Methods of Thermal Analysis

Michael Feist

Berlin Humboldt University, Institute of Chemistry

1. Conventional Thermal Analysis

2. Simultaneous Thermal Analysis

3. Hyphenated techniques in Thermal Analysis (EGA)

4. PulseTA® and catalysis

5. Determination of other thermal properties

1. Conventional Thermal Analysis

History

Cooling curves

The phase rule

Measuring principles of DTA and DSC

Information content

Sample carriers

Aristoteles (384-322 B.C.)

Fire is the general analysator of matter.

Robert Boyle „The Sceptical Chemist“ (1661):

No, it is not, because it is destructive.

_________________________________________________

Thermal analysis: T = f(t)

∆T = f(t or T)Following the change of a physical property of asubstance subjected to a controlled heating programdepending on time or temperature.

Joseph Black (1728-1799)

Latent heat

Antoine L. Lavoisier (1743-1794)

Mass balance of chemical reactions

Henri-Louis LeChatelier (1850-1936)

Pt - PtRh10 thermocouple for measuring T (1887)

William C. Roberts-Austen (1843-1902)

Differential measuring setup with inert reference (1899)

Josiah Willard Gibbs (1839-1903)

The phase rule F = K - P +1 (for p=const)

In many cases, the temperature-depending change of properties characterize a substance in the same unequivocal manner as can be done by its formula or its structure.

K. Heide (1982)

The subject investigated: A phase or a phase mixture

Changing temperature changes the phase state which yields

Information about purity

state diagrams

material constants

ThermogravimetryTG

Mass

DifferentialScanning Calorimetry

DSC

DifferentialThermal Analysis

DTA

TemperatureHeat flow

Thermosonimetry

Thermooptical AnalysisTOA

ThermomechanicalAnalysis

TMA

ThermodilatometryTD

Other parameterse.g. length

Thermal Analysis

Classical thermal analysis -

Heating and cooling curves of one-component systems

T T

t t

m.p.

melting solidification

F=1

F=0

F=1

second phase appears

F = K - P + 2 - E E = 1 for p=const

F = K - P + 1

Differential measuring setup - Generation of the DTA signal

Asymmetricsignal shape

The information contained in the DTA signal

h = u + pvdh = du + pdv + vdp

du = dq - pdvdq = du + pdv

dh = dq + vdpvdp = 0

dh = dqP

DTA sample holders

Pt or Al2O3crucibles (baker)

Metal block with symmetric holes

Pt or Al2O3 crucibles with „crucible shoes“ as thermocouples

Rapidly ∆T=0 :High resolution

„Slower“ :High sensitivity

Minimal sample mass

Heating of the environment(oven) - sample and measuring system follow passively; Sample and refence areconnected via a gold band which sets ∆T=0

Variable heat flow from separate heaters to maintain ∆T=0

Difference in heating power is the measuring signal

Differential Scanning Calorimetry

2. Simultaneous Thermal Analysis

Heating and cooling (CsFeF4 ·2 H2O)

Repeated heating runs (NaClO4 · H2O)

Gas changes (Coal)

TG

DTG

DTA

2mg

0,4mg/min

4µV

T

-10,76 %

81107

597

606659

676

687

636

638 595

594

100 300 500 700 700 600T / °C

heating cooling

CsFeF4 · 2 H2O

301

54

55150

308

307

303

303305144

4,33%

7,91%

TG

DTA

DTA

TG

n(H2O)

1

0,5

0

1st + 2nd heatingin N2

1st coolingin N2

2 hrs isothermallyin moist air

T

endo

T

TG

NaClO4 · H2O

T / °C

TG-Analysis of coal with changing atmosphere

One TA run -

Four parameterscharacterizinga natural product

Ottaway (1981)

S

SS

O OO

O

O

O

O NN+

N–

N+N

–N

NN+

N– hν, ∆

- 3 N2

S

SS

O OO

O

O

O

O N

N

N

Explosive decompositionof a Tris-azide at 130°C15 mg sample

C24H15S3O7N9 25,55% N

3. Hyphenated Techniques in Thermal Analysis (EGA)

Gas flow in thermobalances

Pressure reduction and unfalsified gas transfer

Coupling systems in TA-MS

Examples

Speed and Flow ⇒ delay in detection ?

Flow profile ⇒ laminar, turbulent ?Distribution of evolved gases inpurge gas stream

Flow direction ⇒ influence of convectionControlled convection in verticalTG systems (chimney effect)

Temperature gradients

Position of transfer line connection

Gas flow in thermobalances -The influence of transport conditions

Pressure reduction

103 mbar → 10-5 mbar

Condensation and secondary reactionsUnfalsified gas transferDifferent viscosity - Demixing

Attribution to a chemical processPyrolysis or Evaporation ?

Fragmentation in MS or sample behavior ?

Three main problems for TA-MS coupling devices

OrificeValve

Throttle

Pump

Capillary

AdapterMS

Capillary Coupled System 2400°C

MS analysator

Overflow

Orifice 2

Orifice 1

Cooling Trap Orifice Coupling System

1500°C

MS analysator

Orifice 2 Skimmer

Orifice 1

Skimmer Coupling System

Overflow

1500°C

NETZSCHSTA 409 C Skimmer®

Amorphous ZrO2 · n H2O → ZrO2 (tetragonal)

-40

-20

0

20

40

DTA /uV

75

80

85

90

95

100

TG /%

100 200 300 400 500 600 700Temperature /°C

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

Ionenstrom *10-9 /Am18 H2O

m44 CO2

m17 OH

TG

DTG

DTA

-4.89

-8.77-0.97

426

↑ Exo

Dehydratation,Dehydroxylation

Squeezing out of residualH2O and CO2 moleculesduring the crystallization

0

0.2

0.4

0.6

0.8

1.0

Ionenstrom *10 -9 /A

50 100 150 200 250Temperatur /°C

60.0

65.0

70.0

75.0

80.0

85.0

90.0

95.0

100.0

TG /%

m18

m28

Santos D2O Bohne(in Ar / 20%O2)

m44

m19

TG

196 °C

196 °C

214 °C

Simulation of the roasting of coffee beans

Corundum platecrucible, Pt netArgon / 20% O2

m/z 18 (H2O+) m/z 28 (CO+, N2+)

m/z 19 (HDO+) m/z 44 (CO2+, ... )

m/z 20 (D2O+, Ar++)

m/z 28 (12CO+, 14N2+)

m/z 29 (13CO+, 14N15N+)

i28 : i29 = 16 : 1

65

70

75

80

85

90

95

100

TG /%

50 100 150 200 250Temperatur /°C

0

1.0

2.0

3.0

4.0

5.0

[4]Ionenstrom *10 -10 /A

m28

m29 (x10)

Santos D2O Bohne(in Ar / 20%O2)

m44TG

196 °C

215 °C

• No discrimination between CO and N2• Bean bursts due to water pressure

• Decarboxylation at higher temperature

m/z 68 (C3H4N2+, ... )

m/z 80 (C6H4N2+, ... )

m/z 109 (coff)

m/z 194 (coffM+)

• MAILLARD products together with CO2• Three roasting stages

• Coffeine partially evaporates

0

0.5

1.0

1.5

Ionenstrom *10 -12 /A

50 100 150 200 250Temperatur /°C

60.0

65.0

70.0

75.0

80.0

85.0

90.0

95.0

100.0

TG /%

Santos D2O Bohne(in Ar / 20%O2)

m80

m68

m109

m194

TG

196 °C

214 °C

4. Pulse Thermal Analysis® and Catalysis

Quantitative evaluation of MS or FT-IR signals

Taylor made redox catalysts

Applications

Maciejewski et al. (1997)

PulseTA® with aNETZSCHSTA 409 C Skimmer®

Injection of known voluminaof one or two gases

Quantitative evaluation ofMS and FT-IR signals

Appropriate m/z⇒ chemical composition

TG ⇒ sorption phenomena

Calibration of MS signals for PulseTA® Maciejewski et al. (1997)

1. On line

2 NaHCO3 → Na2CO3 + CO2 + H2O

2 PdO → Pd + O2

2. Off line

Injectionof a known gas volume into the purge gas

SeparateTA run of a suitable calibration substance

Taylor made catalysts with PulseTA® (1)

Maciejewski et al. (1997)

Reduction of CuO

by H2 pulses at 450°C

Oxidation of Pd/ZrO2by O2 pulses at 100°C

Taylor made catalysts with PulseTA® (2)

Maciejewski et al. (1997)

Reduction of PdO/ZrO2by CH4 pulses

and its subsequent

re-oxidation

by O2 pulses at 500°C

Mass change of H-ZSM-5 zeolite after pulses of 1 ml NH3 at 200°C

Chemisorption and Physisorption studied by PulseTA®

Eigenmann et al. (2000)

The HTB structure of β-CrF3 (hexagonal tungsten bronze)

De Pape et al. (1987)

• (CrF6) octahedra

• Hexagonal channels

• Hosting of small molecules

(NH4)3CrF6β-CrF3 · n NH3

β-CrF3 Menz & Bentrup (1992)

β-CrF3 · m H2O · n NH3

PulseTA Calibration of m19with NaHF2 · n H2O

m/z 20 HF+ H218O+ Ar++

m/z 19 F+

NaHF2 · n H2O → NaF + HF + n H2O

(1) Determination of mH2O by PTA after calibration of m/z 18 with NaHCO3

(2) Calculation of mHF using the residue mass

msample = mNaF + mHF + mH2O (3) Calibration factor:

= 1,21mg

0,236 10 As6⋅ − = 5,127 ⋅ 106 mg/As

F HF( ) = mHF calA cal

( )( )m19

100 200 300 400

0,0

0,5

1,00,12 H2O

=0,149 mgA sAm18=0,188 10

-6

A sAm19=0,236 10-6

I.C.

10-9 [

A ]

T [ °C]

70

80

90

100

in N2

NaF + HF + 0,12H2O

Ton112

endo

.0,12 H2ONaHF2

30,07%

DTA

TG

∆m/m

[%]

m20m19

m18

100 200 300 400 500 600 700

0

1

2

3

I.C. [

10 -9

A]

T [ °C]

92

94

96

98

100

0,05 H2O .

DTG

endo

599

Te 593

. 0,06 NH3β-CrF3

0,27mg

1,25mg

0.40mgDTA

TG

∆m/m

[%]

1,22mg

0,10mg0,21mg

151917

18

5x

in Ar • Water release in threetemperature ranges

• Slight endothermal shiftdue to pyrohydrolysis

• β-CrF3 structure collapsesat 593°C ⇒ NH3 loss

• Major part of NH3 lost onlybetween 589 and 604°C

Quantitative descriptionof all details of the thermalbehavior of a fluoride beingsensitive to hydrolysis

1. Transformation of unwanted CFC's

Chlorofluorocarbon (CFC) ⇒ Hydrofluorocarbon (HFC)

Substitution of refrigerants Recycling of refrigerators

__________________________________________________

Hydrodechlorination Removal of Cl from C-Clby use of H2 Replacing Cl with H

Dehydrochlorination Elimination of HCl

1,1 Dichlorotetrafluoroethane , CF3-CCl2F (CFC-114a)

Nomenclature: 1 1 4 aC-1 H+1 F (Cl = rest)

The HTB structure of β-AlF3 (hexagonal tungsten bronze)

De Pape et al. (1987)

• (AlF6) octahedra

• Hexagonal channels

• Strong LEWIS acid sites

• Hosting of small molecules

Catalytic hydrodechlorination of CFC-114a -

via stepwise or carbene mechanism

CF3-CCl2F + n H2

CF3-CH2FHFC-134a

CF3-CH3HFC-143a

CF3-CHClFHCFC-124

CF3-CF:

- 2 HCl

- HCl- HF

- HCl

CF3-CCl2F R114a

CF3-CHClF R124

CF3-CH2F R134a

CF3-CH3 R143a

Fragment

31 (52) 31 (49) 31 (18) 31 (12) CF+

- - 33 (100) 33 (8) CH2F+

- - 34 (100) 34 (8) CHDF+

35 (18) 35 (8) 35 (100) 35 (8) CD2F+ , 35Cl+

- 51 (52) 51 (20) 51 (3) CHF2+

- 52 (52) 52 (20) - CDF2+

- - - 65 (40) CH3CF2+

- 67 (100) - 67 (40) CHClF+, CHD2CF2+

68 (8) 68 (100) - - CDClF+ , C37ClF+

69 (40) 69 (27) 69 (72) 69 (100) CF3+

83 (5) - 83 (65) 83 (2) CF3CH2+

85 (57) 85 (1) 85 (65) - CF3CD2+ , CClF2+

101 (80) 101 (40) 101 (1) - CCl2F+ , CF3CHF+

- 102 (40) 102 (1) - CF3CDF+, CF3CH2F+

135 (100) - - - CF3CClF+

Discrimination

of HFC‘s

via

characteristic

mass numbers

Why D2 used ?

Why D2 used ? (2)

to distinguish the main reaction (DCl) from side reactions (HCl)

m/z 39 (D37Cl+)

(1) Reaction of the freon with Al-OH groupsCF3-CCl2F + 4 OH- → CO2 + CO + 4 F- + 2 HCl + H2O

(2) Reaction of D2 with neighboured surfacial Al-OH groups

D• … HO D• … DO Pd Al → Pd Al + H D

Side reactions :

Main steps of a PulseTA experimentcharacterizing a Pd/β-AlF3 catalyst

1. Thermal pretreatment in N2 at 300°C

2. Cooling down to 25°C

3. CFC pulse at 25°CMS : m/z in SIM modeTG : Chemi- / Physisorption ?

4. D2 pulse at 25°CTG: Chemisorption: activated D...D

5. Heating in N2 with 10K/min up to 270°C withPulses of CFC and/or D2

MS : Products ?

0 20 40 60 80

Ion

curr

ent /

A

m102 (x5)

m18

m33 (x1000)

Time / min

92

94

96

98

100

T

TG

m101

0

100

200

300

∆m/%

1E-9

T/°C

m34 (x1000)

m39 (x1000)

ß-AlF3/Pd + R114a 5E-11

• Freon pulses only

• m101 (CCl2F+) = CFC-114a

• No educt consumption

• No product formation

• Temporary buoyancy effects (freon density)

• No adsorption

• Freon and D2 pulsedsimultaneously

• First educt pulse:chemisorption of D2

• Physisorption of productsat higher T

• Water as byproduct

0 20 40 60 80

E-10

T/°C

Time / min

Ion

curr

ent /

A

ß-AlF3/Pd + R114a / D2

m102 (x5)

m33 (x10)

m34

92

94

96

98

100

E-9

TG0,24%

∆m/%

(x10)m18

m101

0

100

200

300

m39 (x0.5)

0 20 40 60 80

ß-AlF3/Pd + R114a / D2

Time / min

Ion

curr

ent /

A

E-11m68(x0.1)

m65(x0.5)

92

94

96

98

100

T/°C∆m/%

m20(x0.1)

m4

TG

0

100

200

300

T

0.06mg

5E-10

0 20 40 60 800 20 40 60 80

m52(x0.5)

0 20 40 60 80

• m4 (D2+): educt consumption

• m20 (HF+): byproduct

• m65 (CH3CF+): formation ofhydrogenated products

• m52 (CDF2+): bothHCFC-124 and HFC-134a

⇓m102 (CF3CDF+, CF3CH2F+)remains const and low

Catalytic hydrodechlorination of CFC-114a -

via stepwise or carbene mechanism

CF3-CCl2F + n H2

CF3-CH2FHFC-134a

CF3-CH3HFC-143a

CF3-CHClFHCFC-124

CF3-CF:

- 2 HCl

- HCl- HF

- HCl

0 20 40 60 80

R143a

Time / min

ß-AlF3/Pd + R134a / D2

DFm21

m65

92

96

100

2E-9

0.06mg

R134am33

TG

0

100

200

300

400T/°C

5E-12

∆m/%

0 20 40 60 80

Pulsing the intermediateHFC-134a

• m33 (CH2F+): no consumptn.

• m21 (DF+): no reaction

• m65 (CH3CF+): no formationof HFC-143a

⇓No direct transformationof HFC-134a into 143a -

Confirms the carbene mechanism.

5. Other thermal properties

Heat capacity

Thermal diffusivity

3 step procedure fordetermination of Cpby DSC :Heat flow rate of1. Empty crucibles2. Calibr. substance R3. Sample S

Höhne, Hemminger, Flammersheim (1996)

β⋅−=−==∆ )(d

dd

d- RSRRSSRSSR CCtTC

tTCΦΦΦ

β - average heating rate(different heating ratesof sample and reference)

Thermal diffusivity - The Laser Flash Method (1)

ρλ

pctT =

∂∂

q = - λ grad THeat flow

Thermal diffusivity

T = T(x,y,z)λ - thermal conductivity

cp - heat capacityρ - density

Temperatureresponse

Sample diskLaser flash

T = f(t)

Thermal diffusivity - The Laser Flash Method (2)

The t1/2 method

2/1

21388.0

tda ⋅=

Thermal diffusivity - The Laser Flash Method (3)