. Uranium forms several different oxides including UO2, U3O8 and UO3. Depending on its oxidation state the uranium atom offers 4, 5 or 6 chemical bonds to other atoms such as oxygen, chlorine or fluorine. Normal operation of a MSR releases active chlorine or fluorine atoms which will corrode the metal enclosure, heat excchanger, piping and pump components, etc unless these active chlorine or fluorine atoms are immediately captured by a suitable component of the molten salt. The strategy usually used to minimize MSR corrosion is to attempt to keep the uranium in the molten salt in its 4 bond valence state so that when nuclear reactions cause release of active chlorine or fluorine atoms these atoms immediately chemically bond with uranium atoms by causing the uranium atoms to jump into their 5 or 6 bond states. Since release of active chlorine or fluorine atoms is an ongoing processs a MSR needs to have ongoing side arm electrochemistry that constantly removes chlorine or fluorine from the molten salt to return the uranium to its 4 bond state.

Even so, this is a corrosion mitigation process, not a corrosion prevention process. This method of corrosiion mitigation relies on the molten salt being fully mixed at all times. If an inside corner of the enclosure has little salt flow it may become locally depleted in 4 bond uranium, in which case there is nothing to prevent active chlorine or fluorine attacking the enclosure at that location.

In the blanket salt there may be little or no uranium in which case another metal with multiple oxidation states needs to be added to the blanket salt.

The physical viability of small liquid fuel MSRs was demonstrated at the Oak Ridge National Laboratory (ORNL) during the 1960s, as shown in the Molten Salt Reactor Experiment Video.

The apparent advantages of a liquid fuel MSRs and LFTRs are set out in the videos:
Kirk Sorensen in Canada, chemistry of molten fluoride salt blanket treatment
Kirk Sorensen-A Global Alternative (thorium energy via LFTR) TEAC 4
Kirk Sorensen-The Promise of Thorium in Meeting Future World Energy Demand
Short video outlining the promise of liquid fuel MSRs
Long video outlining the promise of liquid fuel MSRs
Molten Salt Liquid Fuel Reactor Videos
The Thorium Problem
Thorium 2017
Thorium 232 - From History to Reactor[2019] including the required ongoing side arm chemistry
LFTR Chemical Processing & Power Conversion
Liquid Fluoride Thorium Reactor
Kirk Sorensen - Flibe Energy LFTR Development Strategy at ThEC 13
ThorCon: Cheap Reliable CO2 - Free Electricity by Lars Jorgensen at ThEC2018 relating to Indonesia
ThorCon's Thorium Converter Reactor 2019-10 Update by Dane Wilson at TEAC 10
Thorium Remix (Thorium Energy Alliance meeting presentations)

However, major issues not addressed in the above videos are power stability, economics, corrosion, ongoing maintenance, inability to breed new fuel fast enough to expand the reactor fleet and nuclear waste disposal.

A UCB 2012 YouTube video in which Kun Chen from the Chinese Academy of Sciences describes the Chinese Molten Salt Reactor development program

Further Thorium MSR information:
Energy From Thorium
Thorium Energy Alliance Proceedings

The molten chloride fast reactor uses a NaCl - UCl3 - CaCl2 salt mixture. The chlorine is enriched in Cl-37.

The thermal (neutron) molten salt reactor used a LiF - UF3 - XF salt mixture. The lithium is enriched in Li-7.

In theory the molten salt could be NaF, NaCl, LiF or LiCl. However, use of Li requires separation of Li-6 from Li-7 because the Li-6 absorbs neutrons to form H-3 and He-4. Use of Cl requires separation of Cl-35 from Cl-37 because Cl-35 absorbs neutrons to form Cl-36 which is a nuclear waste disposal nightmare due to its long half life of 308,000 years. From a nuclear waste perspective the most suitable salt is NaF. However, NaF has a very high melting point of 993 degrees C. At the melting point of NaF there is rapid thermal expansion. A practical NaF based MSR would have to operate with a maximum molten salt temperature of about 1200 degrees C. This temperature is so high that it triggers a host of material problems.

Other salts such as CsF have been considered for use in liquid fuel MSRs. However, for MSRs fuelled by thorium, which breeds to U-233, the selection of salts is further constrained by neutron absorption by the salt. Fissioning of U-233 yields only slightly more than two neutrons per fission. There is very little margin for neutron absorption in a fuel conserving breeding cycle.

LiF/BeF2 (FLiBe)
A molten salt mixture currently receiving attention for possible use in MSRs is LiF/BeF2, which in a suitable ratio is claimed to have a melting point in the range 360 C to 459 C and to have the best neutronic properties of all the fluoride salt combinations that are appropriate for reactor use. The Li component should be monoisotopic Li-7.

If the Li-6 is not removed from LiF/BeF2 the molten salt will likely absorb too many neutrons to support thermal neutron fission breeding and will produce unwanted tritium.

The LiF/BeF2 is superficially chemically stable. However, when Li and Be absorb neutrons they change into He-4 and H-3. The companion F atoms in the salt mixture then become aggressively corrosive to the reactor intermediate heat exchange tubes and metal walls.

A major problem with LiF/BeF2 is that the stable isotope Be-9 on neutron absorption can become Be-10 which is a beta emitter with a half life of 2.5 million years. Once formed Be-10 needs to be safely isolated from the environment for at least 25 million years. The isotope Be-10 could easily become one of the most persistent environmental toxin problems.

Relevant papers are:
NT Fission Fusion 2019
NT Heat Pipe Salt Reactors 2019

An under appreciated issue relating to liquid fuel MSRs is the complexity of the side arm radio chemistry. This chemistry must on a continuous basis:
a) Separate protactinium from the blanket zone salt;
b) Separate fission products from the core zone salt including safe capture of xenon, krypton and other inert gases;
c) Remove and safely capture tritium resulting from neutron absorption by Li-6;
d) Electrolytically remove small amouts of fluorine and/or chlorine to restore uranium to its four bond valence state to minimize corrosion;
e) Add U-233 fuel as required to maintain core criticality;
f) Add caesium (Cs) or a like substance to preferentially absorb fluorine released during neutron absorption by the molten salt;
g) Selectively remove the CsF or a like fluoride resulting from fluorine capture;
h) Provide an electronic means of monitoring the state of the liquid fuel MSR;
i) Control the fissile fuel concentration and hence the reactor reactivity;

Automatic implementation of this complex radio chemistry may be more difficult than implementation of the entire balance of the reactor. It might turn out that the entire concept of distributed liquid fuel MSRs economically fails due to the complexity of the required side arm radio chemistry.

The following text was provided by John Rudesill in a December 28, 2020 enail.

While we are discussing the merits of molten salts and liquid metal sodium coolants and in particular their corrosive potentials, I caution that there are serious environmental issues associated with wide scale use of fluoride salts especially when including the metals Li, and Be. We need to be mindful that in our zeal to eliminate GHG emissions that we don't excuse the increased traffic in toxic elements. Fluorine is highly toxic and fortunately for biology it is inertly bound in minerals like fluorspar CaF2 and phosphate rock. A consequence of mining and processing phosphate rock for fertilizer is the release of large amounts of fluorine as SiF4 and H2SiF6.

There was a time when these toxic byproducts were simply vented or dumped into waste water with little concern for the consequences. The phosphate extraction is carried out with strong sulfuric acid often made on site with its attendant SOx emissions. The workers in the phosphate processing areas often had perforated septums due to inhaling the F laden gases. Environmental concerns made capture of these by-products necessary and disposal was not the preferred option. Conveniently, the assertion that toothe decay could be prevented with fluoride treatments succeeded in creating one new outlet for these orphaned fluoride wastes. Fluoro silicic acid, H2SiF6 is used for disinfectants in breweries, etc. It travels in RR tank cars. These wastes no doubt go down the sewer and to the sewage treatment plant ending up in either the sludge or the outfall into natural water ways. The sludge is often burned, land filled, or used as soil enhancement on farm land resulting in fluoride contamination/pollution of the environment.

A significant amount of fluoride is used in the forms of natural and synthetic cryolite Na3AlF6 necessary for the Hall-Heroult electrothermic aluminum smelter cells. The molten cryolite easily dissolves powdered Al2O3 that is normally very difficult to dissolve directly in aqueous acids. This is a caution about alloys using Al to protect the surface from corrosion that molten fluoride salts will likely have no problem removing that otherwise protective Al2O3 coating. The Al reduction reaction occurs at 950-980 C right at the upper limit of proposed HT MSR's. The smelting process discards spent cryolite to "disposal". Some of which is used as crop pesticide e.g. on grape plants as it is an excellent remedy for very destructive leaf hoppers. The cryolite leaches into the soil and the plants will take some of it into their tissues including the grapes which we consume as wine or table grapes. Repeated applications of this toxic pesticide have to be causing a build up in the soil. The levels in the grape juice are high enough to inhibit yeast fermentation if the level exceeds the limit of ~1ppm F. Grape ranchers would like to use higher doses of cryolite, but the wine makers cannot tolerate the inhibiting effect on the yeast such is the toxicity of F. In animals including humans, F competes with I inhibiting functional thyroid hormone production leading to slowed metabolism and immune function.

If we embrace fluoride based molten salt technology we are increasing the extraction, movement, and disposal of toxic fluorides in the environment. This needs to be thought through very carefully as environmental activist will eventually catch on to this and we need to have credible responses.

As for Li and Be, both have toxicity problems. Small amounts of Li are essential for good health and it is used in treatment of bi-polar disorders. Too much Li is cumulative and destroys kidney function and this can happen fairly quickly. Be can cause berylliosis especially from inhaling dust or fumes. Some people seem to have an inborn sensitivity to Be and are affected worse than others.

Chemical processing industry tenets obligate chemical engineers to design processes with thorough "cradle to grave" accounting for all inputs and outputs. Proposed chemical processes have to be submitted for review and approval or rejection by state and federal EPA's. The permitting process can take years if not thought through ahead of time with counsel of permitting agencies. This needs to be done early during the development phases not later! Usually, conditional permits can be approved for development phase "proof of concept studies". However, one must be careful to avoid the use of the term "pilot plant" as the requirements suddenly become much more onerous and restrictive. "Pilot Plant" connotes intent to produce for at least trial marketing. What I have seen done is to refer to small up scale from laboratory bench processes in the gram scale to what are called "kilo" processes. Permitting agencies are more flexible with this terminology. I expect NRC has ChE's involved in the licensing process who can review chemical processing aspects. If not, then respective EPA's need to be engaged.

John Rudesill

This web page last updated January 1, 2021.

Home Energy Physics Nuclear Power Electricity Climate Change Lighting Control Contacts Links