I don't know your barrel lengths, but Scottish ballistician Geoffry Kolbe has a pretty good chapter on the .22 Long Rifle. He says it starts to lose velocity in a barrel at about 19". The powder charge is about 1.2 grains, according to Kolbe. Assuming similar powder constituent compounds, a .223 charged with 24 grains AR2210 makes about 20 times more gas. Add in that you have about 4 times greater peak pressure in the smaller expansion ratio .223 chambering, and it makes for somewhat disproportionately higher muzzle pressure.
At then end of a 20" barrel .22 Long Rifle will have a muzzle pressure of about 320 psi, because the expansion ratio is high and the gas has cooled a lot by the time it gets there. A 50 grain bullet .223 load of 24 grains of AR2210 or a similar powder from a 20" barrel will have a muzzle pressure of about 7900 psi, or 25 times higher than the .22 LR. If you expand .22 LR muzzle pressure to a 2" diameter piece of PVC, the pressure will be just below 40 psi, which is typical max domestic water pressure in the U.S. If you put that same 2" tube over the muzzle of the .223, you get about 1,000 psi. Clearly above the pay grade of the PVC.
In addition, as mentioned, the gases are hotter. Since smokeless powder doesn't contain enough oxygen for truly complete combustion, when hot fuel gases exit the bore and mix with the oxygen in air, a secondary gas explosion often takes place, appearing as a fireball and raising the pressure further. So I don't think there's much mystery as to why this apparatus blew up?
If you want to make a .223 suppressor in the same size, you need something heat resistant that is about 25 times stronger than the PVC was. Consider large diameter hydraulic tubing rated for the expected pressure.
Also, be aware that while you can suppress the muzzle blast, you can't do anything about the sonic crack of a supersonic bullet. So this isn't going to be much quieter than an unsuppressed .22 LR rifle unless you go to subsonic squib loads.