Advanced Heat Transfer and

Thermal Storage Fluids

Dan Blake, Luc Moens, Dan
Rudnicki, Mary Jane Hale,
Professor Ramana Reddy,
and Greg Glatzmaier

Goals and Objectives

The goal is to find a heat transfer and
thermal storage fluid with a usable liquid
range from near 0 to above 400 C that will
meet the cost and performance requirements
of parabolic trough systems.
The near term objective (FY03) is to
identify a fluid with the potential for service
up to 300 C.

Project to Date
Synthesis and thermal stability of
imidazolium salts and identification of other
possible salts NREL FY2000-present
Physical properties, materials compatibility,
and extended thermal stability testing The
University of Alabama FY2001-present

Production Process and Cost Study - Peak

Design Subcontract FY02

The Challenge for A New Fluid

Low freezing point - <25 C

Low vapor pressure - < 1 atm at Tmax
Low cost - < ~$4.50/Kg
Thermal stability at > 400 C
Compatible with alloys and materials used
in solar plants
Other physical properties compatible with
the solar application

Imidazolium Salts

R4

R3

R4

R5

N
R5

N
R1

R2
X

N
N
R1

R3
R2

Methyl
Ethyl
Butyl
Hexyl
Octyl
Phenyl
Silyl

Cl
OS O 2CH3

B F4
P F6

Table 1. Cost of reactants for synthesis of EmimBF4

Reactant

Cost
($/lb)

Cost per lb of
EmimBF4

Glyoxal (50%)

1.00

.73

(ethylene glycol)

(0.38)

(0.12)

Formaldehyde

.21

.03

Ammonia

.10

.01

Methylamine

.73

.11

Ethylchloride

2.00

.64

(ethylene)

(0.23)

(0.03)

Tetrafluoroboric acid
(50%)

.65

.58

Process Changes to reduce the cost of an imidazolium salt

Peak Design (2002)
H

OH

Variation 2

OH

Me

NH2

Et

Cl

Me

Me

Cl
Et

3H2O

Et

Et

HBF4

N
Me

Cl

Me

HCl

Me

O
NH3

partial
oxidation

Cl

Variation 1

BF4
Et

HCl

Production process for ethylmethylimidazolium

Tetrafluoroborate (process 3)
Air

Ethylene
Glycol
Storage

Tube Reactor 5
Glyoxal

Ethylene
Storage

Reactor 1

Formaldehyde
Storage

Ethyl Chloride

Reactor 2

HCl

Reactor 4
HCl/Solvent
Solvent

Ammonia
Storage

EmimCl

Tetrafluoroboric
Acid Storage

Distillation
Tower

Reactor 3
EmimBF4
Storage

Methylamine
Storage
EmimBF4

Table 3. Economic assumptions for the discount cash flow analysis

Peak Designs (2002)

Economic Factor
Internal Rate of Return
Depreciation
Recovery Period
Plant Life
Construction Period
Working Capital
Federal Tax
State Tax
Salvage Value
Debt/Equity Ratio
Plant size
Annual hours of operation
Labor
Maintenance
Electricity
Reaction yields

Value
15%
Tab. 4
11 years
21 years
1 year
15% of total capital
40% of net income
5% of net income
10% of fixed capital
0/100
10,000,000 kg/yr
8,322 (95%)
$50/hr (loaded including supervision)
7% of fixed capital
$0.06/kWhr
100% (except glyoxal synthesis: 75%)

Table 7. Required costs for EmimBF4

Process
1
2
3

Reactant costs
4.75
3.30
2.02

Other operating
0.24
0.25
0.25

Capital
0.37
0.37
0.39

TotalCost ($/kg)
5.36
3.92
2.66

Table 8. Product cost dependence on plant size and internal rate of return
Process
1
2
3

Process
1
2
3

Plant Size (106 kg/yr)

5
10
50
5
10
50
5
10
50
IRR (%)
10
15
25
10
15
25
10
15
25

Product Cost ($/kg)

5.57
5.36
5.16
4.16
3.92
3.70
2.90
2.66
2.43
Product

Cost ($/kg)
5.24
5.36
5.62
3.80
3.92
4.19
2.53
2.66
2.92

Table 9. Cost of EmimBF4 as a function

of reaction yield (process 3)
Reaction Yields*

100%

Total cost ($/kg)

$2.66

95

2.86

90

3.11

80

3.85

*Glyoxal yield is held at 75% in all cases

Summary and Conclusions

Cost goals can be met with an organic salt
When a viable candidate is identified the
production process must be optimized

Longer term stability test of

[C8mim]PF6
400

300

Wt.%

T = 235.9oC

60

/ hr.

Weight (%)

T = 285.4oC

350

65
-5.3

80

T = 338.1oC

100

250
200

40
150

T = 134.2oC
20

100

0
0

U of Alabama, 2001

500

1000

1500

2000

2500

3000

Time (min)

3500

4000

4500

Temperature ( C)

Constant Wt.%

50
5000

Overall Observations
We may have pushed the upper temperature
limit of imidazolium salts as high as
possible with anions Further
improvements may require changes in the
imidazolium ring substituents
A 300 C salt appears to be within reach
We will begin some work on alternative
types of organic salts and mixtures

THE UNIVERSITY OF ALABAMA

Corrosivity of Ionic Liquids

THE UNIVERSITY OF ALABAMA

Experimental Methods

Tafel extrapolation method was used to determine corrosion

current density and calculate corrosion rates
Potentiodynamic polarization curves to analyze corrosion
behavior of the alloys in ionic liquids
Experiments were performed at room temperature
Alloys tested:
- 1018 carbon steel
- 316 stainless steel
- gray cast iron

THE UNIVERSITY OF ALABAMA

Typical Tafel Plot

0.2

Potential vs. SCE (Volt)

0.1

3.27 10 3 i corr W
r=
d

Type 316 Stainless Steel

o
in [C4mim][Tf2N], 25 C

0.0

-0.1
Ecorr = -0.138 V
-7

icorr = 1.10x10 A/cm

where:
r = corrosion rate (mm/yr)
W = equivalent weight (g)
d = alloy density (g/cm3)
icorr = corrosion current
density (A/cm2)

-0.2

-0.3
10-10

10-9

10-8

10-7

10-6
2

Current Density (A/cm )

10-5

10-4

THE UNIVERSITY OF ALABAMA

Uniform Corrosion Rates of

Solar Materials in Ionic Liquids
Corrosion Rate at 25C, in m/yr

Ionic liquid

SS 316

Gray Cast Iron

[C8mim]PF6

1.3
LC
1.2

1018 Carbon
Steel
3.2
LC
5.6

[C6mim]PF6

0.4

13.0

13.0

[C4mim][Tf2N]

1.1

11.0

20.0

[C4mim]Cl

LC = localized corrosion

8.9
LC
2.6

THE UNIVERSITY OF ALABAMA

Conclusions

Materials used in solar plant technology such as 316

stainless steel alloy, 1018 carbon steel and gray cast iron
were found to be outstanding in corrosion resistance
(corrosion rates: <20 m/yr) against ionic liquids at RT
Localized corrosion was observed on the surface of the
materials exposed to [C4mim]Cl ionic liquid
The potentiodynamic tests showed in most cases an
active/passive corrosion behavior. However, the ionic liquid
[C4mim]Cl prevented the formation of stable passive films

THE UNIVERSITY OF ALABAMA

Heat Capacities

THE UNIVERSITY OF ALABAMA

Properties of Heat Transfer Fluids at 25C

Liquid

Melting
point, oC

Density
g/ml

Viscosity
cPs

Cp
Jg1K-1

Cp, v
Jcm-3K-1

Dowtherm HT

NA

1.01

953

1.42

1.41

Thermal oil

NA

0.89

1.9

1.69

1.90

[C4mim]Cl

57.1

NA

1.58

>1.58

[C2mim][BF4]

5.8

1.20

34

1.12

1.34

[C2mim][PF6]

60.5

1.10

NA

1.00

1.10

[C4mim][PF6]

6.5

1.37

389

1.14

1.56

[C6mim][PF6]

-80(Tg)

1.30

688

1.34

1.75

[C4mim][Tf2N]

-5.1

1.44

53

1.05

1.50

THE UNIVERSITY OF ALABAMA

A comparison of Volumetric Heat Capacity

Performance

Dowtherm MX
Syltherm XLT
Dowtherm G
[C4mim][NTf2]

2.5

[C4mim][PF6]
[C6mim][PF6]

-3

cv / Jcm K

-1

3.0

2.0
1.5
1.0
-50

50 100 150 200 250 300 350

o

Temperature / C

THE UNIVERSITY OF ALABAMA

Viscosities of Ionic Liquids

THE UNIVERSITY OF ALABAMA

Viscosities of Ionic Liquids

2.2

2.4

BMIC

HMIC
2.2

Log 10 (Viscosity/cp)

Log 10 (Viscosity/cp)

2.0

1.8

1.6

1.4

1.2
0.35

2.0

1.8

1.6

0.40

0.45

0.50

0.55

0.60

0.65

X (AlCl3)

C4mimCl at 300K

0.70

1.4
0.40

0.45

0.50

0.55

0.60

X (AlCl3)

C6mimCl at 300K

0.65

THE UNIVERSITY OF ALABAMA

Long Term Thermal Stability of IL

THE UNIVERSITY OF ALABAMA

Evaluation of Long-term Thermal Stability

Experimental Procedure :
Hold the ionic liquid samples at a fixed temperature for
20 hours,
Then raise the temperature to a higher level and holding
for another 20 hours until significant weight changes are
observed.

THE UNIVERSITY OF ALABAMA

Weight Changes of [C4mim][Tf2N] as a Function

of Time at Different Temperatures
400
-0.0068 Wt.% / hr.

100

-0.1062 Wt.% / hr.

T = 332.5oC

60

T = 133.7oC

40

200

r
/h

Weight (%)

.%
Wt
22
91
-3.

T = 235.3oC

100
20

0
0

500

1000

1500

2000

Time (min)

2500

3000

3500

Temperature (oC)

300

80

THE UNIVERSITY OF ALABAMA

Weight Changes of [C6mim]PF6 as a Function of

Time at Different Temperatures
400
-0.0373 Wt.% / hr.
-0.5
092
Wt.
%/

90

350
hr.

300

T = 284.4oC

Weight (%)

80

250

T = 235.3oC
70

200

60

150

T = 133.9oC

50

100

40

50
0

500

1000

1500

2000

Time (min)

2500

3000

3500

Temperature (oC)

-0.0048 Wt.% / hr.

100

THE UNIVERSITY OF ALABAMA

Weight Changes of [C8mim]PF6 as a Function of

Time at Different Temperatures
400
350

65
-5.3

300

T = 285.4 C

Wt.%

T = 235.9 C

60

/ hr.

Weight (%)

80

T = 338.1 C

250
200

40
150

T = 134.2 C
20

100

0
0

500

1000

1500

2000

2500

3000

Time (min)

3500

4000

4500

Temperature ( C)

Constant Wt.%

100

50
5000

Influence of ANION
120

100

Weight (%)

80

N
60

R2

N
R1

40

20

1,3-dimethyl imidazolium Iodide

1,3-dimethyl imidazolium DBS
1,3-dimethylimidazolium chloride
1,3-dimethylimidazolium PF6
1,3-dimethylimidazolium BF4
1,3-dimethylimidazolium OMs
1,3-dimethylimidazolium NTf2
1-bu-3-methylimid NTf2

100

200

300

Temperature (C)

400

500

600
Universal V2.3C TA Instruments

Influence of ANION
120

100

O
O S CF 3
O

Weight (%)

80

60

Et

O
O S CH3
O

N
Me

40

20

1-ethyl-3-methylimidazolium OMs

1-ethyl-3-methylimidazolium OTf

100

200

300

Temperature (C)

400

500

600
Universal V2.3C TA Instruments

Influence of CATION structure

120

100
Me
Me

Weight (%)

80

60

Me

N
Me

Me

Me
P F6

N
Me

40

Me
Me

20

P F6

Me
P F6

N
Me

1,2,3,4,5-pentamethylimid PF6

1,3-dimethylimidazolium PF6

1,3,4,5-tetramethylimid PF6

100

200

300

Temperature (C)

400

500

600
Universal V2.3C TA Instruments

Ramp 20C/min and Isothermal decomposition over 120 min

120

199.97C
99.08%

100

Weight (%)

80

60

N
Me

40

299.63C
58.67%

Bu
N

S O2 CF 3
S O2 CF 3

20

375.46C
0.9011%

100

200

300

Temperature (C)

400

500

600
Universal V2.3C TA Instruments

Schematic of NRELs MBMS

Sampling System
Products
and
Transients

Three Stage Free-Jet

Molecular Beam Source

Reactants

Q1
Reactor
with heat
or photons
or catalyst

Q2

Q3

Collisions

e{P+}

{D+}

EI
Source

Turbomolecular
Pump
Turbomolecular
Molecular
Pump
Drag
Pump
National Renewable Energy Laboratory

Triple Quadrupole
Mass Analyzer

Detector

Argon
Collision
Gas

Turbomolecular
Pump

Thermal Decomposition Pathways

Demethylation
N
+
N
Me

MeX

Hofmann Type Elimination

H

N
+
N
Me

N
X

N
Me

HX

MBMS study of [EtMeIm][BF 4 ]

120

Et

82
100

80

BF 4

N
Me

60

MW = 198

40

96

28
20

110
42
55

68

0
0

20

40

60

80

100

120

140

160

180

200

m/z

Et

Et

N
Me

m/z = 82

Proposed thermal events

Me

m/z = 96

m/z = 110

1) HF liberation at high T
2) de-alkylation of quat. amine salt

One-Step Synthesis of Ionic Liquids

Bu
N
CHO

glyoxal

CHO

Me NH2

PF6

Me
Me
alkylamines

BuNH2

one-pot
synthesis

PF6

N
Me

HCHO

formaldehyde

HPF6

(aq. acid)

(mp= 91C)

Bu
N
N

PF6

Bu
* avoids chloride salts as intermediates !
* 'adjustment' of melting point and thermal stability

T stability of Ionic Liquid Mixture from One-Step Synthesis

120

100

Weight (%)

80

60

40

20

Mixture
1,3-dibutylimidazolium PF6
1,3-dimethylimidazolium PF6
1-butyl-3-methylimidazolium PF6

100

200

300

Temperature (C)

400

500

600
Universal V2.3C TA Instruments

Conclusions
Imidazolium salts offer flexibility in designing melting
point and thermal stability
PF6 salts are easy to prepare and purify, and are probably
least expensive
Onset T for thermal decomposition must be determined
carefully. Kinetic data are needed for evaluation of longterm T stability
Reactivity of ANION has strong influence on T stability
Structure of imidazolium CATION appears to have less
influence, but more work is needed
Influence of IMPURITIES on T stability of ionic liquids is
not completely understood and is case-dependent.
Intermediate chloride salts in synthetic route must be
avoided due to residues in final ionic fluid (corrosion)
One-Step Synthesis of ionic liquid mixtures:
a) no chloride residues
b) lower complexity of synthetic process for ionic
fluids
c) possibly lower production cost
d) synthetic methods must be optimized

Can alternative ionic liquids be developed other than the

imidazolium salts ?