This seems a good reference, from what I can determine: https://www.thermometricscorp.com/typek.html
TYPE K THERMOCOUPLE (Chromel / Alumel)
(US) ■ YELLOW jacket: Positive leg, Chromel, is approximately 90% nickel, 10 chromium.
(US) ■ RED jacket: Negative leg, Alumel, is approximately 95% nickel, 2% aluminum, 2% manganese and 1% silicon. MAGNETIC due to the nickel content.
(US color coding is ANSI and is not universal across all geographies and industries)
Thermocouple wire is not copper wire, and there is a matching type of thermocouple wire for each type of thermocouple. Type K thermocouple wire is two-conductor (for a single thermocouple) and the each of the two conductors is composed of different metallurgy, to match the thermocouple's native composition.
Introducing extra cold junctions opens the potential for those junctions to introduce errors in the signal that cannot be compensated by calibration, especially if there is a temperature gradient across the pair of junctions.
There are a couple of design approaches to extending TC signals. The TC loop (ie the length of the wire run x 2) should not exceed a total resistance of 100 ohms. Thermocouple wire is manufactured in a couple of different grades so you if you want to have a long TC signal wire run you will want to know if you're running standard TC wire, Special Limits TC wire, or Extension TC wire (to name a few) and do the calcs. Generally, Extension grade TC wire has a much lower working temperature range as well.
Best design practice is to either keep the wire run short, or as soon as practical you can convert the signal from millivolt (thermocouple signal) to 4-20ma loop via a converter transmitter. If that isn't practical, using shielded TC wire is a good design element, because when working with millivolt signalling, EMI can be a significant factor.
Besides using TC wire to extend a run, there can be other reasons that you may need to cut/splice/join TC wiring. Sometimes, you need a detachable point in the wiring for serviceability. In this case, quick-disconnects that feature Chomel/Alumel terminals and contacts are widely available.
However, what if you don't want/desire to have a quick disconnect? Maybe you need to extend the TC wiring, cut & splice for repair. Another approach is to use a TC terminal block, either surface-mount or DIN rail mount.
Omega BS16A Thermocouple terminal strip
Omega XBTKK25 DIN rail terminal block
One topic is especially contentious: splicing thermocouple (TC) wires. It's difficult to find good information on a plain butt connector for inline splicing of TC wire, but it does exist. Military and aerospace people have this stuff down.
AC 21-99 Aircraft Wiring and Bonding ("THERMOCOUPLE WIRE SOLDERING AND INSTALLATION", Section 2, Chapter 16, begins at page 351)
Advisory-Circular-AC-021c99-AIRCRAFT-WIRING-BONDING-.pdf]AC 21-99 Aircraft Wiring and Bonding, Sect. 2 Chap. 16 Para. 31 Materials wrote:
Qty 1 D-436-133-01 Chromel Splice, Colour Coded Gray.
Qty 1 D-436-133-02 Alumel Splice, Colour Coded Green.
Qty 2 D-436-133-03 Splice Sealing Sleeves.
Those butt connectors sort of translate to TE Connectivity numbers.
AMP has been part of E Connectivity, since 1999, and this product line is their "Strato-Therm" terminals.
The Type K Positive/Chromel/Yellow (Gray) butt connector is AMP/TE Conn. 1-322325-1 (OnlineComponents.com, $4.26 ea, min. qty. (10))
The Type K Negative/Alumel/Red (Green) butt connector is AMP/TE Conn. 1-322325-0 (OnlineComponents.com, $2.32 ea, min. qty. (10))
I just ordered (10) of each for $80 after shipping and tax.
It seems very common for people to try to solder, use twist connectors, or standard (dissimilar metals) screw terminal blocks to splice thermocouple wiring. From what I have read, the most-correct method is to use TC wire butt connectors whose metallurgy matches the TC wires to be spliced. This method avoids creating extra cold junctions in the TC wiring circuit.